Radio Boulevard
Western Historic Radio Museum

Vintage Longwave Receivers
Restoration and Performance Testing

Part 2 - WWII LW Receivers

9. Mackay Radio & Telegraph Co. - RC-123 Coast Guard Receiver

10. USN-RCA-Andrea Radio Co. - RAK-7 CND-46155

11.  US Army Signal Corps - BC-314-D

12.  RCA-Victor Div. of RCA - CR-91 Receiver

(1944 Camden-build)
Includes NDB Reception Log

13.  USN-National Co. RBL-5

14. Radiomarine Corp. - AR-8510

15.   USN RBA Receivers
CFT-46154 and CFT-46300

Includes Rebuilding Information and NDB reception log for RBA-1 SN:972
Description and performance for RBA-6 SN:181
Photos of RBA-1 with Field Mod installed, Photos of RBA-2 SN:168

Mackay Radio & Telegraph Co. - Type 3001-A Marine Radio Receiver


photo: Shipboard Radio Room in the late-twenties to early thirties - note the vacuum tube transmitter in the right corner of the operator's desk. The photo is from an October 1932 issue of Radio News showing the radio room of the ship "Discovery II."





Mackay Radio & Telegraph Company

for the

U.S. Coast Guard

Radio Receiver Type RC-123   sn: 97

CONTRACT: TCG-34199   ORDER: CG-80609

built by: Federal Telegraph Co.

15kc to 635kc

TRF Stage, Regeneration Amp, Detector, Audio Stages


photo above
: The Federal Telegraph Company building in Palo Alto around 1927

Mackay Radio & Telegraph Company, Federal Telegraph Company and Federal Telephone & Radio Company - Mackay Radio & Telegraph Company was founded by Clarence Mackay, son of John W. Mackay, one of the "Big Four of the Comstock" fame in Virginia City, Nevada. John Mackay made his wealth in Comstock silver but in 1883 he partnered with J. Gordon Bennett and started in the wire telegraph communications business. Mackay-Bennett started Commercial Cable Co. for trans-Atlantic wire telegraph, although the Mackay-Bennett system used a visual indication code rather than an aural system. Soon, Postal Telegraph was created to compete with Western Union (Postal used the standard American Morse code.) When John Mackay died, he was in the process of laying the trans-Pacific cable for service to Asia (1902.) Clarence inherited the business, finished the trans-Pacific cable (1904) and, in 1925, added radio to the world-wide communications system. In 1928, Clarence sold most of his business holdings to International Telephone & Telegraph Company (ITT.) Mackay remained on as head of Mackay Radio and Telegraph Company until his death in the late-thirties.

Federal Telegraph started out in Palo Alto, California mainly dealing in arc transmitters. At one time, Lee DeForest worked for the company but Frederick Kolster was the chief engineer for most of FTC's history. FTC bought the Brandes Company and created a new division of FTC called Kolster Radio Company specifically for selling consumer radios in the mid-twenties. FTC became involved with Mackay Radio in 1926 when Mackay bought a radio station that had belonged to FTC. When Mackay sold his interests to ITT, Federal Telegraph Company continued to do most of the Mackay Radio work under contract to ITT and eventually, in 1930, ITT bought Federal Telegraph. ITT moved the Federal Telegraph operation to Newark, New Jersey in 1931. For awhile ITT tried the consumer radio market with a division called Kolster International but it was a short-lived venture. The name of Federal Telegraph Co. was changed to Federal Telephone and Radio Company during the early days of WWII (about 1942.) After WWII, Mackay Radio equipment usually indicates "built by Mackay Radio & Telegraph Company." Since ITT owned both Federal Telephone and Radio Company and Mackay Radio, it's assumed that ITT believed showing both companies as involved with the production was superfluous information. Later, Mackay Radio became ITT-Mackay Marine.

General Information Radio Receiver Type RC-123 - The Mackay RC-123 is a six-tube, regenerative receiver that tunes from 15kc up to 635kc in four tuning ranges. A single TRF stage is coupled to the detector through a Regeneration tube, an amplifier for the feedback that allows better stability in the receiver sensitivity. Two audio stages are used. The power to operate the receiver was supplied by ship's DC power for B+ (115vdc) and batteries for the tube heaters. It was also possible to operate the receiver entirely on batteries in which case dry cells were used for the B+ requirements. It was also possible to operate the receiver entirely on 115vac. To operate the tube heaters on AC, a small 115vac to 6.3vac transformer was mounted inside the cabinet and could be connected into the circuit using the terminal strip located in the rear of the cabinet. If the RC-123 is operated on AC, an external 1:1 isolation transformer should be used since a power transformer is not part of the B+ circuitry. The tube heaters are connected in parallel and the two dial lamps are also part of this circuit (#47 specified.) The pilot lamp operates from the internal ballast in the 35Z5 tube and is also a #47 bulb. The RC-123 receiver is specified as a Mackay Type 128-AZ with two changes. First, the audio output tube's plate circuit is transformer coupled and provides a 600Z ohm output at the phone jack and at the audio output terminals available inside the cabinet (for routing to a console output or operator's desk jack.) If compared to the common Type 128-AY, that receiver's audio output was coupled via a plate capacitor providing a Hi-Z output. The other change is specific to the difference compared to the 128-AZ and refers to the 6.3vac transformer for the tube heaters that is mounted inside the cabinet (the "AY" version does have this tube heaters transformer.) No reception filters are provided for relief from static crashes or other atmospheric noise. "Mackay Radio" isn't shown anywhere on the receiver. The "triangular shield" logo is the only indication that Mackay was involved in production. However, at the bottom of the panel is "built by Federal Telegraph Company, Newark, N.J., USA" which indicates the RC-123 pre-dates the FTC name change that happened during the early part of WWII.

photo right: top of the RC-123 chassis

Serial Number 97 Inspection - This RC-123 was given to me by my old LW friend, Dave Sampson. The idea was for me to restore, document and test the performance of the RC-123. Superficially the RC-123 looked pretty rough. Only close examination revealed its true condition. The dial cover plastic was very dark amber-brown (looking barely transparent) and it was severely cracked and broken. The panel was darkened with over 75 years of "aged cigarette smoke" deposits that almost obscured the panel nomenclature. The cabinet was extremely dirty and stained along with having been painted light gray over the original very dark gray (almost black) wrinkle finish. However, when the chassis was extracted from the cabinet, the condition assessment started to change for the better. The chassis was in very nice condition with no corrosion at all. All of the components were original with only a few exceptions. The RF Gain pot had been replaced with an incorrect value and the NE-2 Antenna Input protection bulb was missing. It was also noted that one knob, although very close in appearance, wasn't an original type. Closer examination was going to be required on L3, an antenna coil for Band C which appeared burned. The coil still had continuity (I don't know what I measured, actually had an open primary winding) but its actual usability has to be determined when the receiver is functioning. The best surprise was when the darken, broken plastic dial cover was removed it was discovered that underneath the dial itself was in near-perfect condition.

Underneath the chassis, as mentioned, the RF Gain pot was a 25K with a series 25K resistor that needed to be replaced with an original 50K WW pot. There was a length of brown "zip cord" that had the exterior end cut and the interior end was connected across C27 at the component board directly behind the Audio Gain pot. This had the two wires connected to the 6K6 output tube's plate and screen and that also was the connection of the primary of the audio output transformer T2. The purpose of this "zip cord" is unknown, and it wasn't original, so it was removed. As received, the RC-123 had a glass 6SK7GT tube installed for the RF amplifier. This was replaced with a metal 6SK7 tube. All other tubes tested good.

photo left: Underneath the RC-123 chassis. This photo was taken before the rebuild. Note the two upper-most coils, L3 and L4. Note the "wax spatter" on the metal mounting bracket (right coil, L3) indicating where the coil had gotten hot - fast! Note on left coil, L4, how the coil primary is discolored and dark, also indicating high heat.

Panel Cleaning - Tobacco stains are difficult to remove mainly because the discoloration has affected the paint itself. Fortunately, the discoloration normally isn't too deep and if the original paint was applied with heavy coats then one can gently remove just a thin part of the surface which eliminates the tobacco staining and somewhat returns the paint to its original tint. The method used sound harsh (and it can be) but if carefully done it results in an improved appearance to a tobacco-damaged panel. The front panel must be removed and have all of the parts also removed since this process works best if the panel can be kept on a flat surface.

I use plenty of Glass Plus and 0000 steel wool. Keep the panel very wet with the Glass Plus and gently rub the 0000 steel wool in a "figure-8" pattern. This "figure-8" will avoid linear scratching and evens the paint surface removal. Don't scrub the surface and check the progress by washing the panel with Glass Plus and a paper towel frequently. It takes 20 to 30 minutes of this treatment to remove significant tobacco staining BUT it only works on smooth, non-glossy paint, so this process is pretty limited but the Mackay panel paint was this type and the tobacco-stains were 95% removed.

Dial Rebuilding - When disassembling the receiver to remove the front panel one really has to be observant as to how the various parts of the dial assembly are mounted. There are two identical gray colored full-size spacers, one mounts on the front panel and the other is near the back of the "stack" of pieces. From the rear dial frame forward,...first is the dial scale, then one of the two gray metal spacers, then the black spacer. These are mounted behind the front panel. On the front of the panel is the transparent plastic dial cover and then the second gray spacer. The dial pointer has to be in front of the black spacer. There are two washers used as spacers between the rear dial frame and the back gray spacer (two top screws only) to compensate for the thickness of the dial scale material. If the front panel is mounted to the chassis then you can't get the dial scale or the black spacer installed (or removed) because of the dial shaft. The procedure is to loosen the set screws on the dial shaft coupler and slide it back onto the tuning condenser shaft as far as possible. Then loosen the set screw for the vernier reduction and pull it forward off of the dial shaft coupler pin. This allows the entire dial frame and vernier unit to be removed allowing installation of the entire dial assembly. When installing the dial assembly be sure to mechanically "zero" the dial pointer with the fully meshed tuning condenser.

I had to make a new transparent plastic dial cover which is just a flat plastic piece about .020" thick. The vernier reduction unit was lubricated with heavy grease applied with a small brush and worked into the pinch wheels. The dial scale is made out of a celluloid plastic that reacts to water or water-based cleaning. Too much water will begin to dissolve the celluloid. I dry cleaned the dial scale using just a clean flannel cloth to remove the loose dirt. Luckily, the dial scale was in good shape and didn't need a lot of cleaning anyway.

Cabinet Wiring - There are several connections that can be utilized inside the RC-123 cabinet. All four fuses for the LINE+, LINE-, DC B+ and DC A+ are on a bracket mounted to the rear of the cabinet. The fuse holders are the long-shaft type with the screw-in top. These types have a real tendency to break the side connection due to the flimsy design. Two of the four fuse holders had broken side connections. A search of the "fuse holder junk box" turned up two exact replacements that were in good condition. As usual, the fuses inside the holders weren't all the correct current ratings. The LINE fuses should be and were 1A, but the DC B+ should be 1/8A and a 1A was installed and the DC A+ should be 5A but a 10A was installed. Correct fuses were installed in all holders (even though DC operation wasn't anticipated.)

Another problem was the 6.3vac output wiring from the "tube heaters" AC transformer. The wires had be cut and then the ends were wrapped with plastic electrician's tape indicating a non-vintage alteration. That the AC tube heaters supply was removed indicates that some other source was used or intended to be used to power up the receiver. Fortunately the cuts were only about 2" from the transformer and there was ample wire available on the other end to make new connections that were stripped, tinned and soldered to the transformer.

To power the RC-123 on AC requires a power cable be wired to the LINE+ and LINE- terminals on the power strip inside the cabinet. There's a "knock-out" just below these terminals for access. Also, the antenna connections are routed through the back of the cabinet and connect to "M" and "G" on the antenna terminal strip. "E" and "EXT" are connections for the emergency antenna (E) and a connection for another receiver (EXT) to share the emergency antenna. There is another terminal strip that provides an external connection for the 600Z audio output to run to a console or to a phone jack on the operator's desk.

Cabinet Paint - The interior and the exterior bottom were still the original dark Mackay gray but the two sides, the back and the top had a coat of light gray paint applied. The paint job was definitely of poor quality but not bad enough for a total repaint. Besides the correct color would require a wrinkle finish base coat and a matched color top coat. The existing paint was thoroughly cleaned using a brass brush and Glass Plus. This was allowed to dry for several hours. I then matched the dark Mackay gray using Artist's Acrylic paint. I mixed a fairly large amount in a small cup and thinned the paint with water. Then using a paint-dampened sponge I dabbed the thinned paint onto the cabinet. This process takes awhile because the paint goes on so thin. The sponging action allows the "wrinkle" to still show through the paint. It usually takes a couple of coats of sponging the acrylic paint onto the cabinet to get the color correct. Sometimes the paint will seem to separate and look a little "foamy" but when it dries it will look correct. The disadvantage of this type of paint treatment is that Artist's Acrylic isn't a very durable paint. Once it's dry, you can wipe the cabinet with a dry flannel cloth but you can't use any dampened cloth for at least a week or so. Also, Glass Plus (and certainly Windex) will attack the paint so dry-cloth dusting is the safest approach. The cabinet will have a matte finish and the wrinkle will show. It's about the easiest way to get a "color match" on wrinkle paint without going through an entire repaint job.

Shock Mounts - The manual shows the typical WWII type shock feet installed directly to the bottom of the cabinet. I used a set of WWII shock feet I had purchased for another receiver but never used.

Power-on Testing, L3 Primary and other Problems - As mentioned in the "inspection" section above, L3 looked like the primary winding had gotten very hot. L3 actually splattered wax on the band switch shield panel because it had gotten so hot so fast. Probably lightning damage, especially since the NE-2 static drain bulb was entirely missing (wires were there but the bulb was gone.) Only band C seemed to be affected. Bands D, B and A functioned more or less correctly (I thought.) I measured the B+ at only +12vdc but after re-reading the manual I noted that if an isolation transformer is used then the LINE- should be tied to chassis ground. When that connection was made the B+ went up to +40vdc,...just a third of what it should be. As a test, I "tack soldered" two 10uf electrolytic caps in parallel with the dual 8uf can electrolytic. That got the B+ up to +130vdc about where it should be. The receiver then began to function somewhat normally. Regeneration, RF Gain and AF Gain all worked normally. I tuned in WWVB 60kc (strong signal) and KPLY 630kc (strong signal, sort of) as a test that the RC-123 was basically working with the exception of Band C and that fault was the open primary winding on L3. 

Rewinding L3 Primary - When attempting to repair a coil the first thing to try is a close examination to see if perhaps it's something easy, like a break. Unfortunately, the primary was wound with Litz wire and the heat-damage made it difficult to find any single break. Several breaks were found and the heat apparently had made the wire very brittle. Examination looked like the Litz wire had about ten strands. The coil looked like it was about 12 layers deep. The DC resistance should be 2.8 ohms. I salvaged a Litz wire wound LF coil out of a derelict RBH receiver coil box. Since this coil measured about 8.0 DC ohms there would be plenty of wire to wind a new primary coil for L3. The burned primary winding was removed and the form cleaned. The original wrap was the "zig-zag" type that requires a mechanical coil winder (like a Morris, which I don't have.) Since this was only a primary winding, I just layered the windings and tried to keep all of the layers equal. To measure the DCR of Litz wire all of the strands have to be soldered together for an accurate measurement. Two times checking and the second check measured 3.0 ohms which was close enough. I installed new black sleeves and then soldered the ends to the terminals. I then coated the entire primary coil with bee's wax, as original. Before installing the repaired coil, I rechecked the DCR and the primary was 3.0 ohms and the secondary was 14.8 ohms, both close to what is shown in the manual. With L3 installed, the RC-123 now "came alive" on Band C, not that there's an abundance of signals from 100kc up to 240kc but the signal generator proved that Band C was working correctly. An alignment of Band C improved reception significantly. Band D was also aligned. Band B and A are aligned together using a trimmer located on the tuning condenser stator B.

Rewinding L4 Primary - In comparing the repaired L3 signal levels along with the levels on Bands B and A, it was apparent the Band D signals level were far less than those Bands and the Antenna Trimmer had no affect on signal level on Band D. The Antenna Trim worked fine on Bands A, B and C. I measured the DCR per the procedure in the manual and the L4 primary was twice the resistance it should have been (measured 1.8 ohms, should have been 0.8 ohms.) Also, there was very little wax left on the primary wires. I suspect that when L3 got the lightning "hit" there was enough "flash-over" (because of the common chassis connection and low DCR) that it over-heated the L4 primary winding and changed its characteristics.   >>>

photo above: The rebuilt L3 Antenna Coil for Band C in the foreground. The coil at the upper-left is L4, note the burned look of the primary. This photo was taken before L4 was rebuilt.


photo above: L4 primary rewound, rewaxed, reinstalled and working

>>>   Once L4 was removed from the receiver, I had to melt the bee's wax off of the coil to see how it was made. It's a little different in that the primary winding is actually two coils, one on each side of the secondary. The two primary coils are wired in series and consist of a single layer of Litz wire. Although the primary coils had continuity, close examination showed that the cotton (or silk) had burned up completely and had mixed with the lightning-melted wax to make a black matrix that the wire was embedded in (a carbonaceous mixture?) Perhaps this might have been affecting the primary coil field and coupling. It's also possible that the extreme instantaneous heat changed the wire characteristics itself. At any rate, L4 wasn't working as it should.

The secondary winding had to be disconnected from the coil form in order to unwind the lower primary coil. Both primary coils were then unwound. Underneath, the white cloth adhesive tape (that provided a cushion for the coil wire against the coil form) was totally burned and destroyed. I used one layer of white masking tape as a replacement cushion. I rewound the primary coils with Litz wire (as original.) When each single layer was complete the DCR of each was 0.4 ohms, so the two in series would be 0.8 ohms.

