Howard Radio Company - 1926 to 1949
Howard Manufacturing Company - 1922 to 1926
Electric Specialty Company - 1921
 

Austin A. Howard was born in 1872 and, by 1905, was already building radio apparatus. He was a ham radio enthusiast before WWI and, in 1918, the Navy performed transmission tests using his station in Chicago. He had an experimental call, 9XG, up until 1922. He was at the First National Radio Exposition in Chicago in 1921 where he represented "Electric Specialty Company" which was probably the forerunner to Howard Manufacturing Company. Howard Manufacturing Company began building and selling radio equipment around 1922.

In 1924, Howard obtained a Neutrodyne license from Hazeltine (he was the last company to obtain a Neutrodyne license thus becoming the last company to be accepted into the Independent Radio Manufacturers, a group of fourteen radio companies that were the sole legal builders of Neutrodyne radios and the exclusive holders of the Hazeltine Neutrodyne license.) Howard Manufacturing Company was located in Chicago, Illinois. Austin Howard was already independently wealthy so Howard Radio Company was operated a little differently than the other 1920s radio companies. Howard built modest quantities of very high-quality radios at a fairly leisurely pace and sold them for hefty prices. Just the opposite of another 1920s independently wealthy radio builder, Arthur Atwater-Kent, who built literally millions of radios and sold them for more or less average prices.

Howard changed the company name around 1926 to Howard Radio Company. Howard was able to obtain a RCA TRF license in 1927 and then later the Superheterodyne license. Howard had to build a tremendous quantity of radios per month to be able to afford the minimum royalty to qualify for the RCA licenses. Howard's low production rate had to change and, beginning in 1927, Howard's work force dramatically increased along with the company offering several different models of radios.

About this time (1929,) Howard Radio Company was purchased by the Everett Piano Company and they moved manufacturing to South Haven, Michigan. Everett Piano bought Howard just before the Stock Market Crash and then tried to work their way into the radio business during the depths of the Depression. Everett Piano Company wasn't making any money building radios so they sold Howard Radio Company back to Austin Howard in 1934.

McMurdo Silver, Silver-Marshall and the RCA Licensing - Howard moved the company back to Chicago in 1934 and ended up operating in McMurdo Silver's building that had been Silver-Marshall. McMurdo and his cousin, John Marshall, had started Silver-Marshall in the late-twenties. RCA licensing became a necessity of being in the radio business after the 1930 "Radio Group" Anti-trust suit. The settlement of that suit put RCA directly controlling almost every aspect of radio manufacturing. The Superheterodyne circuit was part of the settlement. It had been Westinghouse's property since 1920 but the government settlement gave the Superhet patent to RCA with the stipulation that licensing had to be provided for qualified manufacturers. Large radio companies that could produce huge quantities of radio, such as Zenith, Atwater-Kent or Crosley, didn't have a problem producing the large amount of radios required but smaller companies couldn't directly qualify and that was Silver-Marshall's situation. Becoming a subcontractor was one way around the RCA licensing and that was how McMurdo Silver dealt with his lack of a RCA license. By inviting Howard Radio to move into the Silver-Marshall building and manufacture radios there, any other radio building for other smaller companies involved would be covered by Howard's RCA licensing. It was a solution that benefited not only McMurdo Silver but later Bill Halligan and others, not to mention Howard Radio.

Bill Halligan and Hallicrafters - Ultimately, Howard found that it was much more profitable to build radios for other companies as a contractor and his RCA and Hazeltine licenses allowed him to do just that. Halligan was friends with McMurdo Silver and had been selling his very early radios by mail-order since this could bypass the license requirements. Halligan made arrangements to have his receivers built at Silver-Marshall by Howard Radio. That allowed Hallicrafters to develop their radio circuits and use RCA TRF and Superheterodyne circuits, protected by Howard's RCA license. The Hallicrafters business ended about August 1936 when Halligan made arrangements with Case and bought the Echophone Company to acquire their RCA license and manufacturing facilities.

Sears-Roebuck and Howard Radio - Howard's biggest and most profitable customer was the Sears-Roebuck Company. Of course, Sears-Roebuck's "Howard-built" radios were labeled "Silvertone" as all Sears' radios were. The Howard Radio Company identification was usually only a manufacturer or contractor number. Howard Radio almost always used the number #1731 as their identification. This number was supposedly based on the Silver building address of 1731 Belmont in Chicago. Typically, a stick-on paper model label was applied to the radio chassis that had the "Howard ID" on it. Sears had always been reluctant to reveal that Silvertone radios were built by contractors, although almost everyone knew that fact. In the later thirties, Sears Silvertone radios have no indication of the contractor-builder of their radios. In the later half of the thirties, Howard began to build and sell ham-communication shortwave radios. Some of the communication receivers were built for Sears-Silvertone but Howard did sell their own brand of ham receivers with a lineup that featured examples from entry-level receivers up to advanced-level communication receivers.

Howard Radio Co. Receiver Design and Build Quality - Howard's communication receivers were designed and built to be affordable, therefore the chassis design used what were essentially broadcast-entertainment radio components. Some of the small Howard ham receivers were obvious AM-BC-SW radio conversions to entry-level, metal cabinet, shortwave receivers that could have been usable in the stations of neophyte hams.