Tinning Litz wire is difficult but very necessary to assure that all of the strands are connected together at the terminals. I think new Litz wire would have been easier to work with but the junk RBH coil was easy to access and was the correct size even though it was 80 years old. To tin the Litz wire, careful removal of the enamel insulation is required. I used a razor blade and gently scraped the enamel off. This has to be carefully done so none of the wires are broken. Once clean and shiny, the strands are twisted together, again lightly scraped and then the solder is applied. Even with all of this care, the solder joints at the terminals didn't easily flow over and though the Litz wire. It could have been wax contamination since the bee's wax residue was in every part of the coil. Eventually, with more light scraping, the solder flowed and the joint was decent. Once all the Litz wire ends were soldered to the correct terminals then the entire coil was coated with new bee's wax.

The rewound L4 was reinstalled into the RC-123. First thing noticed,...the Antenna Trimmer now worked. Next, slight changes in the coil itself due to the reworking required a touch-up on the alignment. Now, Band D has much better sensitivity and behaves more or less like Bands C, B and A.

Quick Reception Test - About 2200 hrs (Aug 31, 2020) I performed a quick check to see how the RC-123 would function, even though the capacitors hadn't been replaced. I had performed a resistor check earlier and surprisingly all of the ERIE resistors were within 25% tolerance. During the R-check, I discovered that one end of R5 had never been soldered (grid leak for RF amplifier) and had been relying on a loose wrapped joint for the past 75+ years. Cleaned joint and soldered to correct. During the reception test, I tuned from 405kc down to 325kc to see how certain known NDBs could be heard. MOG 404kc was very strong, QQ 400kc was moderately strong, ZP 368kc moderate signal and DC 326kc was moderate. Conditions were pretty noisy and it's still Summer, so it's not the best time for MW DX. I was using the 135' "T" antenna. The test showed that, more or less, the RC-123 is operating correctly and does receive signals of moderate strength in the MW portions of the spectrum.

Capacitors - There are twelve paper-wax tubular capacitors in the chassis and two in the cabinet. The wax was cracked on many of the cap ends indicating a loss of the sealing property of the wax. Even if the paper dielectric is sealed, leakage current can still develop since it's basically caused by contamination of the paper at the time of manufacture. However contamination problems are accelerated with moisture ingression so these capacitors undoubtedly have some leakage current problems. Any new polyfilm cap is about 1000 times better. All I had on hand in sufficient quantity were "yellow jackets." I usually paint these types because the yellow-color is so terrible-looking,...way too bright, looks like they belong in an imported toy. I painted the new capacitors black and left it at that. There are ten .05uf caps and two .1uf, so it's not too hard to keep track of which value is which. I didn't replace the two capacitors in the cabinet because they are bypass caps for A+ and B+ and are only in the circuit if DC operation is going to be used.

The wrong value RF Gain pot was replaced with a NOS 50K 4W WW Clarostat pot that was almost identical to the original. The missing original pot was an audio taper. I could only find a linear taper style so the RF Gain adjustment is more towards the top 25% of the range.

Post-Recap Test - At 2220hrs to 2235hrs (Sept 3, 2020) I tuned in fifteen NDBs in fifteen minutes of listening. Antenna was the 135' "T" and I was using 600Z phones for the reproducer. Tuning range was from 408kc down to 296kc. NDB stations heard were the following:

1. MW 408kc - Moses Lake, WA                          9. AA 365kc - Fargo, ND
2. MOG 404kc - Montegue, CA                           10. RPX 362kc - Roundup, MT
3. QQ 400kc - COMOX, BC,CAN                      11. MEF 356kc - Medford, OR
4. PNA 392kc - Pinedale, WY                             12. NY 350kc - Enderby, BC,CAN
5. YWB 389kc - Westbank, BC,CAN                  13. XX 344kc - Abbotsford, BC,CAN
6. PI 383kc - Tyhee, ID                                        14. DC 326kc - Princeton, BC,CAN
7. GC 380kc - Gillette, WY                                 15. LGD 296kc - LaGrande, OR
8. ZP 368kc - Sandspit, QC Is, BC,CAN

Conditions were very good for early September with relatively low noise levels. All stations were easy copy with moderate to very strong signals.

The RC-123 is now functioning as it should. During the LW season peak in late-November through mid-Jan, DX NDBs should be plentiful and fairly easy copy. The receiver is very easy to operate and the tuning rate, although fast, is able to fine-tune signals in the MW portion of the spectrum. Regeneration doesn't have to be adjusted at the feedback point for good sensitivity. This is probably because of the "Regeneration Tube" that amplifies the feedback. Maximum sensitivity is still at the feedback point but slightly beyond that gives good stability and ample feedback. A couple of minor negatives would be that the RC-123 is a very "basic" receiver with no filters for static crashes and no selectivity filters. What you hear is pretty much everything as far as spectrum noise. The other minor point is the dial resolution is extremely vague with 25kc separation of the index marks on Band D. Considering the RC-123 was designed in 1941 and that it was primarily a commercial receiver that was "pressed into" military service, it does a good job with no "fluff" or "frills."



Radio Corp. of America

Navy Department - Bureau of Ships

RAK - 15kc to 600kc,  RAL - 300kc to 23mc

2 TRF Stages, Regen Detector, 2 AF Stages

Contractors: RCA-Victor, Andrea Radio Corp. or The Magnavox Company

photo right: Andrea Radio Co. CND-46155, RAK-7 - Accepted by USN Jun 1945
photo below-left: Andrea Radio Co. CND-46156, RAL -7  - Accepted by USN Mar 1945

The Navy wanted to replace the RAG and RAH receivers that had been contracted in 1933. The RAG/RAH were TRF receivers with non-regenerative detectors and tracking BFOs built by Hygrade Sylvania Corporation. The Navy wasn't satisfied with the questionable sea-going durability and ultimate reliability of the RAG/RAH (or maybe they weren't satisfied with Sylvania) and only one contract was issued. The RAK/RAL came along by 1935 and the design (RCA with Navy Dept. input) and manufacturing must have impressed the Navy since ultimately eight subsequent updated versions were contracted, with production going on until the end of WWII. There was another version of the RAL that was supplied to the USN that was designated as the TBR-1, a portable transmitter-receiver outfit. RCA built the receiver for the TBR-1 and it was essentially a RAL with slight modifications for portable operation. RCA-Montreal made a version of the RAK receiver designated TE236. It was essentially a RAK without meters or elaborate filters.

Any receiver built for WWII shipboard use had to be "bullet and bomb" proof, in other words, the ship had to take a couple of torpedoes, be sinking fast and the radio gear would still be working. Additionally, steel and iron were kept to a minimum in shipboard radio construction to reduce corrosion problems that were common on marine equipment. The RAK/RAL series were built like the battleships they served on. The construction is something to marvel at - so robust, so over-built, so heavy-duty,...and with no expense spared - it's no wonder that most RAK or RAL receivers still function with all original parts even though they are pushing eighty years old. The electronic design concept was to provide maximum reliability in severe service by simplicity of design - and it paid off since the receivers were in use up until the end of WWII with their last service onboard submarines.

RCA was the primary designer with Navy Department input and RCA-Victor was builder of the first contracts of RAK and RAL receivers. The demand during WWII required another company, Andrea Radio Corp., to build RAK and RAL receivers however the "contractor" was still RCA. The RAK, (aka CND-46155 by its Andrea-build/Navy designation, substitute "R" for the "N" for the RCA-build /Navy designation) covers 15kc up to 600kc in six tuning ranges while the RAL covers 300kc up to 23mc. There was a RAK-8 and RAL-8 produced with The Magnavox Company as the builder for RCA. The Magnavox versions have a slightly different panel nomenclature plate design but otherwise are identical.

The RAK and the RAL used glass tubes that were large six-pin type, 6D6 tubes for the two RF amplifiers, a 6D6 for the regenerative detector, a 6D6 for the first audio amplifier, a 41 for the audio avc amplifier and another 41 for the audio output. The power supply, CRV/CND-20131, was a separate unit that used a 5Z3 rectifier, an 874 regulator tube and an optional 876 ballast tube. The 876 can be "switched out" of the power transformer primary input circuit if the AC power is stable and noise free. An internally mounted toggle switch routes the 120vac to a different tap on the power transformer and also disconnects the 876 from the circuit. When the 876 is switched in, the 120vac actually is IR dropped through the ballast and a different tap on the power transformer is used (~70vac) thus providing the regulation of the AC to the transformer if the line voltage is not stable. Since the ship had to generate its own power and most of the equipment onboard (including motors to rotate gun turrets) ran on this power, the varying switching loads were what caused the line voltage fluctuations that required using the ballast regulators. In shore set-ups, on standard AC line power, ballast regulators were not required and could be "switched out" of the primary circuit. However, the Navy manual (NavShips 900,480) recommends that the RAK can be operated without the 876 ballast if the AC line maintains 10% regulation, BUT the RAL receiver should always be operated with the 876 ballast tube switched "in" regardless of the AC line stability. The RAL had to operate at much higher frequencies than the RAK and slight AC line variances could cause receiver frequency instability. I've only experienced frequency instability when tuning above 10mc and only when receiving CW (or SSB nowadays.) With AM reception above 10mc it's difficult to detect the instability since the detector isn't oscillating. With modern AC line stability the only time slight variances occur are when heavy-load appliances turn on or off within the user's home. While the instability can be noticed, it is not to the extent that it would affect copy. Operating the 20131 power supply with the 876 ballast "switched in" will generate quite a bit of heat since it dissipates about 140 watts. It's up to the user to decide whether the slight instability only on CW above 10mc is that much of a problem versus the power dissipated as heat resulting from having the 876 in the circuit.   

photo left: CND-23073 Control Unit - This was a necessary device if the shipboard set-up required one radio operator to monitor (guard) two radio frequencies simultaneously. The RAK and the RAL were tuned to different frequencies that needed to be "guarded" and by selecting "MIXED" the audio lines from each receiver were combined at the radio ops 'phones. The Control Unit also allowed the radio op to power the receivers individually. All wiring cable connections are accessed at the rear of the unit and the unit was normally "bolted" to the table within easy reach of the radio operator.

photo right: CND-20131 Rectifier Power Unit - Each receiver required its own power supply connected to the receiver by a ten foot long four-conductor shielded cable (there are five spade lugs since one lug is for the shield-ground connection.) The AC line to the power supply could be switched using the toggle on the power supply front panel or, if used and wired, the Control Unit could be used to switch the AC line to the power supply. There's also a DC Power toggle switch on the receiver if batteries are used to power the receiver. The Rectifier Power Unit has a 5Z3 rectifier tube, the voltage regulator is an 874 and the ballast tube (if used) is an 876.

The RAK was designed primarily for CW, ICW or MCW reception. The receiver has a low pass filter that is permanently connected in the audio circuit to roll off the upper audio frequency starting at about 1200 hz. An elaborate audio avc circuit allows the user to limit the audio or noise peaks at an adjustable level. This was to provide the radio operator relief from "static crashes" common near storm fronts. Also a selectable-frequency audio bandpass filter was provided to enhance the CW tone reception for noisy conditions. Voice transmissions can be received (nowadays in the 540kc to 600kc portion of the AM-BC band) but the RAK audio response is restricted to about 400hz to 1300hz, so voice or and music transmissions sound "hollow" with little depth. The manual states that another receiver should be used if voice reception is required, implying that the RAL should be used since its audio filters can be switched out of its circuit. The tuning of the RAK is heavy duty, gear driven and the tuning dial readout is shown on two circular dial scales of 0 to 10 and 0 to 100. The actual tuned frequency has to be correlated with a graph that is in the manual. The receiver does provide a logging chart on the front panel for a "most used frequencies" reference. A RF trimmer, an antenna trimmer, sensitivity and regeneration controls are on the lower panel of the receiver. The meters monitor audio output level in db and tube heater voltage (doubles as the "power on" indicator.) The RAL receiver is almost identical construction but has nine bands covering 300kc to 23mc. Additionally, (as mentioned) the low pass filter and the tuned audio filter can be switched out of the RAL circuit for voice reception. Also, a vernier tuned frequency control was provided. In some installations, the two receivers operated together through a Control Unit (CND-23073) that allowed the radio op to monitor two frequencies simultaneously. The Control Unit also could be used to switch the AC to the receivers on or off.

photo left: The RAK and RAL in use aboard a US Navy ship. Also, National RBL and RAO receivers far left,  the LM-type frequency meter by the telephone handset and a Scott SLR-type receiver below the order binders.

Nowadays, a complete RAK and RAL set up will require a heavy-duty table for the set-up since the total weight of the two power supplies and the RAK and RAL receivers is well over 200 pounds. In my installation I had the power supplies for the RAK/RAL receivers bolted to the underside of the table. I provided for a space of about 3.5" above the supplies to allow good ventilation for the ballast tubes. I ran the power supplies with their ballasts even though wasn't necessary. The actual difference in power consumption using two power supplies is significant - as mentioned, the ballast dissipates about 140 watts - each! I had run the receivers both with and without ballasts and I noticed that the received noise seemed to be slightly less with the ballast in use. Of course, this was operating the RAK and RAL in Virginia City, a city notorious for noisy, decrepit, unstable and usually low voltage AC lines (nowadays, here in Dayton, NV, the AC line is way too high at 124vac although it seems to be fairly quiet and stable.)

In actual operation, the RAK is a very sensitive receiver that spreads the LF tuning range over several bands. This bandspread action is nice for tuning in weak stations or trying to separate several stations that are on the same frequency - as many NDBs are. The major problem is that calibration is relying on the readout versus a graph and that graph is in the manual. The first thing to do is make a copies of the frequency graphs to keep with the receivers (RAK and RAL have separate frequency to tuning dial charts.) Then it's easy to keep track of where you are in the LF spectrum. If it's important to know the exact frequency, use a heterodyne freq-meter set up (or a RF Signal generator with DFC can also be used as a calibrated heterodyne freq-meter.) The Audio AVC will help with static crashes and to a certain extent, noisy conditions but, like most output limiters, if it is advanced too far it severely clips the audio with high distortion. With the AVC control, 10 on the scale is minimal AVC action and reducing the setting (CCW rotation) will increase the AVC action. A setting lower than 3 will usually begin to affect the CW audio tone. The adjustable frequency audio bandpass filters seem at first to be almost useless due to the seemingly high audio frequencies chosen of 450hz to 1300hz. All NDBs use MCW with a 400hz tone. When tuning a NDB, with the detector oscillating to provide a heterodyne, the tendency is to tune for carrier zero beat but that will only allow the 400hz note to be heard. By selecting a higher tuned audio frequency, for example 800hz, and tuning to enhance just the MCW note (that won't be 400hz anymore,) the selectivity of the filter will allow hearing the Morse in the clear but the carrier will be attenuated. It takes some practice and a moderately strong NDB signal. For DX NDBs, the Tuned Audio Filter should be switched off.

A remotely-tuned loop antenna, with its high Q, really helps reduce the noise and increase the signal to noise ratio. I used a ten foot diameter remotely-tuned loop with the RAK-7 when I was in Virginia City and the signal reception was excellent. Unfortunately, that large of a loop didn't survive the mountain winds. I have yet to test the RAK-7 using the Pixel Technologies shielded magnetic loop that I now use but, as soon as I do, I'll add the results here. The easiest access to the audio output is from the front phone jack. It's 600Z ohms and, while the RAK will easily drive a 600 ohm loudspeaker, many more weak signals can be copied using 600Z earphones rather than using a loudspeaker. The 600Z audio out the back of the receiver was normally routed to the Control Unit. Both the rear audio line and the front panel phone jack have identical filtering inline. I have tuned in all of the normal LF signals with my RAK-7. The best NDB DX were several in North-Eastern Canada and Puerto Rico's powerhouse transatlantic beacon, DDP. At lower frequencies, the RAK seems to get better and better with JJY at 40kc being a regular copy. The Navy MSK signals from 19.8kc up to 25.2kc are always present. WWVB, JJY and all of the USN MSK stations will require using the Tuned Audio Filter set to about 800hz for best sounding signals.

photo above:
  "Women Marines - USMC PHOTO - 24-8" -  showing a Women Marines Reserve radio op using two RAK receivers. She's tuning the right-side RAK and is ready to copy on the "mill" (the typewriter.)  RAK power supplies are on top of the receivers in this set up. Note how she has the 'phones slightly in front of her ears. This helps reduce "ringing ears" from static crashes or unexpected loud "pops" common on the lower frequencies. The RAK (and RAL) have audio AVC circuits for an output limiter function. Also, note that her 'phones are plugged directly into the Phone jack of the RAK indicating that a Control Unit isn't part of this dual-RAK set up.



U.S. Army Signal Corps

Farnsworth Television & Radio Corp.



BC-344-D - LF/MW Radio Receiver - 1944

150kc  to  1500kc

Double Preselection Superheterodyne


SN: 5     Contractor: CFN (Farnsworth) 

  Order No. 10651-PHILA-44

The BC-344 belongs to a family of radio receivers that were designed by the U.S. Army Signal Corps at Fort Monmouth, New Jersey in the late-thirties. The early versions of the receiver used more aluminum in the chassis and some early versions were painted with a "leatherette" finish paint (properly called "Crackle Finish.") By the time WWII contractor manufacturing had started, the receiver used an all-steel chassis and the paint used was the standard black wrinkle finish. Farnsworth Television & Radio Corporation built the AC operated versions, those being the BC-342 and the BC-344. RCA Manufacturing Co., Inc. built the DC operated versions, those being the BC-312 and BC-314. The BC-312 and BC-342 are shortwave receivers with six frequency ranges that cover 1.5mc up to 18.0mc. The BC-314 and BC-344 are LF (low frequency) and MW (medium wave) receivers with four frequency ranges the cover 150kc up to 1500kc. The BC-344 has two RF amplifiers and two IF amplifiers (four 6K7 tubes.) A separate Mixer (6L7) and LO (6C5) are used. The BFO uses a 6C5 triode, the detector-AVC-first AF is a 6R7 and the AF Output tube is a 6F6. There are ten tubes used in the BC-344 including the 5W4 rectifier tube used in the RA-20 power pack that mounts internally. No selectivity filter is used in the BC-344. Audio output is generally 4000Z ohms but may be set to 250Z ohms by moving a jumper inside later versions of the receiver. The LS-3 loudspeaker was commonly used with the BC-344 receiver although a headset would allow copy of weaker signals.

photo above: Army communications set-up for local ground transmissions. The telephone box allowed interfacing a telephone contact through the BC-191 transmitter. Note that the vernier dial locks are removed from all three BC-342 receivers.

The Signal Corps utilized the BC-312/314/342/344 receivers in many applications. Fixed stations generally used two or more receivers paired with at least one BC-191 transmitter. A typical Army station is shown in the photo to the left. Note that three BC-342 receivers are utilized with an AC operated BC-191 transmitter (the RA-34 power supply is barely visible under the transmitter table.) Mobile stations ranged from small vehicle set-ups to the SCR-299 that used a BC-342 and BC-344 set up with a BC-610 transmitter, all powered by an AC generator towed in a trailer. Most DC set-ups operated on 12vdc from vehicle battery-charger systems. There was a 28vdc BC-312 version, the BC-312-NX.

BC-344-N on Long Wave? - I've been interested in how well the BC-344 would perform on MW and LF reception for some time. I have a "rough-condition" receiver that had been painted "metallic" green. I even had it on the workbench once but I lost interest when I saw some "hamster" rework inside. Recently (2016,) in a trade that involved a "drop in ready" AC power supply for a BC-348 receiver, I obtained this BC-344 receiver that happened to have the scarce shock mount installed. This receiver seemed to be pretty original and virtually complete. Replacements for the few missing or poor condition parts were easily located in my BC-312/314/342/344 junk box.

Reworking the BC-344-D - Since the BC-312/314/342/344 were essentially "workhorses" for the Army, many are in rough condition today. All will require some work and many require a full restoration in order to function correctly. The receiver is difficult to disassemble and this has saved many BC-344s (and 342s) from being modified too much by hams. Most hams preferred to modify the easy-to-work-on BC-348-Q. Most 344/342 receivers will be very close in alignment because all of the adjustments were "locked" if they were easy to access, like the IF transformers, or hidden by covers or plugs in the case of the RF adjustments. Most receivers will be missing the dial lock since this piece often times interfered with the rotation of the vernier knob. It's not unusual to do mostly a cosmetic restoration and have the receiver work with all original parts with the exception of the dual electrolytic filter capacitor located in the RA-20 power pack.

Front Panel Dismounting - One of the common problems with any of this series of receivers was present on this BC-344. That is, the plastic dial index that always seems to have warped, cracked, shrunken in size and discolored to the point where it's no longer covers the dial itself or is even transparent. Luckily, I had an excellent condition replacement. Unfortunately, to install the index requires completely dismounting the front panel.

I say "unfortunately" because this task is unbelievably complicated by the Army's mechanical design that never seemed to even consider the possibility that the receiver might need to be disassembled someday. Besides the abnormal amount of front accessed screws there are three screws that mount from the backside of the panel that must be removed. The two fuse holders must be unsoldered and removed. All of the phone jacks and controls must have their mounting nuts removed. The wires to the dial lamps have to be unsoldered. All of the wires going to the front panel "trunk" connector have to be unsoldered. The power input terminal strip has to be dismounted. All screw connections to Antenna and Ground connections have to be dismounted. Knobs and control nuts must be removed. At this point, you'd think the front panel would easily come off, but not yet! The Fast Tuning gear has a pinned shaft that is flanged. You can't remove the front panel unless the gear and shaft are taken apart and that requires driving out the pin. However, once the gear and flanged shaft are apart, the front panel finally can be dismounted.

photo right: A rear view of the chassis showing the stout construction of these receivers

At this point, installing the replacement dial index is easy (it's mounted with eight screws!) Since the front panel is off, now is the time for a thorough cleaning and touch-up. I use jet black nitrocellulose lacquer that is thinned about 3:1 for touch-ups. After the lacquer has set for a while, I rub down the panel with light weight machine oil. Usually, this will blend the color of the touch-ups with the original black wrinkle paint making the touch-ups invisible.

When remounting the front panel, it will be noted that all of the 6-32 screws are the same length. However, there are three different length 4-40 screws. The four long 4-40s are for the wire mounts. There are two short 4-40s and the rest are all the same length. The two short 4-40s are very important. There are two places where, if long 4-40 screws are used, they will protrude far enough thru the rear panel mounted nut to contact the dial mask with possible scratching of the mask when the band switch is operated. The "short" 4-40 screws must be mounted as follows: One "short" screws is by the band change switch and is the screw nearest the "G" in "CHANGE" in the nomenclature of that switch. The other "short" 4-40 is used near the lower left side of the data plate (the screw head almost is contacting the data plate edge.)

The other observation is the small component board that is mounted to the back of the front panel uses two different length stand-offs. The reason is that the lower screw must past thru a part of the chassis mounting flange and the top screw doesn't. The longer stand-off is the upper one.

Patience is required in any BC-312,314,342,344 rework. The work is tedious and none of the component parts are easy to access without removing another part or assembly first.

  Electrolytic Filter Capacitor - Most military gear used oil-filled paper dielectric capacitors as the filters. The disadvantage of this type of capacitor was its large size but the advantage was long-term reliability. Since space was at a premium in the RA-20 power supply, a dual electrolytic filter capacitor was used. The advantage was a small size for two 8uf 450vdc caps but the disadvantage now is, after 75 years, the dielectric paste has dried up and drastically reduced the value of the capacitor.

There's ample room to mount new electrolytic capacitors inside the old capacitor can. There are many methods to accomplish this "re-stuffing." I cut the cap in two and remove the old cap and black wax using a heat gun to soften the wax. This allows the old capacitor to be easily pulled out of the can. I install new capacitors connecting them to the correct terminals and then epoxy the can back together. I remount the rebuilt cap and connect it into the circuit.

The photo to the left shows the underside of the BC-344 and in particular the RA-20 power pack. Note that the RA-20 is "swung out" on the right side hinge mount. Also note the wiring harness that is connected to the power terminal strip. 

Miscellaneous Rework - Sometimes the OFF-MVC-AVC switch contacts don't allow the operator the ability to turn the receiver on. These are special build, stacked switches that have the MVC and AVC switches enclosed. Luckily, the AC on/off switch is only covered with a fiber disk that can be bent to allow access to the arm and contacts. It's usually only dirt or maybe minor corrosion that is causing the problem. This can be cleaned off using a small piece of Alu-Ox paper or a very small wire brush. It might be necessary to bend the arm slightly to have better contact. If the AC contacts are damaged beyond reconditioning then the entire switch has to be replaced with a used-good original. These are sometimes difficult to find except from parts sets. The same switch is used on all of the versions of the BC-312, 314, 342 and 344. Alignment Notes - IF is 92.5kc. Note in the photo above left that the large LO box (far left) has four plugs that are covering access to the trimmers for each band. Also note the shield cover over the two RF amp boxes and the Mixer box that is preventing easy access to these trimmers. Each IF transformer adjustment has a lock nut on the threaded shaft to prevent tampering (see photo above right showing top of chassis.) Other than the lower frequencies involved, the BC-344 is straight forward in its alignment procedure.
Performance on a Wire Antenna -  The antenna used was my 135' CF Inv'd Vee with 96' of ladder line with the two feed line wires tied together. This is something like a "T" antenna and it performs fairly well on LF. The test listening was on January 29, 2018 from 1915 to 1945 PST. I only tuned the BC-344-D from 405kc down to 300kc. Within the frequency span and time period I tuned in 25 NDBs. Greatest USA DX was IN 353kc in International Falls, MN. Greatest DX was YMW 366kc in Manawaki, Quebec, CANADA. Another Canadian NDB tuned was YXL 346kc in Sioux Lookout, Ontario. Listening QTH was Dayton in Western Nevada.

All signals were heard over a headset, not by loudspeaker. BFO was on and stations generally tuned for zero beat of carrier to then hear the MCW tone correctly. Sometimes it's advantageous to tune the BFO about 400hz to 1000hz from the IF frequency. This will enhance the 400hz on some weak NDB stations. The AVC was off. Noise level wasn't too bad and I'd rate the conditions as "very good."

The BC-344-D is a capable LW receiver that has ample sensitivity in the medium wave portion of the spectrum using a wire antenna. There is no crystal filter or any other method to reduce IF bandwidth so many signals are "heard" over what seems to be a fairly wide IF passband. It's easy to select one particular signal and then concentrate on that tone or sound of the signal to then successfully copy the call. Most of the time this process will be with very, very weak signals that are within the passband with stronger signals. If the received noise seems to be covering up some weaker signals it's possible to slightly "detune" the antenna trim (ALIGN INPUT control) to reduce the noise while not affecting signal copy.

While the BC-344-D is kind of a "basic" superhet with no fancy filters or output limiters, it does a good job with the few controls provided. It's certainly sensitive enough and the reduction of the VERNIER tuning mechanism allows for easy tuning of all stations, including NDBs. Without an output limiter though there isn't any way to reduce "pops and clicks" or heavy static. Wearing the 'phones just ahead of the ears is recommended. I'm sure if the BC-344-D was used with a wire antenna in a modern RFI-noisy, urban-type location the weak signal DX reception would be impossible. But, in the rural RFI-quite area here in Dayton, NV, the BC-344-D does a very good job of pulling in NDB DX using just a wire antenna.

Performance on  a Tuned Loop Antenna - The remotely tuned loop is six feet "point to point" of its diamond-shape and tuned using varactor diodes with a variable bias voltage source. The receiver is connected to a pick-up loop that is mounted within the main loop. Test listening was on February 1, 2018 from 1930 to 2000 PST. I only tuned the BC-344 from 325kc up to about 390kc. 23 NDB stations were tuned in during that time. Conditions were very good. Greatest USA DX was FIS 332kc in Key West, FL. Greatest Canadian DX was GW 371kc in Jarpik, Kuujjuarapik, Quebec, CAN.

All signals were heard over a headset, not by loudspeaker. BFO was on, AVC was off. Loop Antenna was pointing NE/SW for the entire test listening period.

The BC-344-D running with the tuned loop antenna is a surprisingly good performer. Without any bandwidth controls, one would expect noise to be a problem but the loop does keep the noise level somewhat lower than the wire antenna. Additionally, the "loop tuning" provided the ability to slightly detune the peak adjustment which lowered the noise without causing loss of the signal. This was a nice advantage for signals that were "in the noise." Also, the ALIGN INPUT (antenna trimmer) could be used to slightly detune the antenna and reduce the noise peaks allowing some weaker signals to be copied. These two features, loop tuning and antenna trimmer, do a lot to make up for the "wide open" IF bandwidth which does have a tendency to be somewhat noisy.

As mentioned in the "wire antenna test" section, the BC-344-D is a pretty basic receiver with no filters or no bandwidth control. Yet, I was still able to copy NDBs all the way to the east coast and Canadian NDBs out to Quebec. I don't think any vintage LW receiver enthusiasts will be going out to purchase a BC-344-D to use as their main LW receiver, but I was impressed with its performance and would rate the receiver as one of the better medium wave and low frequency receivers. It is fully capable of receiving DX LW signals and, with its incredible "WWII Military" appearance, makes a fine addition to any collection of vintage LW receivers. Additionally, if you get tired of listening to MCW signals on 'phones, you can always tune in almost the entire AM BC band and listen on an LS-3 loudspeaker - very cool.



Radio Corporation of America
RCA-Victor Division, Camden, N.J., USA



 Double Pre-selection Superheterodyne Receiver

LF, MW and SW

75kc to 550kc and 1.5mc to 30mc

(includes NDB reception log)

CR-91 sn:050068 during its 2022 MW reception testing using the Pixel Loop Antenna.

The CR-91 and the CR-91A are not the same receiver. During WWII, the CR-91 was the USA-built version of the RCA-Montreal-built AR-88LF. After WWII ended, the Camden CR-91 production stopped. The Montreal AR-88LF was updated and continued to be built but designated as the CR-91A.

Brief History of the AR-88 Series - RCA's AR-88 planning may have chronologically followed their AR-77 ham receiver but the AR-88 owes much of its design concept as a replacement for RCA's aging commercial-military receiver, the AR-60. The AR-60 had been introduced in 1935 and was still being built as late as 1940. RCA did produce an AR-80 for the military that was ruggedized AR-77 that was field-portable. RCA had to update their "cost no object," highly reliable military/commercial product and the AR-88 was the result. Design stages probably date from as early as 1939 but urgent demands of  WWII in Europe pushed RCA into having the AR-88 ready by early 1941. The finalized AR-88 was a 14 tube superheterodyne that covered .54 to 32MC in six tuning ranges, featuring incredible sensitivity (even up to 10 meters), excellent stability and high fidelity audio along with mechanical and electronic reliability that couldn't be found in any other receivers of the day (except for those over-built RCA creations for the Navy.) The electronic design was the work of Lester T. Fowler while George Blaker handled the mechanical design. The actual production during WWII was handled by RCA's Export Sales under Charles Roberts in Camden, New Jersey. Additional receivers were produced at RCA facilities in Bloomington, Indiana and Montreal, Quebec, Canada.

Lend-Lease - During the AR-88 design phase, it became apparent that demands of WWII would supersede any commercial market and that a reliable high-performance receiver was going to be required for surveillance and intercept,...not in the USA but in England and for other Allies. To say that ALL of the early AR-88 production was sent to England wouldn't be too much of an exaggeration. RCA wanted to export all of the receivers and didn't want any of them returned after the war (Lend-Lease options were to return, destroy or pay for, at a reduced price, the equipment that had been sent to the end user.) The AR-88 was produced at Camden, New Jersey by the RCA-Victor Division of RCA but another version of the receiver that featured 75kc to 550kc coverage, skipped the AM-BC band and then continued coverage from 1.5mc up to 30mc, was also being concurrently produced for export by RCA-Montreal in Canada. These receivers were designated as AR-88LF and again almost ALL of these receivers were exported to England though many LFs did remain in Canada for their military. The early versions of the AR-88 and the AR-88LF are seldom encountered in the USA. Almost all AR-88 versions built during the early part of WWII were exported to our allies by Lend-Lease.  >>> 

>>>  Diversity Receivers - During WWII, RCA developed the DR-89 Triple Diversity Receiver that utilized three AR-88F receivers. Triple Diversity was being used commercially for long-distance, fade-free, reliable communications (usually electro-mechanical data transmissions) but these elaborate receivers and receiving installations were ideal for Allied radio communications during WWII (also primarily data transmissions.) The US Navy also wanted their own version of the DR-89 which RCA supplied as the RDM receiver, again using three receivers designated as CRV-46246A (AR-88F.) RCA-Camden also produced the AR-88D that became the "standard" version of the receiver. The AR-88D and the AR-88F are the versions usually encountered in the USA but many AR-88D receivers were also exported to the Allies during the later part of WWII.
   "D" does NOT indicate "diversity" - "F" was the diversity receiver.

The CR-91 - RCA may have believed that there was a commercial or military need for an AR-88LF in the USA, possibly for government surveillance (the FCC-RID did use some AR-88s) and also for various other purposes that included commercial shipboard use (referenced in the manual) and use at Coastal Radio Stations. RCA designated the Camden-built AR-88LF as CR-91. The production quantity was relatively small with the reported serial numbers ranging from 050068 up to 050958 indicating that probably about 1000 "Camden" CR-91s were built. RCA serial numbers use the first digit to indicate the build location, a "0" in this case indicating Camden. The second digit indicates the model of the receiver, in this case a "5" which was used for the CR-91. The remaining four digits are the receiver's build number, "0068" for the receiver shown in this article. However, there's also the more likely possibility that the AR-88LF production quantity out of Montreal needed to be supplemented with additional AR-88LF receivers built at Camden in order to meet the WWII demand. This thought is based on the fact that nearly all known surviving CR-91 receivers are in countries other than the USA (3 in the UK, 2 in Brazil, 1 in France, 1 in Italy and 1 in the USA.)

Post-WWII - Since the Camden CR-91 receivers were just a requirement of WWII, production ended when the war ended. This then allowed RCA-Montreal to continue to build the AR-88LF but in an updated configuration. These later Montreal-built receivers were designated as CR-91A. These versions have a RCA Umber color smooth finish front panel, a front panel mounted Crystal Filter Phasing control, a different gear box and several other changes from the old AR-88LF. RCA-Camden updated the AR-88D in a similar manner and provided the CR-88 receiver after the war and the CR-88A to replace the AR-88F.

The CR-91 Circuit and Construction Details - The CR-91 tunes from 75kc to 550kc and from 1.5mc to 30mc in six bands. It uses 14 tubes in a double-preselection superheterodyne circuit. The SW front end coils are wound on polystyrene forms that have extremely low losses allowing the receiver to maintain high sensitivity up to 30mc. The gear reduction tuning is substantial at 100:1 and was often referred to as "continuous bandspread." A logging system using two separate dials provided 4400 logging divisions per band for accurate tuned frequency resetability.

The CR-91 provides five steps of selectivity with position 1 and 2 being rather broad for good fidelity while positions 3,4 and 5 use the crystal filter for increasingly narrow bandwidth. The receiver uses three stages of 735kc IF amplification with stagger-tuned IF transformers. Two under-coupled IF transformers and two over-coupled IF transformers are utilized when the receiver is operated in the "BROAD" POS. 1 selectivity position. To assure that the passband is symmetrical usually requires a sweep generator and oscilloscope for proper alignment. However, if fidelity is not an issue, there is a procedure to align the IF section using just a VTVM but the IF bandwidth will usually be narrower than if the sweep method is used. The 735kc for an IF was to allow for continuous coverage in the 75kc to 550kc range and is also why there is a gap in the frequency coverage from 550kc to 1500kc.

The CR-91 receiver also has a clipper-type Noise Limiter and a High Frequency (limiter) Tone control. The audio output is from a single 6V6GT providing about 2.5 watts of power to a 2.5 ohm Z output transformer (600 ohm Z and Hi-Z phones outputs are also provided.) The use of the 6V6GT audio output tube differs from the standard AR-88 that uses a 6K6GT in that function.   >>>

>>>  The receiver power supply uses a potted power transformer and two potted filter chokes for dual section filtering using an oil-filled, multi-sectional unit with three paper dielectric capacitors. Only the AR-88LF and the CR-91 receivers have AC line fuses that are chassis mounted. All other versions required the user to provide a fused AC line. A VT-150 is used to provide a regulated +150vdc for the LO and BFO plates and for the RF/IF screens to improve stability and reduce drift.

A minor difference in the front panel nomenclature is that the CR-91 uses PHONE and BFO on the mode switch while all other AR-88 versions used REC MOD and REC CW for those mode positions. Since the receiver is a bit later in WWII production from Camden, the front panel nomenclature is silk-screened (very early AR-88LF versions were engraved) and the main tuning dial is alternating black and yellow scales (early dials were all yellow.)

Mechanically, the receivers were stoutly built. Heavy steel chassis and an almost quarter of an inch thick, copper-plated steel front panel were the foundation for component assembly mountings that are entirely put together with screws, lock washers and nuts. This was to allow extensive disassembly to be easily and quickly accomplished, the repairs performed, followed by easy and quick reassembly. The only rivets used in the receiver are for the clips that mount the adjustment tools. Early receivers had the chassis side panels bolted in place but late receivers will have the chassis side panels spot-welded to the chassis. The ultra-heavy duty construction made for a stable receiver but also added to the weight. The AR-88 weighs in at just about 100 lbs. when installed in its cabinet. Out of the cabinet, the receiver weighs about 75 pounds.

Finding the CR-91 - In late-2009, I was lucky enough to purchase CR-91 SN: 050068 from Reno's last surviving surplus store, Twin City Surplus.  Unfortunately, after decades of supplying the military surplus needs for Northern Nevadans, Twin City Surplus was sold by its longtime owner around 2015. The new owners did a great job of running Twin City for a short time but the City of Reno's downtown 4th St. renovations and Reno's "anti-military surplus" stance (along with the cancellation of "Burning Man-2020" event in the Black Rock Desert) forced Twin City Surplus out of business in late-2020.

I was told about the CR-91 receiver by a visitor to our Western Historic Radio Museum (I was open in Virginia City, Nevada from 1994 to 2012.) The visitor had seen the receiver at Twin City the day before. To my surprise, in Reno the following day, the CR-91 was still there. It looked in reasonably good shape and was installed in an old Navy RCH receiver cabinet. After a little haggling, I purchased the receiver for $125. The CR-91 was setting up on a high shelf that was a little over six feet off the floor at the rear section of the surplus store. To my surprise, the store employee (not the owner) grabbed a small (but stout) wooden box, placed it on the floor in front of the shelf, stepped up on top of the box and then proceeded to grab the CR-91 and pull it off the shelf (that was about shoulder-level when standing on a box.) Keeping his arms extended and holding the CR-91 out at chest-height, he stepped off of the wooden box and, without pausing, carried the CR-91 (still chest-high with arms extended) up to the front counter, 25 yards away. I was impressed, even though the guy did look like he could bench-press 300 pounds pretty easily.

Initial 2009-2010 Restoration - The CR-91 was missing the bottom chassis cover which isn't surprising since the receiver had been installed in the old RCH receiver cabinet. The RF cover was also gone but this is missing on almost all '88 receivers and was really only present to cover and protect the alignment adjustments. Luckily, the tuning condenser cover was installed. The 735kc crystal for the crystal filter was missing. It had apparently been "cut out" of the circuit for some unknown reason. There was a small cut-out section at the rear of the chassis that was for a Jones-type socket. Also, a SO-239 was mounted on the rear-side of the chassis. These parts were removed and, luckily, there wasn't any wiring associated with these connectors. When dismounting the front panel it was noted that several of the 1/4 x 20 studs used in the mounting the front panel to the chassis side-gussets were broken at the welds to the back of the front panel. I repaired these studs using J-B Weld with careful prep of the mounting surfaces. The resulting repair, while not as strong as the original weld, is certainly strong enough for the front panel mounting (there are eight studs and four were broken.)

photo left: The CR-91 SN: 050068 (in an Apr 2010 photo) after its rebuild and installed in an original AR-88 series cabinet (the rack screw slot edge covers aren't original.) This photo shows the replacement main tuning dial and the replacement logging dial installed. Also in the photo, the MI-8303D matching loudspeaker.

A Self-induced Problem - I hadn't yet reinstalled the front panel when I had to move the chassis off of the bench for some reason. In doing this, the unthinkable happened! I was going to carry the chassis without the front panel installed. The front of the chassis was up against my stomach and, as I was moving the receiver, I heard "snap!" I hated to look but there it was,...I had broken the main tuning dial. Bummer. I tried all sorts of repairs but the break was always obvious. Luckily, Garth Carmen, VE8NSD in Canada, was able to supply a good used-condition CR-91A main dial, which is physically an exact replacement for the CR-91 dial. Garth also supplied a logging dial that I needed for an AR-88 receiver I was restoring. I finished up by cleaning the gearbox, installing the dials and then reinstalling the front panel.

No More Custom f Crystals? - I ordered a 735kc crystal from International Crystal that took about one month to be delivered (and cost $109, even then, in 2010, mainly because it was a custom cut frequency. NOTE: International Crystal is no longer in business. In fact, new "custom frequency" crystals are becoming increasingly difficult to find,...maybe even impossible.) For some reason, I never aligned the IF or even tried to measure the active frequency of the new crystal,...reason? Unknown. 

Relegated to the Back Burner -  I then used the CR-91 a few times for testing always figuring I'd get back to completing the restoration. It turned out that the logging dial I had gotten for the AR-88 restoration "matched" the color of the CR-91A main dial I had gotten from Garth so this logging dial was installed in the CR-91 and the darker, original CR-91 logging dial was used for the AR-88. I had also acquired some other AR-88 spare parts about that time. Along with the spare parts was an original AR-88 cabinet that unfortunately had been powder-coated (it's black wrinkle,...sort of,...kind of glossy though.) About the same time, an original loudspeaker cabinet and chrome grille, the MI-8303D, showed up on eBay and was purchased. With the move to Dayton, Nevada in late-2012, the CR-91 project was put on the "back burner" and it has remained there, stored in the shop on one of the shelving-units, since that time.

photo above: The replacement dial is in very good condition

2022 CR-91 Project

Fast forward to January 2022. I had always wondered how the CR-91 would have worked using a remotely tuned loop antenna. Well, I didn't have that loop anymore but I did have a Pixel Shielded Magnetic Loop that would be interesting to try. After nine years of just setting on a shelf out in the normally unheated shop, careful checkout was going to be required and ultimately the Micamold and oil-filled tub capacitors might have to be dealt-with, resistors checked for drift and a complete alignment performed. The CR-91 had to be moved from the shop to the house and upstairs to the test & alignment bench. Since the receiver was in the cabinet, its weight was probably about 90 pounds since the bottom cover wasn't present and neither was the RF cover. Extracted from the cabinet, the chassis weighed about 70 pounds and that's not too difficult to carry upstairs. 

Chassis Photos 2022

photo above: Top of the CR-91 chassis. The square filter choke is a replacement from another type of AR-88 Series receiver. Looking at the IF transformers from left to right, 4th IF transformer and then the BFO coil, then the top pair of transformers are for the 3rd IF, next pair down is for the 2nd IF, next down is a single transformer for the 1st IF, furthest to the rear is a single can for the Crystal Filter Load coil. The slotted adjustment to the right of the Crystal Load can is the Phasing control for the Crystal Filter. The four hex standoffs surrounding the tuning condenser housing and front end assembly are for mounting the RF Cover, which is almost always missing on AR-88 and CR-91 receivers.



photo above: Under the chassis with the RF coil box shields installed. The six tub capacitors are multi-section units that are oil-filled paper dielectric types for high reliability. The other capacitors are unfortunately Micamold brand, notorious for their dielectric problems (especially after almost 80 years of aging.) The notched cut in the rear corner of the chassis is visible at the lower right. Also, note the holes for mounting a SO-239 next to the Jones-type socket cut-out. These are the only indications of a "hamster attack."

photo left: The CR-91 RF coil box with the shield covers removed. The LO is a separate shielded box located towards the front of the receiver. The Mixer, RF and Antenna coils are in their own individual compartments within the larger shielded box. The band switch uses all ceramic segments. The small diameter metal tubes are the plunger trimmer capacitors and the powdered-iron slugs for the individual coils can be seen in some of the inductors. The metal cylinder shielded coil at the upper right is the 735kc wavetrap. There are a few Micamold caps in the larger RF box (one has been replaced with a blue molded cap of some type.) The remaining caps are all "real" mica capacitors.


photo right: All of the shortwave coils used in the CR-91 are wound on polystyrene coil forms. The close-up to the right shows the Antenna Stage coils noting that the SW coils are wound on polystyrene forms to reduce losses. The MW and LF coils are wound on fiber-type coil forms. 

2022 Power up and Testing - Like a lot of WWII radio gear, the CR-91 doesn't have any electrolytic capacitors. Either oil-filled paper dielectric, just paper dielectric or mica capacitors are used. Even the 600wvdc multi-section filter capacitor is three oil-filled paper dielectric capacitors in one can. So, nothing needs to be reformed and power up is mainly watching that no shorts develop as the line voltage from the Powerstat is increased. The AC line voltage here is 122-4vac so the input voltage switch is set on "125" which should be okay for direct operation on the AC line after testing is finished. The MI-8303D loudspeaker was connected along with the antenna. I had the receiver on 40M and several strong signals were heard. Switched to 20M and also had good signal strength and no hum. I felt the chokes after about 15 minutes of operation and they were still cold. Tracking was not very good but I remember that I never aligned the front-end of the receiver. From the operation of the Crystal Filter, apparently I didn't align the IF to the crystal frequency,...even the BFO seemed to be off by about 90 degrees,...maybe I didn't do anything to this receiver other than mechanical and cosmetics. A quick test later that evening on MW with about 160 feet of wire for the antenna produced some decent signals from MOG 404kc Montegue, CA, RYN 338kc Tucson, AZ and DC 326kc Princeton, BC, CAN using the loudspeaker (not phones) although the frequency readout was substantially off. Pre-alignment Mechanical Check - The first thing to check is to be sure that the position of the tuning condenser mechanically matches the main tuning dial. There are index marks at each end of the main tuning dial that are used for mechanical alignment. The main tuning is run down to the lowest frequency end which is also the lowest numbers in the main dial logging scale. The tuning condenser has stops at full mesh and the main tuning dial should align on the low-end index when the tuning condenser is at full mesh. The top end of the tuning range should also be checked and if set correctly the top end index will align when the tuning condenser high end (full un-mesh) is reached. This mechanical alignment is very important and is often overlooked but it assures that the electronic alignment will be "as designed" and not skewed to accommodate a misaligned tuning dial.

I dug out the CR-91 manual also since there are so many adjustments. IF transformers are adjusted top and bottom so the receiver has to set on its side for the IF alignment. The manual procedure has each IF stage adjusted individually rather than just injecting the RF Signal Generator output to the Mixer grid (not that I followed the procedure.)

IF Alignment - I cover the procedure to sweep align the IF of the AR-88 in part 3 of the "The Amazing AR-88 Receiver" article on this website. Certainly, if one is looking for a flat, broad IF bandwidth in the "BROAD" position then sweeping is the way to go. I was mainly going to use the CR-91 for NDBs and 630M CW, so a standard "peak" alignment would do. Although the manual's procedure can be followed, I didn't use it except to locate where the adjustments were and for the Crystal Filter adjustments. I first found the active frequency of the crystal which was 734.8kc. I capacitive-coupled (0.1uf) the RF Signal Generator to the grid of the Mixer tube and adjusted the frequency to 734.8kc. I performed the IF alignment in SELECTIVITY position 5 which has the Crystal Filter active and most selective. Each IF transformer from IF4 to IF1 was "peaked" using an Audio Output meter connected to the 600Z audio output to measure the gain. This requires a modulated signal which the RF sig gen provided. The receiver was set to Manual Gain Control. Minimal signal levels were used for all adjustments. The Crystal Filter requires a 7kc offset for POSITION 3 crystal load adjustment. The wave trap is adjusted for minimum response to a 734.8kc signal on the Antenna terminal.

I must have never adjusted anything on the CR-91. The IF wasn't "way off" but it certainly wasn't aligned for 734.8kc either. The major IF improvement was in the Crystal Filter which is now working like it should. The Crystal Filter is a very important QRM-fighting tool it must work correctly to be a benefit to copy. Since the Phasing Control is a "chassis adjustment," finding maximum selectivity requires slight Main Tuning adjustments to locate the very narrow bandwidth center. The Crystal Filter is a "must-use" when operating AM on the 75M band and it does help a lot when scanning the MW band for NDBs.

RF Tracking - The RF Tracking alignment is tedious but at least the receiver can be returned to the right-side up position for the RF tracking alignment. The adjustments are all over the RF chassis and some are accessed through holes in the rear apron of the chassis. Without the manual it would be difficult to figure out what adjustments did what. The C adjustment is a "plunger capacitor" that has a locking collar that has to be loosened to then slide the adjustment shaft up or down as needed for the correct C. When adjustment is completed then the locking collar is tightened. The collar requires a 5/16" extra-deep socket or a 5/16" open end wrench to tighten. There were special tools supplied for pulling up or pushing down the plungers and for locking and unlocking the collar but they are almost always missing. I just use a long Allen wrench that fits the hole in the plunger along with a 5/16" open end wrench for the collars. The L adjustments are slotted threaded shafts typical for slug tuned coils, just use a small blade screwdriver to adjust.

I don't see how I ever heard anything on either Band 1 or Band 2. On Band 1, the plunger capacitor for adjusting the upper end LO wasn't tight and was all the way down. On both Band 1 and 2 all of the plunger capacitors weren't tight which allowed any upper end adjustments to not be "locked in." Most of the L adjustments were fairly close except Band 1 LO. Everything needed adjustment on Bands 1 and 2 which cover 70kc to 200kc for Band 1 and 200kc to 550kc for Band 2. Band 3 or the 1500kc to 4300kc range was moderately out of alignment. Probably some former owner only wanted a 160M and 80M receiver and the signals on those two ham bands are particularly strong anyway so the poor alignment probably didn't affect his reception too much. Band 4 or the 4300kc to 12000kc range was also moderately out of alignment. Again, with this band covering 40M, it's not too much of a surprise that it was possible to copy some signals but the dial readout was far from accurate. Band 5 or 12mc to 19mc was getting worse and the pre-testing of the receiver did show that signals were down on this band. Band 6 or 19mc to 30mc was badly out of alignment (probably why I never really heard much on that band.) On all six bands, the LO tracking was significantly off which caused the dial readout inaccuracies. After the RF Tracking alignment, the dial readout is very, very close but, like all 1940s receivers, the dial resolution is vague (the logging dial works great for f reset.)

One thing that I should have done (and it is mentioned in the "Amazing AR-88 Receiver" article) was to lubricate (with DeOxit) the shafts of the plunger caps the night before the alignment. This helps to provide smooth adjustment and eliminates erratic contact. Once I started the alignment, I noticed that most of the plunger shafts had minor oxidation. Sliding the plunger up and down a couple of times usually cleans the shaft pretty well but DeOxit works better. I did have to go back and put DeOxit on the RF2 C plunger on Band 5 to get the sensitivity back up where it should be.

photo above: This shows some of the alignment adjustments on the CR-91 with the long shafts with a hole in the end being the C adjustments and the slotted threaded rod being the L adjustments


MI-8303D - When I got this off of eBay in 2010 it was only the speaker cabinet and grille, no loudspeaker included. The paint isn't the greatest and one speaker-mounting stud is broken but MI-8303D loudspeakers are scarce and pretty difficult to find in any condition. I found a NOS (in original box) Cletron 3.2 Z ohm 8" PM loudspeaker that fit just fine and that's what I've been using all along. However, the MI-8308D originally had a back cover which was missing on this cabinet. The back was probably mild steel and 20 gauge would be an adequate thickness. The back cover mounts with 6-32 screws in each corner that thread into welded corner pieces. With the back cover installed the cabinet becomes an "infinite baffle" type of enclosure. These were popular with the military with the best known "infinite baffle" loudspeaker being the Signal Corps LS-3. NOTE: 2.5Z is specified and the manual recommends 2.5Z to 3.2Z. 4.0Z will also sound fine. 8Z will require a slight increase in the Audio Gain but speakers with a lower Z sound better. The original RCA loudspeaker voice coil was rated at 2.2Z measured at 400hz.

Using 'Phones - The CR-91 is designed to work with Hi-Z phones. When the phones are plugged into the jack the contacts connect a 5 ohm resistor across the 2.5Z audio output as a load. That way the loudspeaker is disconnected when using a set of phones. One can insert the phone plug in about half way and both the loudspeaker and the phones will be operational. It's also possible to operate 600Z phones from the 600Z output directly but the loudspeaker will be operational or a 4.7 ohm load resistor could be installed on the speaker terminals. The audio output transformer has a completely separate winding for the 600Z output and a tapped winding for the 2.5Z speaker and Hi-Z phones outputs. I tried a set of WE 518W phones but they were pretty shrill sounding. Much better response from an older set of TRIMM phones (actually a pre-WWII version.)

The TRIMM phones are actually kind of "bassy" sounding. In fact, the TRIMM phones reproduce so much bass, the diaphragms can "rattle" a bit on loud signals. However, the bass response also helps to reduce certain types of noise. Also, I've found that going to SELECTIVITY 5 SHARP really helps for reducing noise and allows hearing the audio tones that are generally around 400hz. 

Shortwave Performance - I've been listening on 20M quite a bit using the 135' tuned dipole antenna. A lot of South American DX on both SSB and CW. Also copied two stations from Finland on SSB, both OH8 calls. I've also been receiving the Coastal Station Frequency Markers CW from South Korea and from China in the 12.4mc to 12.9mc region (late-afternoon.) Trenton Military on 15.035mc was copied easily (USB from Trenton, Ontario, Canada.) Listening in the evening, I received TAH from Istanbul and SVO from Athens both around 8.4mc (CW frequency marker beacons.) Also, Gander Air USB 8mc from Newfoundland.  

MW Using the Wire Antenna - This test tuned stations from 248kc up to 408kc. I copied 20 NDBs, of which the best DX was probably WG 248kc in Winnepeg, Manitoba. I was using a set of Hi-Z 'phones for reproducers. The antenna was one leg of the 135' tuned dipole plus the feedline or about 162' of end-fed wire.

MW Using the Pixel Loop - I had to buy a 50' length of RG-6 Quad Shielded coaxial cable in order to reach the test bench without having to move the Pixel Loop. I got everything connected up Saturday at just before noon. My test listening tuned in Coastal Station KPH 425kc running their Saturday CW broadcast just before they signed off at noon. Strong signal from Point Reyes, CA during the daytime. Nighttime listening resulted in 24 NDBs tuned in 25 minutes within the range of 415kc down to 305kc. Noise level greatly reduced with the loop. I had the Selectivity on 4 with BFO offset about 400hz. Dial readout is very accurate given that the resolution is poor. Subsequent listening resulted in changing to position 5 SHARP which was much better for NDBs.

Pixel Loop vs Tuned Dipole - While the Pixel Loop is great on MW, like any loop, it doesn't provide stronger signals than a matched or resonant full size dipole antenna. In fact, the signals on the Pixel Loop are noticeable weaker than on the wire antenna. BUT, on MW, the RFI and atmospheric noise are so strong that most weak signals are completely covered up by noise. The Pixel Loop, being a shielded magnetic loop, doesn't respond very well to electrical noise so the RFI level is much lower when compared to a wire antenna. This gives a much improved "signal to noise ratio" allowing much weaker signals to be heard. However, SW reception in rural Nevada isn't plagued by RFI noise and the 135' tuned dipole antenna provides very strong signals with a low noise level in that part of the spectrum. At this QTH, on SW, the tuned dipole always out-performs the Pixel Loop. In order to take advantage of each antenna type, I hooked-up a dual antenna selector switch to allow easy selection of either the Pixel Loop or the Tuned Dipole antenna depending on the frequency of operation.

CR-91 NDB Reception Log - 2022

Jan 21, 2022 - 2205hrs to 2225hrs

WG 248kc Winnepeg, MA, CAN
AM 251kc Amarillo, TX
UNT 312kc Penticton, BC, CAN
DC 326kc Princeton, BC, CAN
MA 326kc Midland, TX
OIN 341kc Oberlin, KS
XX 344kc Abbotsford, BC, CAN
MEF 356kc Medford, OR
AL 353kc Walla Walla, WA
GGF 359kc Grant, NE
RPX 362kc Roundup, MT
6T 363kc Foremost, AB, CAN
SX 367kc Cranbrook, BC
GC 380kc Gillette, WY
CNP 383kc Chapelle, NE
PI 383kc Tyhee, ID
PNA 392kc Pinedale, WY
FN 400kc Ft. Collins, CO
MOG 404kc Montegue, CA
MW 408kc Moses Lake, WA

Cndx - Noisy with Wire Antenna, using phones, 20 stations in 20 minutes.

Jan 22, 2022  -  2200hrs to 2225hrs

GRN 414kc Gorden, NE
HRU 407kc Herrington, KS
SB 397kc San Bernardino, CA
TW 389kc Twin Falls, ID
YWB 389kc West Bank, BC, CAN
AA 365kc Fargo, ND
ODX 355kc Ord, NE
RG 350kc Will Rogers AP, OKC, OK
AFK 347kc Nebraska City, NE
YXL 346kc Sioux Lookout, ON, CAN
PMV 329kc Plattsmouth, NE
QT 332kc Thunder Bay, ON, CAN
FIS 332kc Key West, FL
RO 305kc Roswell, NM

Pixel Loop Antenna pointed E-W, good cndx, phones. 24 stations tuned in 25 minutes, 14 are new for the CR91 and shown above. One CW station on 630M sending "test" over and over - no call.

CR-91 Total is 34

Jan 25, 2022 - 2150hrs to 2220hrs

LGD 296kc LaGrande, OR
QD 284kc The Pas, MB, CAN
XE 257kc Saskatoon, SK, CAN
HLE 220kc Hailey, ID

Pixel Loop pointed E-W and N-S, cndx good, phones. Concentrated on the 200kc part of the spectrum. 10 stations copied, 4 were new to the CR-91 shown above. This region has less NDBs and the noise level is somewhat higher than the 300kc to 400kc part of the spectrum.

CR-91 Total is 38



Jan 26, 2022 - 2145hrs to 2215 hrs

Conditions were absolutely horrible. I couldn't copy any NDBs other than MOG 404kc which is an extremely strong signal here. I switched from the Pixel Loop to the wire and was only able to copy a few more NDBs. Atmospherics? Local RFI? Receiver problem?



Jan 27, 2022 - 2150hrs to 2220hrs

YAZ 359kc Van.Is., BC, CAN
PND 356kc Portland, OR
LLD 353kc Lanai City, HI
YKQ 351kc Waskaganish, QC, CAN
POA 332kc Pohoa-Hilo, HI

As bad as conditions were last night, they were just the opposite tonight, very good, low noise cndx. 24 stations copied, 5 new for the CR91 shown above. Pixel Loop was pointed E-W, phones. Best results are with the SELECTIVITY in POS. 5 SHARP - Crystal Filter Narrowest bandwidth. Receiver installed into its cabinet for better shielding from local RFI.

CR-91 Total is 43

Jan 29, 2022 - 2155hrs to 2225hrs

YYF 290kc Penticton, BC, CAN
GUY 275kc Guymon, OK
GEY 275kc Greybull, WY
VR 266kc Vancouver, BC, CAN
YCD 251kc Nanaimo, Van.Is., BC, CAN
EL 242kc El Paso, TX
ATS 414kc Artesia, NM

Good Cndx, tried the 200kc region again
with much better results, 12 stations heard and 7 of those were new to the CR91 and are shown above - Pixel Loop pointed E-W, phones, SELECTIVITY on 5.

CR-91 Total is 50

Feb 3, 2022 - 2205hrs to 2227hrs

LYI 414kc Libby, MT
JDM 408kc Colby, KS
EX 374kc Kelowna, BC, CAN
WC 332kc Abbotsford, BC, CAN
LW 257kc Kelowna, BC, CAN
QL 248kc Lethbridge, AB, CAN
XC 242kc Cranbrook, BC, CAN

Good Cndx, phones, Pixel Loop pointed N-S, 27 stations heard and 7 stations new to the CR91 shown above. SELECTIVITY on 5

CR91 Total new is 57

If the stations tuned logs included the "night-to-night" repeats, the total number of tuned signals would be 117.



Navy Department - National Company, Inc.

RBL-5   CNA-46161-B

MW, LF & VLF  TRF-Regenerative Detector Receiver from 1944

  3 TRF Stages, Regenerative Autodyne Detector, Limiter, Audio Output Stage

15kc to 600kc


National produced the RBL series of longwave TRF regenerative receivers for the Navy during WWII. Following the tradition of supplying the Navy with National NC-100A type receivers, like the RAO or RBH receiver, the RBL has the same general appearance, a similar-looking dial and a familiar-feeling band switching system. Though the band switch looks like that used on the RAO or RBH receivers, it isn't a moving coil catacomb system. The band switch actually uses a set of gears to simultaneously rotate two parallel ceramic band switch assemblies to change tuning ranges. The RBL is the same approximate size as the RAO and was certainly intended to be paired with that receiver. That pair, the RAO and the RBL, would allow reception from 15kc up to 30mc - continuous. Unlike many of the earlier LW receivers, the RBL has its own built-in power supply. It also has direct frequency readout on its illuminated dial. As with the RAO, Wells Gardner & Co. was a second contractor supplying the RBL-3 and RBL-4 versions. The RBL-6 was supplied by National in a larger, one-piece cabinet that was similar to the RAO-7 cabinet with shock mounts already installed on the bottom of the cabinet (so the cradle-type shock mount used on earlier RBLs wasn't required.) The RBL-6 panel was 19" wide but it wasn't intended for rack mounting but the construction allowed for easy removal of the receiver from the cabinet.

The RBL circuit uses a cascade of three 6SK7 TRF amplifier stages. The detector is a 6SG7 used as a regenerative autodyne detector (untuned grid input) followed by a 6H6 audio limiter circuit followed by a 6K6GT audio output stage. The earlier RBL receivers used a 5U4G as a rectifier but the later versions changed that to a 5Y3GT. An Audio Filter was included in the circuit that allowed a "BROAD" or "SHARP" selectivity (the switch is located below the "ON-OFF" switch.) An adjustable Output Limiter circuit could be employed by using the two controls to the right of the dial. The RBL Output Limiter is one of the best operating OLs and will reduce static crashes to a whisper without distorting the audio excessively. The direct frequency readout dial is one of the RBLs advantages over older types of LW receivers that have 0-100 dials and frequency charts. The RBL illuminated dial is very accurate but like all 1940s receivers, the resolution of the dial scale is vague at best. The lower controls on the receiver are (l to r) GAIN, REGENERATION, BAND CHANGE, ANTENNA TRIM, OSCILLATION push button and FREQUENCY TRIM.

This RBL-5 was acquired from a neighbor when I lived in Virginia City, Nevada. It required a little bit of work before it was performing to specifications. The tubular Navy coaxial antenna connection input that attaches to the box that bolts to the back of the cabinet had an internal short which resulted in any antenna connected to the receiver at that point was essentially shorted to chassis. I removed the tubular connector and just ran the coaxial cable through the box to connect to the antenna input terminals on the rear of the chassis to correct the problem. Underneath the chassis there were a few wires on the audio output transformer that had been soldered to different terminals under the chassis as an attempt by a former owner to unsuccessfully correct the problem and these wires had to be returned to their proper connections in the circuit and resoldered. This got the RBL-5 working quite well. An alignment, although not really needed, was performed.


photo left: Inside the RBL-5 is in immaculate and original condition. RF stages are to the right of the tuning condenser. Power supply, detector, limiter and audio stage are to the left of the tuning condenser.



I've logged a lot of NDBs with this RBL-5 primarily because it's very easy to set up and use, it is very sensitive and the direct-frequency readout really helps in locating stations. The Output Limiter is one of the best found in WWII receivers. The OL will reduce static crashes to the point where long listening sessions don't result in headaches (or ringing ears.) I use 600Z phones connected to the PHONES jack on the front panel for best copy on weak signals (like all DX NDBs are.)

I used this RBL-5 with a remotely-tuned loop antenna that I built when I was in Virginia City. Using both the Loop Tuning and the Antenna Trim on the RBL-5 allowed slight enhancing of multiple signals that were on the same frequency, like a lot of NDBs are, and this usually allowed copy of all of the stations on frequency (unless their timing was exactly the same but that was rare.)

I have yet to try the RBL-5 with the Pixel Loop antenna.




photo right: Underneath the chassis is also immaculate. Note the gears that actuate the band switching.

Shown in the photos right and left is the Wells Gardner & Co. RBL-3. This receiver is also in excellent condition. Note the differences in the chokes and transformers used in the WG version. It was thought that WG used National parts for their contracted versions but the parts are actually made by sub-contractor companies in the Chicago area for WG and are very close (but not exact) copies of the National parts. Even the knobs aren't National-built but are copies made by the sub-contractor companies in Chicago (where Wells Gardner & Co. was located.) Note that the rectifier tube in the RBL-3 is a 5U4G.  This RBL-3 is owned by NU6F of Reno, Nevada.



Radiomarine Corporation of America


Model AR-8510

MW, LF & VLF Shipboard Receiver - 1944

15kc  to  600kc

2 TRF Stages + Regen Det + 2 AF Stages


The AR-8510 was the replacement receiver for the AR-8503 (aka RAZ-1 for the USN - profiled in a section above) and is a five tube regenerative receiver that tunes from 15kc up to 650kc in four tuning ranges. Two TRF amplifiers are used with a Regenerative Detector and two stages of audio amplification. The RF amplifiers use a combination of tuned grid input and tuned plate output using a three-section ganged tuning capacitor. The antenna switch allows the user to select which receiver will be connected to the antenna - either the AR-8510 or an emergency receiver. The audio output can drive the panel mounted loud speaker or headsets either simultaneously or, using the Loudspeaker switch, the panel speaker can be turned off. The receiver requires a separate power source of which many types were available. Various types of battery combinations could be utilized with either the RM-2 or the RM-4 Battery Control panels. These functioned on ships that provided 115vdc or 230vdc power. If 115vac was to be used then the RM-23 Rectifier (power supply) was used. There was also an RM-37A unit that provided 90vdc B+ output with a 115vdc input from the ship's power. This was to be used if it was necessary to conserve the B batteries that normally provided the +90vdc. The AR-8510 requires 6.3 volts at 1.8A (AC or DC) and 90vdc at 15mA. The vacuum tubes used are four 6SK7 tubes and one 6V6G or GT.

photo right
: Top of the chassis showing the antenna connections (far left front of chassis) and the power input connections (far right back of chassis.) The tuning condenser is under the central cover.

photo above: Under the chassis showing the bee's wax impregnated coils.

The AR-8510 could be provided with a cabinet and shock mounts if it was to be used as a "stand alone" receiver. However, if it was going to be installed into a shipboard communications console (as most were) then the cabinet and shock mounts were not provided. Many AR-8510 receivers were part of the shipboard 3U transmitter console that included a 200W transmitter, an emergency crystal receiver, battery charger switching, clocks and more. The 4U console used the RMCA AR-8506 MW-SW receiver with a larger transmitter. The 5U console had both receivers installed along with transmitters and auxiliary equipment. Mackay Radio and Telegraph Company also supplied Marine Radio Consoles MRU-19 or 20 that had their equipment installed.

The AR-8510 was approved by the FCC for shipboard use in 1942 (concerning minimum radiation from the antenna.) The schematic drawings are dated 1943. It's likely that it was at least 1944 before any AR-8510s were in use and this particular AR-8510 is dated NOV. 1944 (with a serial number of 2774) making it an early example. It seems that most of the installations during WWII were onboard Liberty ships. Post-WWII installations were generally on commercial ships. The AR-8510 found a lot of use and longevity with production unbelievably still going on in 1960. Some receivers were still in use onboard some old oil tankers as late as the 1990s. Most ship owners wouldn't replace the radio gear as long as it still functioned or could be repaired.

Unfortunately, most AR-8510 receivers led a pretty hard life and the sea environment didn't help preservation. Most examples have been worked on or have missing parts (or non-original parts.) The AR-8510 shown in the header photo is cosmetically restored with nearly all original parts. The exception is one capacitor under the chassis, the speaker grille and the RCA pointer knobs. The paint job on the front panel is VHT Gray wrinkle finish which is slightly darker than the original RMCA gray.

Later manuals and some Internet sites will show a slightly different AR-8510 that has embossed nomenclature on the panel including the information on the data plate embossed onto the panel in the upper right part of the panel. The B&W photo (shown right) in the 1950 manual shows this later version with a date on the panel of 1947. It's probable that the WWII version (early version) used the easy-to-replace nomenclature tags as an ease-of-maintenance function. Later post-war receivers were probably not going to be subjected to the rigors that the wartime versions experienced so the embossed panels could be used and provided an excellent appearance.

I was given the AR-8510 shown in the photos as payment for some radio repair work. It probably was taken off of one of the Liberty ships that were part of the "moth-balled" fleet that was moored outside of Benicia, California since the receiver originally was obtained from the SF Bay Area. The "as received" condition was fairly good considering how the ships were taken care of - they weren't. Of course, the front panel has been repainted in the past - probably with a brush. The perf-metal grille had more than its share of paint applied (and it wasn't original either.) The receiver came without any type of power supply (the RAZ-1 power unit RM-6 can be used as a power source.)

This AR-8510 required a little bit of work to get it operational. Bands 1 and 2 functioned okay but needed alignment. Band 3 and 4 were non-functional due to broken leads from the coils that are in the plate circuit. The open coils resulted in an absence of plate voltage to the first RF amplifier when bands 3 or 4 were selected. I had to remove the coils and rebuild them then finishing them off with a re-waxing job. After reinstalling, bands 3 and 4 had to be aligned.

Performance using a "T" antenna of 98 vertical feet running to a 135 foot horizontal section was very good. Since the AR-8510 was the replacement for the AR-8503, it's fair to compare the two receivers. First, with a direct readout dial there's no need for the charts and graphs that are necessary for finding where you're tuned on the AR-8503. The preselector is built-in with the AR-8510. Also, only a single band switch is necessary on the AR-8510 while two band switches are used on the AR-8503 plus a band switch on the preselector. Sensitivity on the AR-8510 is about the same as the AR-8503 with preselector. Regeneration action is very similar in that it's a very sharp adjustment between maximum sensitivity (either non-oscillating or oscillating) and any adjustment below either point greatly reduces sensitivity (this is typical of regenerative detectors though.) The AR-8510 seems to hold its adjustments better across the band especially the Trimmers that only require a slight adjustment from one band end to the other. This is expected since it's part of the alignment process. I find that the loudspeaker is actually pretty good for some reception. If you want to use Hi-Z phones, it's better if you leave the loudspeaker on. Without the speaker load the 'phones seem to respond to more noise than signal.



RCA-Federal Telephone & Radio Corporation


CFT-46154, RBA-1
MW, LF & VLF Table Top Radio Receiver - 1941 contract
SN: 972     Accepted by USN 6-26-1943

CFT-46300, RBA-6
MW, LF & VLF Rack Mount Radio Receiver - 1945 contract
SN: 181   No Acceptance Tag

CFT-46154, RBA-2 SN: 168

15kc to 600kc

3 TRF Stages, Detector, Tracking BFO, 3 AF Stages

(includes NDB reception log for RBA-1 SN:972)

1943 RBA-1 CFT-46154

In the late thirties, it was becoming apparent that a replacement receiver was going to be necessary for the aging series of longwave receivers used by the Navy at that time. The receivers included the RAA superhet from 1931, the RAG TRF from 1933, the RAK TRF with regenerative detector from 1935 and the various upgraded versions of these receivers that were built on later contracts. The new LW receiver design was going to blend the advantages of the TRF plus tracking BFO design of the RAG receiver but built with modern circuits. Using a TRF with tracking BFO was advantageous in keeping the leakage radiation on the antenna to a very low level that prevented enemy direction finding equipment from determining the location of the receiver. Additionally, the low-level of radiation allowed the receiver to operate in the presence of other receiving and transmitting equipment along with radar equipment without interference. The tracking BFO design utilized a section of the main tuning condenser so the BFO tuning condenser was ganged to the main tuning. Since the new receiver was not a superheterodyne, the BFO had to track at 1kc above the tuned frequency allowing a 1kc heterodyne to be heard thus allowing CW to be readily copied. There were a couple of very good reasons for not designing the new LF receiver as a superheterodyne. First, was to provide complete coverage of the tuning range of 15kc to 600kc. Most IF amplifier sections utilized around 400kc to 500kc for the intermediate frequency, right in the middle of the most used portion of the medium wave band (as far as the Navy was concerned.) Operation of the IF amplifier at, for example 455kc, would eliminate a section of frequency coverage of about 20kc either side of the intermediate frequency. Some superhet LW receivers moved the IF above the intended tuning range (15kc to 600kc) but there were disadvantages to this solution to the problem. For example, the RBH receiver uses an IF of 1500kc but any transmitting activity around 1500kc will "leak into" the IF section of the receiver and cause heterodynes throughout the tuning ranges.

At $3000 each cost to the Navy, the new RBA receiver was certainly an expensive receiver but a look inside reveals the incredibly high level of electro-mechanical design and construction. The tuning range is from 15kc up to 600kc in four bands. The illuminated dial readout is direct in kilocycles along with a logging scale that uses two scales - one on the main dial and a separate "units" logging scale. The mechanics of the tuning system design allows for super-smooth operation. The Gain adjustment controls the RF stage sensitivity of the receiver but there is also a "tracking" gain control (also called the auxiliary gain control) that is in series with the Gain control but geared to the main tuning. This allowed a variable gain compensation to frequency tuned that provided a constant gain level across the tuning range of each band. To protect the operator's ears, since most operation would be via earphones, an Output Limiter circuit allows an adjustment of the maximum output of the receiver based on the setting of the Output Level control. The O.L. can be switched on or off via a toggle switch on the front panel. The O.L. could be used in high level static conditions or for unexpectedly strong local signals. Two meters are provided, one to monitor Audio Output Level in db and one to monitor the B+ voltage.  Two levels of audio selectivity are provided, Broad selectivity rolls off the audio response at about 1300hz with an internal LP filter and the Sharp position is provided by a 1kc bandpass filter for CW in noisy conditions or in cases of interference. Voice modulated signals sound somewhat "muffled" due to the 1300hz roll off but the only voice signals encountered anymore will be from 530kc up to 600kc (bottom of the standard AM BC band.) Below 530kc, all signals are either data streams, MCW, MSK or other types of coded signals. 630M hams occasionally operate CW. The audio output is 600 ohms Z and is intended for earphones although the RBA can drive a matched loud speaker if necessary. The audio Z can be anything from 30 Z ohms up to 600 Z ohms. Negative feedback is utilized in the last two audio stages to reduce level changes as multiple earphones were added to the output (up to 20 sets of 600Z phones could be operated by a single RBA receiver.)

1945 RBA-6, CFT-46300 SN:181 - The RBA-6 came from the factory painted Navy Gray and in a rack mount cabinet.

The separate power supply is the CRV-20130, which is the same power supply used for the 15 tube RBB and RBC superheterodyne receivers. The CRV-20130 provides the filament voltage and B+ requirements via an armored cable with heavy-duty connectors. The power supply will easily operate two RBA receivers for emergency conditions and two separate connectors are provided. The power supply has a cold-cathode regulator tube (OC3) and a HV rectifier (5U4.) The RBA uses eight tubes, three 6SK7 RF amplifiers, one 6J5 Triode Detector, one 6SK7 BFO, two 6SJ7 AF amplifiers and one 6K6 AF Output. Table top versions of the RBA receiver were identified as C(FT)-46154 (FT would indicate that Federal Tele. & Radio Corp. was the contractor) and CFT-46154-A (RBA-5) while the rack mount versions are identified as CFT-46300. Internally, all versions of the RBA receiver circuit are the same.

 photo left: Inside the RBA-6 receiver showing the immaculate condition of the top of the chassis. All of the TRF coils are in the cylindrical shielded cans to the left. The moveable covers on top of the cans allow access to the trimmers for alignment. The two front coils are for the tracking BFO. Note the gear-driven potentiometer coupled to the main tuning. This is the auxiliary gain control that allows for constant gain across the tuning ranges. Also note the shielded meters. The blue "dots" on the tops of the tubes are my indicators that I have tested these tubes and they are in good operational condition.



 photo right: The underside of the RBA-6 chassis is also immaculate. Full shielding of each RF section is provided when the bottom cover is installed. The heavy-duty band switch uses ceramic mounts with .25" silver-plated contact buttons. All components are mounted on terminal boards or are mounted directly to the chassis. Tracking BFO is the front section of coils. RF3, RF2 and RF1 are next going towards the rear. The rear-most set of coils are the Antenna Input coils.

The RBA-6 shown is from a 1945 contract. This version is identical to the RBA-5 internally but the RBA-6 is a rack mounted configuration only and is painted smooth Navy gray rather than black wrinkle. Judging by the condition of this receiver, it is unlikely that it was ever put into service. It is all original except for the substitution of an SO-239 UHF connector in place of the Navy coax connector for the antenna input. The RBA-6 is an impressive performer with ample sensitivity, direct dial read-out with illumination and a tracking BFO rather than regenerative-autodyne detector. The tracking BFO actually works quite well for finding the carrier on weak NBDs. The dial accuracy is excellent and allows tuning the NDBs by frequency rather than constantly referring to charts or graphs. The LP filter does limit the audio frequency response on BC stations but not to the point where the voice is incomprehensible. The O.L. works quite well at limiting the maximum output (which is usually due to static bursts) while not distorting the signal. The RBA-6 is a first-class longwave receiver capable of receiving any of the signals found below 500kc if used in an RF-quiet area with an appropriate antenna.

RBA-6 Performance - I have been using this RBA-6 during the morning hours for late September through most of October 2009 and have found the receiver to be a phenomenal performer. I can usually separate LLD 353kc in Hawaii from local Reno AP NDB NO 351kc and this is using just a wire antenna and not relying on the directional characteristics of a loop antenna. That's amazing selectivity for a TRF receiver. I've probably tuned in well over 100 NDBs but so far only two were new copies and they were West Coast NDBs. The wire antenna I'm using is the 135' center-fed dipole with 43' of open feed line that is shorted together at the receiver antenna terminals. This antenna, while not really something I designed for LW, seems to work quite well with all of the LW receivers. Radio Rossii 279kc is received every morning coming in very strong - Russian LW station located on Sakhalin Island. JJY also can be received every morning on 40kc coming out of Japan. Noise is the only limitation on reception and for better noise reduction I have to run a loop antenna.

photo above: This USN radio op is listening to the RBA receiver and is ready to copy on the mill. Other equipment is the General Radio LR-1 Frequency Meter, the RBB and RBC receivers. This photo of the U.S.S. Mugford in 1946

January 21,2010 - I finally got the RBA-6 up to the top floor of the house where it can be used with the six-foot loop antenna. I had tried using the loop antenna in the basement but the concrete floor and the rock foundation were a serious detriment to the loop's performance. The top floor of the house is actually about 30 feet above ground level and allows the loop to function quite well even though it is located indoors. The performance of the RBA-6 on the loop is amazing. The signals just jump out of a fairly quite background noise level. Much quieter than running on the wire antenna. Most frequencies seem to have at least three NDBs active and by tuning the loop I can usually enhance one or the other to allow copy. Quite an improvement in performance.

November 7, 2013 - QTH is now Dayton, Nevada and the antenna is a 300 foot long end-fed wire up about 50 feet. This antenna works quite well with the RBA-6 although the noise level is probably higher than using the loop. However, the actual ambient noise level is so low in Dayton that the 300 ft wire seems to provide better signal levels than the loop ever did in Virginia City. More info to come,... Nov. 8, 2013 0615 PST tuned in JJY at 40kc, very loud signal.


RBA-1, CFT-46154 - I had the receiver shown to the left for a short time. It's the early table top version RBA-1, the CFT-46154. This particular RBA-1 has had a common Field Change Modification Contract installed that changes the toggle switches for Reception mode and Output Limiter to knob-type controls. There are additional changes internally. The B+ meter has been replaced on this receiver. Additionally, like many receivers that have gone through a field rebuild, the cabinet is black wrinkle (it is for an RBA-1) while the receiver's panel has been repainted gray. This RBA-1 was traded off in 2015.

RBA-1 CFT-46154 SN: 972  -  Restoration Log

I acquired RBA-1 CFT-46154 SN: 972 (shown to the right) from a fellow in Texas in October 2017. It was shipped to Nevada via UPS. No damage was sustained during transport. I purchased this particular RBA-1 because it was the first all-original, black wrinkle finish, complete example that I had ever come across. Most RBA receivers found around here look like the RBA-1 shown in the photo above. That is, most have the Field Change modifications and have been repainted gray. Not that SN: 972 didn't have problems. For example, the DB meter glass was cracked to several places. The meter, however, worked fine. Band switching from Band 1 to Band 2 was normal. To Band 3 felt rough, like fine gear teeth meshing and switching to Band 4 was stiff but it would switch. All toggle switches seemed to function okay. There was a little bit of dial drag on the logging dial which was slightly rubbing against the housing. Inside was very nice with all shields present. Some very minor corrosion in just a few locations - nothing serious. Even the antenna input was still the original Navy coax fitting. Cosmetically, the cabinet was very scratched up with quite a bit of paint missing, especially on the top. Front panel was very good with some minor blemishes and old touch-ups. The dial mask on Band 4 needs some attention as there are some "pin points" of corrosion present. All knobs were present but the spinner was missing from the tuning knob. The two nomenclature panels were in good condition. All tags were present and in good condition on both the panel and cabinet. The original tuning chart was present with NSS (WWII Radio Central USN, Annapolis, MD,) shown at 18kc.

This RBA-1 didn't come with the CVR-20130 power supply or the armored power cable. Fortunately, I had a spare 20130 and a spare cable to apply power to this receiver. Normally, I wouldn't apply power without a thorough check out first, but the seller told me he had the receiver operating, so there was some confidence that nothing serious was going to happen. I did check all of the tubes first and ended up replacing three. I had SN:972 out in the shop, so I connected my 275 ft CF dipole with the 77 ft of feed line shorted to act as the antenna. Upon power up and warm up, the receiver was receiving KTHO 590kc up a Lake Tahoe quite loud. I had a 600Z ohm modified LS-3 connected to the phone jack on the front panel. Further tuning had MOG 404kc, a NDB from Montegue, CA and FCH 344kc from Fresno, CA coming in strong. This was about 3PM, so no DX NDBs were heard.  However, when switching to Band 3 the output was nil, same for Band 2. Switching to Band 1 the receiver again had output. NAA on 24.0kc was strong but NLK on 24.8kc was extremely strong. NPM on 21.4kc was strong from Lualualei, Hawaii.

So, this RBA-1 is somewhat functional. The most serious problem is the non-operative Band 2 and Band 3, however since the receiver works on Bands 1 and 4, the problem is certainly in the RF stages ahead of the detector. Additionally, repairing the mechanical issues of the band switching "roughness" and logging dial drag will be necessary. The DB meter glass repair will be accomplished by replacing the entire housing - it's a Weston Type 506 and they're common. Finally, when fully functional, a complete alignment. Cosmetics will be necessary too but that looks to mainly consist of cleaning and touch-up.


photo above: RBA-1 SN: 972 "as received" Oct. 11, 2017. Note the bubble-wrapped tubes piled up on the RF coils. Some of the cabinet can be seen to the right. This photo was taken before the power-up test.

Dismounting the Front Panel - This is an arduous task that is not well-described in the manual, at least in the NAVSHIPS 900,708 manual I have. In order to repair the logging dial drag, the rough feeling band switch and restore the dial mask, I had to access the cast metal housing that has the dial, the dial mask and the band switching inside. The only access to into this housing is from the front and the entire front panel has to be dismounted to even see the front of this housing. To get the front panel off I first removed all of the knobs. The ANT. COMP. and INPUT CLPG. control shafts needed to be removed by loosening the set screws at the flexible coupling for each control and withdrawing the shafts out the front panel. All controls that have panel nuts needed the nuts and washers removed. The bottom cover needed to be removed. Then the two chassis-withdrawal knobs needed to be removed. The meter covers needed to be removed so the meter leads could be disconnected. The meters can stay mounted to the panel. The tuning chart frame needed to be taken off. The screws underneath mount a resistor board to the rear of front panel and these screws had to be taken out. The PHONES jack nut has to be removed and the shield around the phone jack itself has to be dismounted by the four screws on the side of the chassis. The dress nuts and the hex nuts that mount the four toggle switches had to be removed. I had to dismount the housing around the logging dial. Underneath there are two set screws that mount the logging dial to the tuning shaft. I loosened the set screws to remove the logging dial. The spacer around the band switch shaft had to be removed. Now came the difficult part,...there were six screws across the front panel that secure the front panel directly to the chassis. These had nylock nuts on the inner chassis side. I had to use a 3/8" open end wrench to hold the nut while the screw was removed. Two nuts are difficult to access because of the harness prevents seeing the nut but it can be done by "feel." The four vertical screws on the left side also had nylock nuts to mount them. On the right side, two screws had nylock nuts and two had pem nuts. The pilot lamp assembly had a 5/8" nut on the back side. When loosened, then the red jewel front could be unscrewed and dismounted. Now the front panel could be removed. Photo right shows the RBA chassis with the panel removed. All removed parts were "bagged and tagged" in small plastic bags. This "bagging" method is extremely helpful to prevent loosing any parts and for easing reassembly.    Oct 16, 2017

Accessing the Dial Mask and Band Switching - With the front panel off, I could see the housing was a two-piece assembly. The front part of the housing was mounted with four screws. The next step is to remove the four screws and lock washers that mount the front cover of the housing. Now the front cover can be removed and this exposes the tuning dial and the band switch arced gear. The dial mask is mounted to the front cover. In addition to the dial mask there is also a gear on the back side of the mask that engages the band switch arced gear and actuates the band changing function. However, since the mask needed to be restored it had to be separated from the front cover. This required removal of the retaining collar that was mounted to the band switch shaft. The set screw wasn't a hex like all the other set screws had been. This set screw was a spline or Bristol-type. Also, just to make removal difficult, the collar was pinned to the shaft. This required driving out the securing pin with a proper size punch. If done correctly with the proper tool, the pin comes right out. The collar has to be placed on a large vise that has the jaws slightly open. With a long thin diameter punch, the pin can be driven out. The set screw was only to keep the collar in place while the pin hole was drilled and the pin driven in. It doesn't have much of a purpose now and was easy to loosen. This allowed the collar to be removed and the dial mask, band switch shaft and drive gear could be separated from the front cover.

Painting the Dial Mask - Once the dial mask was removed its condition seemed a little worse than I thought. The minor flecks of corrosion were numerous but most of them seemed to be around BAND 4. I removed the black paint with stripper and wiped everything down with denatured alcohol. I worked over the surface with steel wool to be sure it was smooth and no blemishes would show when repainted. I used Krylon Flat Black for the paint as the initial coat.

photo right : The dial mask/cover assembly removed exposes the dial and the arced band switch gear.

Restoring the Mask Nomenclature - Of course, painting the dial mask covered up the frequency band nomenclature. These letters and numbers are embossed so they are somewhat higher than the flat surface of the mask. To restore the nomenclature requires carefully removing the paint just from the top surface of the letters and numbers. I had to do the nomenclature recovery in two steps. After the initial paint in flat black, I used a special angled tool (that I made) that could hold a very small piece of 320 grit Al-Ox paper. This was used to carefully remove the paint just from the nomenclature. Of course, there was some slight scuffing and other minor blemishes that happened to the paint surrounding the nomenclature. The next step was to carefully cut masking tape pieces to exactly cover only the nomenclature. It sounds time-consuming but it only takes about 25 minutes to do. Next, I applied a coat of Satin Black, which is much darker black than flat black. When this paint had dried, I removed the masking tape. Now, only very small areas between the letters and numbers had the flat black paint showing. These areas were so small that I touched them up using a "ultra-fine point" Pilot pen with black ink. The entire mask was then wiped down with a clean cotton cloth to even out the paint surface. It sounds like a lot of work and it is fairly time-consuming but the results are worth the effort since the dial mask is quite visible, directly in front of you, on all bands, whenever you're using the receiver.     Oct 19, 2017

Dial Mask and Band Switch Reassembly - I put a small dab of grease on the mask shaft and installed into the front cover of the housing. The washer and collar were then installed onto the shaft and the collar pin installed. The spline set screw was tightened. The band switch was set on Band 2 so the mask was rotated to show Band 2 also. The front cover was carefully placed in position and moved slightly until the gears meshed while the mask remained showing Band 2 centered. The four screws were installed. The adjustment for the gear mesh clearance is accomplished by moving the front cover upwards (reduces mesh) and tightening the screws. Testing switching showed that the mesh was correct as the switching was ultra-smooth and not binding.

photo right: Dial mask after restoration

Front Panel Restoration - This is really just a thorough cleaning with a brass wire brush (suede shoe brush) and Glass Plus to remove all dirt, grime and cigarette residue. The original wrinkle finish is very tough and can take this type of cleaning but you can't be real aggressive. You just want to remove the dirt not the paint. The front panel had been stripped of all tags and the two dial covers. The silver (engraved) nomenclature needed substantial cleaning. The front panel must have been pretty dirty as it took about five cleanings until the paper towels didn't turn gray with dirt and stayed fairly white. After this many cleanings, the nomenclature was very legible now. Next is the touch-up. I've been using jet black nitrocellulose lacquer for the last year or so. This method replicates how the USN did "touch-ups" on equipment that was being repaired but was not in the depot for an echelon-type rebuild. The lacquer looks too dark when being applied but with the final wipe down with 10W machine oil, the jet black matches the original wrinkle paint color quite well. The touched-up front panel is shown to the right.   

DB Meter Repair - The meter glass was broken in this meter. Luckily, it was a Weston Type 506, which are very common meters. The mechanical zero mechanism on a 506 is mounted in the glass making replacement glass difficult, if not impossible. The easiest method is to find a similar 506 meter that has the meter case with mounting flange. It doesn't matter what the meter scale is,...all I wanted was the metal case front with the glass. I dug around in the junk boxes and found a good candidate. The swap of the cases is easy only requiring removal of the three mounting screws on each meter, aligning the meter zero adjuster-pin with the adjustment slot while sliding the new housing glass in place and then reinstalling the screws. The only difference in the meter I used was that the glass was standard while the original USN meter glass had a "non-glare" treatment. Other than that, the transplant was perfect.      Oct 20, 2017

Plastic Used in Logging Dial Window and Tuning Chart Cover - These transparent plastic pieces are made of a celluloid-type of plastic that is very sensitive to water. Just plain old H2O will "fog" the plastic. The more water added, the worse the fogging gets. With enough water, the surface of the plastic will begin to dissolve. The only way to clean the plastic used in the Logging Dial cover and the plastic used for the Tuning Chart cover is to use oil. I used 10W machine oil (3 in 1 oil, actually.) The oil cleaned the plastic surface and didn't cause any change in the transparency. Another note on the Logging Dial window, if this piece is dismounted only leave it dismounted for enough time to clean it and the dial housing. This plastic piece is fairly thick and if not secured by mounting screws it will quickly begin to warp. I noticed the warp starting in about 15 minutes, so I remounted the plastic to allow it to retain its proper shape. Front Panel Remount - I had to touch-up all of the panel mounting screws with black lacquer. Several were missing all of their paint but most just had chips of missing paint. I mounted the two meters after I had touched-up their mounting screws (and washers.) The most difficult part of the front panel remounting is installing the two chassis shims. These two pieces are mounted by the six horizontal chassis screws. I had to have the six screws protruding thru their respective panel holes to hold the two shims in place while I guided the front panel into the proper position to fully seat against the chassis and the side panels. I also had to guide the four toggle switches along with the Gain control and the Output Level control thru their respective panel holes in order for the panel to fully seat. Once the panel is seated all of the other panel screws were installed. All but two of the panel to chassis mounting screws have nylock nuts that complicate the panel installation, especially the six horizontal chassis screws that have four of the nuts difficult to access with a wrench. I first installed the other panel screws and nuts but didn't fully tighten them. They were holding the panel in place but allowed for minor movement. For the six horizontal screws, the easiest method was to have the receiver upside down as this allowed the best view of where the nuts had to go. Once the nuts were installed I still didn't fully tighten the screws yet. I still had to install the screws that mounted the chassis withdrawal knobs. Once I had all of the panel mounting screws and nuts installed and threaded together, then I could go ahead and tightened all of the panel screws.

Each toggle switch had an flat washer, a hex nut and a dress nut for their mounting. The switches and two controls use external star locking washers that are mounted on the back side of the panel and flat washers and hex nuts are used on the front. The phone jack uses a special fiber insulator that has a locating pin that has to be inserted into the mating hole in the back of the panel. The phone jack body front just fits into the recessed channel which keeps the phone jack from turning when the panel nut is tightened. I had to make sure that two fiber washers and one metal dress washer were used when installing the phone jack nut. This was to provide the correct chassis insulation and the proper spacing so the phone jack nut would tighten up.

Reassembly Notes - When I installed the metal spacer for the band switch shaft I had to make sure that the spring washer was mounted on the band switch shaft with the "fingers" facing up so it pressed against the recessed area behind the spacer. This spring provides a thrust load on the band switch shaft. When installing the logging dial, it also had a spring that mounted on the tuning shaft and loaded the logging dial. To synchronize the logging dial, I first had to set the main tuning to the minimum frequency on any band. I noted that the logging scale actually is a little longer than the tuning scales but I had to set the main dial to "0" on its logging scale. I then installed the logging dial on the tuning shaft with the load spring in place. I set the logging dial to read "0" on its scale and pressed the logging dial firmly as far back onto the shaft to load the spring and then tightened the set screws. I installed the logging dial cover and checked the operation for any rubbing or binding while tuning. I also checked that each 100s segments on the main dial's logging scale matched the logging dial at "0" or within a couple of divisions of "0." I then installed the meter covers and then cleaned and installed the knobs.

I was pretty sure that during reassembly, with tightening up the panel screws and the other mounting screws, some paint chips were going to "pop off." I ended up with two fairly large chips and several smaller chips. These were touched up with black lacquer.

With everything back in place, retesting and troubleshooting the Band 2 and Band 3 problems could begin.    Oct 22, 2017

Minor Troubleshooting - Problem one was with the DC Voltmeter reading very low. Pressing on the meter face gently would result in the voltage reading going up to normal, +220vdc. The problem was that I just reinstalled the lug connections and didn't clean them first. Cleaning with a brass wire brush got the terminals and lugs clean and when reassembled, the voltage reading was normal.

Problem two was that the ANT. COMP. control wasn't indicating the air variable position correctly. The ANT. INPUT shield has to be removed (easy to do - only four screws) to see the plates of the condenser. I set the condenser per the manual,...full mesh is 10 to the right side of the scale. There's a rotation limiting pin on the extension shaft thru the front panel that only allows 180 deg. of rotation so the air variable position has to match the panel scale.

On to the serious problems,...

Repairing Split Hex-Collars on Air Variable Trimmer Capacitors - So far, the problems had been minor stuff. I was now going to find out why Band 2 and Band 3 were inoperative. I could hear a few signals on Band 3 but they were very weak. Band 2 was dead. I used a RF Signal Generator to inject a RMS voltage into the Antenna Input to act as a steady, high-level signal. When adjusting the trimmers to resonate the RF coils, the problem became "painfully" obvious,...on Band 2, two of the air variable trimmers had split hex-collars and the trimmer rotor had dropped down and was "shorting" the rotor to stator plates which shorts the RF coil which in turn takes the grid of that RF amplifier tube to ground. I adjusted the rotors until they were fully un-meshed which removes the short circuit but leaves that RF coil "not tuned" and the rotor too low to mesh with the stator. Enough signal leaks thru to tell that these are the defective components keeping Band 2 from operating. I found that Band 3 had one split hex-collar trimmer that was already un-meshed. This was why Band 3 seemed to work but had weak signals. Only Band 1 and Band 4 had all good trimmers.

The repair of these split hex-collar trimmers requires the removal of the defective RF coil assemblies. First remove the shield-can by removing the top nylock nut and the four screws at the base of the shield. Removal of the actual RF coil assembly can be fairly difficult due to the 16 gauge buss wire used for connections. This requires a fairly large soldering iron (or gun, I used a Weller 250W gun.) On most of the buss connections the wire is wrapped a half turn around the terminal. On some wires, the joint can be heated and the buss wire lifted off of the terminal with large needle nose pliers. Other joints may require the use of a small pointed tool (like a small awl) to get the buss wire bend started once the solder was molten. At this point, long needle nose pliers can slightly unwrap the buss wire from the terminal and then it can be lifted off of the terminal. There are four buss wires and two stranded wires per coil assembly that need to be removed. Once desoldered, then the nylock nuts can be taken off and the coil assembly removed. Be careful when doing the reinstallation of the buss wire onto the terminals. If the buss wire was unwrapped too much it may break when bending it back into position. Try to only unbend the buss wire slightly when removing and then no problems should be encountered when reinstalling.

photo above
: The hex-collar on the right is correct and not split. The hex-collar on the left is split and has been taken off to show the trimmer rotor shaft.

photo above:   Cardboard shims are installed between the rotor and stator plates to maintain proper spacing for repair

An ideal hex-collar repair would require access to a machine shop to make a new hex-collar from 1/4" brass hex stock. The shaft on the trimmer condenser is slightly tapered from 0.152" up to 0.154" and this requires that the hex stock have a slight taper also. The height of the collar is also important since the press fitting of the collar will stop when the adjusting slot of the rotor is at the top of the collar. This fit then has to result in the proper spacing of the rotor and stator plates. Original dimension for the height of the hex-collar was ~0.190" and the new collar would also have this same dimension. The final measuring involves having the hole perpendicular to the top and bottom surfaces of the hex-collar to allow even movement of the adjustment without binding.

It's also possible to correctly position the split hex collar on the shaft and then drill and pin the collar to the shaft. The split should be perpendicular to drill hole for highest strength. A very small diameter steel pin will be necessary since the shaft is only about 0.154" in diameter. The pin should be an easy "press fit" since it's only holding the collar in position on the shaft.

Either of these two repair methods are impractical for most restorers because we don't have access to a machine shop. The next repair method described doesn't require any special tools.

An easier repair of the split hex-collar is to use solder to hold the hex-collar in place with the proper spacing for the trimmer plates. Cardboard shim stock installed between the rotor and stator bottom plates will provide a firm stationary spacing for the plates. A minimum of four shims should be used for best support of the rotor plates. I used .022" thick cardboard that can be salvaged from small boxes or the cardboard backing from tablets or from many other sources. As long as the shim thickness is about .022" the plate spacing will be correct.

The first step is to remove the RF coil assembly from the receiver. I've tried to do this repair with the coil assembly in place (just the shield-can removed) and it is almost impossible to get the cardboard shims installed. There's just not enough space for access and there's limited visibility, making the whole operation impossible. Although it's difficult to unsolder the 16 gauge buss wiring to remove the RF coil assembly, it's easier than trying to do this procedure with the coil assembly still mounted in the receiver.

Once the RF coil is removed, installing the shims is easy. I use a pair of channel locks to gently lift the rotor (with the hex collar installed so it will only move just high enough.) I then slide in four shims. Now, release the channel locks and remove the hex-collar. The slotted end shaft will be at the correct height (for the hex-collar to be sweat-soldered in place.) The slotted end of the shaft has to be cleaned with a small jeweler's file but don't clean all the way down to the stationary spacer. Just the upper half of the shaft should be cleaned. Add some slight grooves with the file for the solder to adhere to better. The inside of the split hex-collar must also be cleaned with the small file.

I used a large soldering iron to "tin" the slotted shaft (250W Weller.) If the hex-collar now won't fit on the shaft then file off some of the solder until the hex-collar does fit entirely onto the shaft with the bottom of the hex-collar flush with the stationary spacer. Heat the hex-collar and shaft with the soldering iron until it's hot enough that a very small amount of solder can be melted. This solder will be wicked between the hex-collar and the shaft and should provide a very strong bond. Add a little more solder to the top slot area and create a slight solder "dome" on the top. Let the solder cool and remove the shims.

I used a Dremel tool with a small cut-off wheel to make a slot in the solder. This slot has to be widened by gently using a hacksaw blade (blade alone - not the entire hacksaw.) Test the trimmer rotation using a blade screw driver. It should now adjust correctly. Check all of the wires and components on the RF coil assembly to make sure there's been no damage during the repair. If it looks okay, then reinstall the RF coil assembly into the receiver.  Oct 27, 2017

photo right:   Finished hex-collar repair showing the slot cut to allow adjustment of the trimmer when RF coil is fully assembled.

Another Problem with Band 3 - Band 2 was performing very well after repairing the two defective trimmers. However, Band 3 still seemed to be lacking sensitivity. Injecting a modulated signal into the Antenna Input revealed that the RF trimmers were all working correctly on Band 3. Switching off the modulation and switching to CW (BFO on) should have resulted in a heterodyne signal but no change was noted in the signal. I adjusted the frequency to the top end of Band 3 (sig gen and receiver) and adjusted the BFO trimmer and what happened wasn't unexpected. Of course, the trimmer rotor dropped as soon as it was unmeshed. The non-operational BFO had seemingly reduced the sensitivity because no signals could be demodulated. To repair the BFO on Band 3 is very similar to repairing the split hex-collars on the RF coils since it's exactly the same type of trimmer.   >>> >>>   The BFO coil assembly for Band 3 and Band 4 was removed from the receiver. The trimmer was repaired per the procedure shown above. After the BFO assembly was reinstalled, the BFO didn't work on either Band 3 or Band 4. This could only have been caused by the common return for both trimmer capacitors. I ended up removing the BFO coil assembly again, just to check it over closely. I found a questionable solder joint on the common buss wire return that looked like it might have been cracked but that should have only affected Band 3. Before I reinstalled the coil assembly, I carefully resoldered all of the coil wire wraps to the terminals as a precaution. I installed the coil assembly and soldered the four buss wire connections and the stranded wire return connection. Upon testing, both BFOs worked. The problem was probably due to the coil wire wraps around the terminals as these looked like questionable solder joints. The shield cover was installed and then the BFOs adjusted for 1kc above the tuned frequency on Band 3 and Band 4.         Oct 29, 2017
Trimmer Capacitors Split Hex-collars -  Notes - This splitting of the hex-collar on this style of trimmer condenser is fairly common. This type of trimmer was very popular in the mid-thirties up into the mid-fifties. They're found in Hammarlund Super-Pro receivers (they are a Hammarlund part - "ACP" Trimmer,) all BC-348 receivers other than Q, N and J versions, Navy RBB and RBC receivers, on and on. It's likely the problem is due to poor storage conditions and metal fatigue due to age. When the metal was new the hex-collar could expand and contract with varying temperature cycles. However, years and years of thermal cycling will cause the metal to become brittle and eventually it will crack. Poor storage in garages or sheds only results in thermal cycles that are even more extreme and rapidly cause the metal to become brittle. There isn't much that can be done today except to repair the trimmers as they are encountered.
Alignment - The hex-collar trimmer repairs resulted in the RBA-1 now being fully functional. The receiver requires a specific "dummy antenna" to load the input regardless of the impedance of the signal generator. Luckily, the schematic of the "dummy antenna" is included in the NAVSHIPS manual. The parts required are one 20uH inductor, one 400 ohm carbon resistor, one 200pf capacitor and one 400pf capacitor. The schematic is shown to the right.

The trimmers are all adjusted at the top end of each band. For example, on Band 4, set signal generator to 600kc. Set receiver to 600kc. With signal input to antenna input thru dummy antenna, adjust all Band 4 trimmers for maximum deflection of the DB meter. Reduce the GAIN control to compensate as the signal level increases. Use lowest signal input that gives an indication on the DB meter set at the lowest scaling. This is repeated for each band.

The BFO is aligned at the top end and the bottom end of each band. With 600kc unmodulated input and receiver tuned to 600kc, switch on BFO. A 1kc heterodyne should be heard. Adjust BFO 4 trimmer to 1kc heterodyne, if necessary. Tune to 235kc and set generator to 235kc. A 1kc heterodyne should be heard.   >>>

>>>   There are adjustments for the BFO inductance for each band that are "thru the front panel." Two holes are under the tuning chart (Left = BFO 3 and Right = BFO 1) and the two other holes have acorn head screw plugs that are removed to access the holes (Left = BFO 4 and Right = BFO 2.) Adjust BFO 4 inductance for a 1kc heterodyne, if necessary. Check both the upper and the lower frequencies to verify that the BFO is actually 1kc higher than the tuned frequency. Tune the receiver up about 1kc higher in frequency and the BFO should be zero beat if it is set 1kc higher than the tuned frequency. Repeat the adjustments until BFO tracks. Repeat the procedure for each band.

Is the Dummy Antenna Really Necessary? - When the RBA-1 was in service and being aligned at a USN depot, it was unknown to the radio repair technicians what type of antenna was going to be connected to the receiver. Shipboard antennae were quite different from USN land station antennae. What the Dummy Antenna does is provide a somewhat constant impedance load to the receiver that behaves more like a typical antenna would than a direct connection to the Signal Generator output would. The variability of what type of signal generator was used probably also entered into the alignment problems. Using the Dummy Antenna generally ensured that every RBA would come thru alignment with the same performance specification regardless of where it was going and what type of antenna it was going to be used with. Today, most of us are going to use one specific antenna with the receiver. If it's an end-fed wire of considerable length, it might be better to align the receiver with this antenna connected. Couple a signal generator output to the receiver's antenna input by using a small "antenna" on the signal generator output consisting of a three foot wire that's near the receiver. This will provide a frequency specific signal that can be adjusted for amplitude, frequency and modulation (if needed.) This simulates the receiver actually "picking up" a transmitted signal and aligning the receiver for the best operation with the antenna the is going to be used. In this type of set-up, the load of the antenna is just as it is when the receiver is in operation. This then provides an alignment that is exactly for the actual antenna that is going to be used. Tuned loops are a bit more difficult and experimentation would probably be best with these types of antennae. Usually, the loop can be tuned for the highest frequency on the band and then the receiver aligned to that impedance load. Since there is no alignment adjustment for the low end of the band, check to see if when the loop and the receiver are tuned to the low end of the band that the performance is the same as the upper end. When experimenting with various antennae and alignment, only the antenna coils will be affected and only those coils will require adjusting.
Cabinet Restoration - I really didn't want to do an entire respray of the cabinet. If I did, it wouldn't match the receiver front panel. There was a lot of paint missing, especially on the top. The first step was to remove the two data plates and then to clean the cabinet's remaining paint with a brass wire brush and Glass Plus. I then used black nitrocellulose lacquer to repaint all of the areas that were bare metal. I let this set for a day. The next step was to mix a very thin batch of lacquer and thinner - about 2:1 ratio. With gloves on, I then used a cotton pad saturated in thinned lacquer to "go-over" the entire cabinet. This hides the "touch-ups" and ends up with the cabinet looking in good original (although "touched-up") condition. I then reinstalled the data plates and installed the receiver into the cabinet.  Antenna Connection to Navy Coax Fitting - These Navy coax fittings are not common making them difficult to find. There are two easy ways to connect an antenna feed line to the RBA (or the RBB or RBC) that allow use of coaxial cable and provide a good connection to the receiver. Use a double female SO-239 adaptor. The outer diameter will just fit into the Navy coax barrel. It will slightly "thread" itself and make good electrical contact. However the pin at the bottom of the Navy fitting is too far away to make positive contact with the center conductor of the SO-239 adapter. I usually roll .010" thick brass or copper sheet metal into a tube. The tube can then be inserted into the SO-239 adapter center barrel. The other end of the tube should be about a .375" extension and the open end of it will fit over the Navy center conductor pin.

Another way to connect coax is to use a female banana pin plug (these do fit onto the Navy center conductor pin.) The coax center conductor can be soldered to the female banana plug. The shield can have a drain wire soldered to it and it can be connected to the ground lug on the back of the RBA cabinet. 

Performance - Since I had the RBA-1 set up out in the shop, I had to use the 270 foot long, center fed dipole. Since the antenna is fed with open transmission line, I just disconnected the two wires from the antenna coupler and, using clip leads, shorted them together and then used a single 14 gauge stranded wire to connect the feed line to the receiver. Perhaps this acts like a large "T" antenna that has the 77 feet of feed line acting somewhat as a vertical with a large 270 foot long "top hat." The "T" configuration was a popular antenna during the early days of radio.

During some of my initial testing and troubleshooting I tuned in BO 359kc, the NDB from Boise, Idaho - and this was during the day! Quite a surprise,...but when I aligned Band 3 to the antenna, I tuned in HDN 211kc in Gooding , ID, HLE 220kc in Hailey, ID and TCY 204kc in Tracy, CA - during mid-day! Additionally, the normal daytime copy of FCH 344kc in Fresno, CA, MOG 404kc in Montague, CA and CC 335kc in Concord, CA were all present. None of these were newly copied NDBs but I've never heard BO, HLE, HDN or TCY during the day. The next day, I tuned in DC 326kc in Princeton, BC, Canada. Incredible! This was at 13:30 in the afternoon and I was hearing British Columbia. Another note, these stations were being copied on loudspeaker, not 'phones. More experimenting with other LW receivers will be necessary to determine if this is RBA-1 performance (unlikely) or if it's antenna performance (likely.)

Other daytime copies included the USN MSK stations NLK 24.8kc in Jim Creek, WA, NAA 24.0kc in Cutler, ME, NPM 21.4kc in Lualualei, HI and, although substantially weaker, HOLT 19.8kc in Exmouth, Australia. These are very strong stations that are easy copy, day or night.

The INPUT CLPG and ANT COMP. controls are very important and must be adjusted for each band. In fact, the ANT COMP. should be peaked every 10kc or so.

The OUTPUT control and OL switch are used as an audio AVC. Once set, the audio output won't increase beyond the level set. When there's a lot of static crashes or "popping" noises, the OUTPUT (limiter) control helps keep your ear drums intact when using 'phones.    Nov 1, 2017

The following is a log of the stations tuned with the RBA-1 using the 270' CF with 77' of ladder line with wires shorted together. There are six listening sessions within about a seven week period that include both daytime and nighttime listening periods. With the six sessions, I tuned in 100 different NDBs of which six were newly heard stations. Greatest DX was DDP 391kc in Puerto Rico which isn't unexpected since it runs 2KW. Other DX (for a western US location) was FIS 332kc in Key West, FL, YMW 366kc in Maniwaki, Quebec and DB 341kc in Burwash Landing, Yukon. Daytime copy of British Columbia stations YZA 236kc and DC 326kc was unexpected but was certainly due to the time of year. Best LW reception is usually in November and December. The large antenna was probably also a factor. Maybe the RBA-1 was a factor, too.

RBA-1, CFT-46154, SN: 972   NDB Log

Antenna is 270' CF with 77' ladder line with feedline wires tied together.   Height is 40 feet.   Stations are listed in "as heard" order.   Pacific Time.   Newly Heard NDBs - Maroon with asterisk. Unless otherwise indicated, all reception is using 600Z ohm 'phones as audio reproducers.

Nov. 3, 2017  05:55-06:35 - Cndx Xlnt, Quiet

Nov. 9, 2017 19:10-19:40 - Cndx Poor, Noisy

Nov. 18, 2017  13:30 - Cndx Xlnt

Dec 23, 2017 19:05-19:50 Cndx Xlnt

1. UAB -200kc - Anahim Lake, BC, CAN
2. IP - 201kc - Mobile, AZ
3. YBL - 203kc - Campbell River, BC, CAN
4. TCY - 204kc - Tracy, CA
5. HDG - 211kc - Gooding, ID
6. LU - 213kc - Abbottsford, BC, CAN
7. QU - 221kc - Grand Prairie, AB, CAN  *
8. HLE - 221kc - Hailey, ID
9. YKA - 223kc - Kamloops, BC, CAN
10. LWG - 225kc - Corvallis, OR  *
11. CG - 227kc - Castlegar, BC, CAN
12. VG - 230kc - Vermillion, MB, CAN  *
13. OKS - 233kc - Oshkosh, NE
14. BR - 233kc - Brandon, MB, CAN
15. OJ - 240kc - High Level, AB, CAN
16. AVQ - 245kc - Tucson, AZ
17. YCD - 251kc - Vancouver Is., BC, CAN
18. LW - 257kc - Kelowna, BC, CAN
19. HCY - 257kc - Cowley, WY
20. SLE - 266kc - Salem, OR
21. YK - 269kc - Castlegar, BC, CAN
22. XS - 272kc - Prince George, BC, CAN
23. GYZ - 280kc - Guernsey, WY - USCG
24. CEP - 278kc - Ruidoso, NM
25 QD - 284kc - The Pas, MB, CAN
26. EKS - 286kc - Ennis, MT
27. YYF - 290kc - Penticton, BC, CAN
28. TOR - 293kc - Torrington, WY  *
29. TV - 299kc - Turner Valley, AB, CAN
30. ONO - 305kc - Ontario, OR
31. UNT - 312kc - Penticton, BC, CAN  *
32. DC - 326kc - Princeton, BC, CAN
33. POA - 332kc - Hilo, HI
34. RYN - 338kc - Tucson, AZ
35. ZU - 338kc - Whitecourt, BC, CAN
36. DB - 341kc - Burwash Landing, YK, CAN
37. BKU - 344kc - Baker, MT
38. FCH - 344kc - Fresno, CA
39. SBX - 347kc - Shelby, MT
40. NY - 350kc - Enderby, BC, CAN
41. AL - 353kc - Walla Walla, WA
42. MEF - 356kc - Medford, OR
43. BO - 359kc - Boise, ID
44. YQZ - 359kc - Quesnel, BC, CAN
45. RPX - 362kc - Roundup, MT
46. 6T - 362kc - Foremost, AB, CAN
47. DPY - 365kc - Deer Park, WA

WH2XVN copied on 183kc at 05:57

1. ZQ - 410kc - Sir Wilfred Laurier CCGS, BC, CAN
2. MW - 408kc - Moses Lake, WA
3. MOG - 404kc - Montegue, CA
4. QQ - 400kc - Comox, Vancouver Is, BC, CAN
5. SB - 397kc - San Bernardino, CA
6. ULS - 395kc - Ulysses, KS
7. DDP - 391kc - San Juan, Puerto Rico
8. YWB - 389kc - West Bank, BC, CAN
9. JW - 388kc - Pigeon Lake, AB, CAN
10. WL - 385kc - Williams Lake, BC, CAN
11. PI - 383kc - Tyhee, ID
12. GC - 380kc - Gillette, WY
13. OT - 378kc - Bend, OR
14. EHA - 377kc - Elkhart, KS
15. EX - 374kc - Kelowna, BC, CAN
16. LV - 374kc - Livermore, CA
17. MF - 373kc - Rogue Valley, OR
18. HQG - 365kc - Hugoton, KS
19. DPY - 365kc - Deer Park, WA (2nd copy)
20. RPX - 362kc - Roundup, MT (2nd copy)
21. BBD - 380kc - Brady, TX
22. AP - 378kc - Active Pass, BC, CAN
23. MR - 385kc - Monterey, CA
24. HAU - 386kc - Helena, MT
25. CSB - 389kc - Cambridge, NE
26. PNA - 392kc - Pinedale, WY
27. DQ - 394kc - Dawson Creek, BC, CAN
28. FN - 400kc - Ft. Collins, CO

29. MM - 411kc - not on Navaid (MW w/o the dit?)
30. INUU - 395kc - I've copied this beacon many times but have never found it on any Navaid. Oil Rig?

NOTE:  Conditions were particularly noisy due to a passing storm front and wind causing severe antenna static. One of the problems with using an outdoor antenna. Even indoor loops will respond to weather fronts however.

 Nov. 17, 2017,  19:00 - Cndx Poor

1. FS - 245kc - Sioux Falls, SD
2. YZA - 236kc - Ashcroft, BC,CAN
3. XC - 242kc - Cranbrook, BC , CAN

High level of static from passing front. Static was so bad that OL didn't help. These are new copies for RBA-1 only.

As an illustration of how fast signal cndx can change, I was running the RBA-1 in the early afternoon on the loudspeaker. Copied for long time periods the following stations,...

1. HLE -221kc - Hailey, ID
2. HDG - 211kc - Gooding, ID
3. YZA - 236kc - Ashcroft, BC, CAN

The static level was very low and these stations were easy copy on loudspeaker. This was at 1:30PM in the early afternoon!

Nov. 22, 2017   21:10-21:40 - Cndx OK

1. LU - 213kc - Abbotsford, BC, CAN
2. RL - 218kc - Red Lake, ON, CAN - #
3. QU - 221kc - Grand Prairie, AB, CAN
4. YKA - 223kc - Kamloops, BC, CAN
5. DN - 225kc - Dauphin, MB, CAN - #
6. CG - 227kc - Castlegar, BC, CAN
7. VG - 230kc - Vermillion, MB, CAN
8. YZA - 236kc - Ashcroft, BC, CAN
9. AVQ - 245kc - Tucson, AZ
10. YCD - 251kc - Van. Is, BC, CAN
11. HCY - 257kc - Cowley, WY
12. LW - 257kc - Kelowna, BC, CAN
13. YK - 269kc - Castlegar, BC, CAN
14. GUY - 275kc - Guymon, OK - #
15. CEP - 278kc - Ruidoso, NM
16. YYF - 290kc - Penticton, BC, CAN
17. TOR - 293kc - Torrington, WY
18. ONO - 305kc - Ontario, OR
19. VC - 317kc - LaRonge, SK, CAN - #
20. MA - 326kc - Midland,TX - #
21. DC - 326kc - Princeton, BC, CAN

Static level was moderate even though weather and wind were calm. Most are repeat copy but five new stations to this receiver (marked with #.)

Total NDBs copied with this RBA-1 = 100

Total newly heard NDBs = 6
1. GDV - 410kc - Glendive, MT #
2. MW - 408kc- Moses Lake, WA
3. QQ - 400kc - Comox, BC, CAN
4. ENS - 400kc - Encenada, BCN, MEX #
5. FN - 400kc - Ft. Collins, CO
6. DQ - 394kc - Dawson Creek, BC, CAN
7. YWB - 389kc - West Bank, BC, CAN
8. WL - 385kc - Williams Lake, BC, CAN
9. HAU - 386kc - Helena, MT
10. YPW - 382kc-Powell River, BC,CAN#
11. GC - 380kc - Gillette, WY
12. K2 - 378kc - Olds-Didsbury, AB, CAN#
13. EX - 374kc - Kelowna, BC, CAN
14. YK - 371kc - Yakima, WA#
15. YBV - 370kc - Berens River,MB,CAN#
16. SX - 367kc - Cranbrook, BC, CAN#
17. YMW - 366kc - Maniwaki, QC, CAN#
18. HQG - 365kc - Hugoton, KS
19. AA - 365kc - Fargo, ND#
20. BF - 362kc - Nolla-Seattle, WA #
21. PTT - 356kc - Pratt, KS#
22. IN - 353kc - International Falls, MN #
23. PG - 353kc - Portage, MB, CAN #
24. SBX - 347kc - Shelby, MT
25. MNC - 348kc - Shelton, WA #
26. FCH - 342kc - Fresno, CA
27. RYN - 338kc - Tucson, AZ
28. LF - 336kc - LaSalle, MB, CAN *
29. FIS - 332kc - Key West, FL #
30. DC - 326kc - Princeton, BC, CAN
31. MA - 326kc - Midland, TX
32. YQF - 320kc - Red Deer, AB, CAN #
33. VC - 317kc - LaRonge, SK, CAN
34. YYF - 290kc - Penticton, BC, CAN
35. EKS - 286kc - Ennis, MT
36. CEP - 278kc - Ruidoso, NM
37. XS - 272kc - Prince George, BC, CAN
38. YK - 269kc - Castlegar, BC, CAN
39. ZSJ - 258kc - Sandy Lake, ON, CAN #
40. ZYC - 254kc - Calgary, AB, CAN #
41. YCD - 251kc - Vancouver Is.,BC, CAN
42. FS - 245kc - Sioux Falls, SD
43. YZA - 236kc - Ashcroft, BC, CAN
44. FOR - 236kc - Forsyth, MT

Xlnt cndx. One newly heard "LF 336kc" and 19 new copies for this receiver marked with #.

RBA-2/CFT-46154 SN: 168 - This receiver was being hauled around Carson City, Nevada in the back of a Ford Explorer for many months (maybe longer.) The storage inside a vehicle actually was pretty safe and much better than if the receiver had been stored in a shed or garage where all sorts of contaminates and accidents could be possible. At any rate, I purchased the receiver and power supply right out of the back of the Explorer. Condition is typical "unrestored" which is okay since that means it hasn't been worked on other than maybe at a Navy depot sometime in the past. That's an obvious conclusion since the D.C. VOLT meter is a field replacement "Simpson" in place of the original Weston Type 506 meter. The oval brass tag just above the dial bezel is a Navy asset tag. The glass on the DB meter is cracked. The data plate on top of the cabinet is not mounted but it is the original piece with matching serial number. The USN original acceptance tag is also present on top of the cabinet. The crank-handle is missing from the tuning knob which is typical of nearly all RBA receivers. There wasn't the armored power cable although there is an unused connector on the power supply. Also, no AC power connector or cable.

Inside, the RBA-2 looks very clean appearing to be complete and original.

The photo to the left shows the RBA-2 and power supply "as found" without any cleaning or dusting.

Purchased October 2019.

More details when this project gets started.

NAVY-RADIO.COM - For the most detailed information WWII Navy Longwave gear and on all types and all vintages of Navy LW radio equipment, Navy LW radio stations with vintage photographs - go to   Nick England's incredible Navy-Radio website has the most information available. 



Pre-WWII LW RCVRs (Part 1)                       LW RCVRs - What to Listen To (Part 4)                     Home-Index




Radio Boulevard
Western Historic Radio Museum

 Vintage Radio Communication Equipment Rebuilding & Restoration Articles,

 Vintage Radio History and WHRM Radio Photo Galleries

1909 - 1969

- 60 years of Radio Technology -



This website created and maintained by: Henry Rogers - Radio Boulevard, Western Historic Radio Museum 1997/2023