The quality level of Howard communication receivers was similar to the receivers produced by Hallicrafters, Radio Manufacturing Engineers or Breting Radio Company. Howard's receivers sold for about the same price as the Breting receivers but were priced less than comparable Hallicrafters or RME receivers. Breting receivers in particular were fairly crude in their wiring and their mechanical assembly where as Howard receivers were more professional assemblies but of an average quality, somewhat similar to Hallicrafters and RME. Wiring lead dress in Howard chassis is mostly "point-to-point" with an almost "messy" quality to it. Some self-tapping screws are used along with some rivet-mounting. Copper-plated chassis were used for many of the receiver models.

The component quality is comparable to Howard's competition with quality capacitors and some well-known component manufacturers being standard. Most of Howard's components and mechanical parts came for manufacturers located around Chicago. This local purchasing of components was also common with Hallicrafters, Wells-Gardner, Zenith and other Chicago-based companies.

The engineering design of Howard receivers is similar to Hallicrafters in that the design had to use purchased parts. The majority of components had to be standard issue that could also be used in AM-BC-SW consumer entertainment radios and preferably available from Chicago companies. Most of the communication receiver manufacturers built their products in the same manner. There were really only two communication receiver builders at the time that built all of their own assemblies and purchased only passive components. The two companies were well-known for that fact,...National Company, Inc. and Hammarlund Mfg. Co., Inc.

Howard never seemed to put a lot of money or effort into advertising and that might be why Howard communication receivers and his smaller SWL receivers never really gained a foothold in hamdom. Nowadays, you'll see Howard Radio communication receivers once in a while though not nearly as often as Hallicrafters or RME receivers.

Howard Radio really never came back after WWII ended. The company offered a few table model consumer radios and a FM converter for a short time. Howard himself was in his mid-seventies by this time so it's not surprising that the Howard Radio Company was out of business by 1949.

Profiling Six Ham Receivers Built by Howard Radio Company
 

Sears-Roebuck - "Silvertone"  Model 5656A - built by Howard Radio Co. - 1936-37

I actually performed this restoration about twenty years ago. But, for this 2024 write-up, I took this 5656A off the shelf and put it through its paces to see how well it receives here in Dayton Valley, Nevada. I also shot all new photos of the 5656A showing the chassis in detail. As I wrote in 2005, the 5656A is a surprisingly good receiver. Now, in a fabulous radio-reception QTH with a large array antenna, the 5656A does an even better job. Although the top tuned frequency is only 18mc, the reception on 20M is excellent. Perhaps the biggest disadvantage is the lack of band spread. Tuning CW or SSB stations can be difficult. 

General Circuit Description - During the thirties, Sears-Roebuck offered a few different communications receivers built for them by Howard Radio Company. Their Silvertone Model 5656A was offered from late-1936 through 1937. The 5656A had a similar appearance to some of the very early Howard-built Hallicrafters receivers, particularly the Super-7 receiver (there was also a similar looking receiver built for Montgomery Wards.) The 5656A receiver was an eight tube superhet with three tuning bands covering .55 to 18.0 MC. The circuit featured a 6K7 TRF amplifier, 6L7 Mixer (called a Translator by Howard,) a 6C5 LO, a 6K7 IF amplifier, a 6Q7 Det, AVC and 1st AF amplifier, a 6F6 audio output, a 6C5 BFO and a type 5Y3G rectifier. The IF operated at 465kc and featured a variable coupled IF transformer for adjustable selectivity. The tuning dial was large at eight inches in diameter with a convex glass cover with fancy "leafed" bezel. Both the dial and speaker bezels appear to have originally been nickel-plated but have worn to the bronze-like appearance they now have. The beautiful multi-colored dial scale somewhat compensated for the lack of any type of bandspread. An AVC switch was included along with a separate Sensitivity Control (IF Gain only) and a built-in 8" Jensen speaker. It's likely that the screen-grille was originally flocked with some kind of mohair as that was typical of the screen grilles of the day (but, of the several vintage photos showing various 5656A receiver grilles, some appear flocked with a light-color felt, one had grille cloth installed and some grilles appear to just be the screen. There wasn't any felt or glue residue on this screen which appears to be the original.) The BFO was built on a separate small chassis that was "cable connected" to the receiver chassis for power but was electro-statically coupled to the detector. The Phone jack was capacitor coupled from the 6F6 plate. The 5656A provided the basic communication necessities and performance that was adequate for many entry-level hams. The 5656A was fairly popular and can be sometimes be spotted in vintage ham station photos. The 5656A selling price was probably around $49.50 based on that exact same selling price being asked for either the Hallicrafters Super-7 or the 1936 Silvertone Super-Eight (also built by Howard Radio.)

Documentation - The 5656A is documented in Riders Perpetual Troubleshooter's Manual, Vol. X, SEARS page 33. The basics are covered in the single page schematic plus the side notes. The schematic is dated 8-1-36. Although Howard Radio isn't specifically mentioned, there are several clues to Howard Radio's contractor identity. The schematic shows the Mixer as "Translator" which seemed to be something that only Howard Radio called the Mixer tube function on their schematics. The use of the 465kc IF, although not exclusive to Howard, was the usual IF that they used. The Candohm resistor in the receiver is from The Motor Co. of Chicago, the usual supplier of Candohm voltage dividers for Howard Radio. There may have been further confirmation on the rear chassis paper stickers but they are mostly missing on my 5656A.

Silvertone Model 5656A built by Howard Radio Co. for Sears-Roebuck 
SN: 8-A 1853   ca: 1937

Delivered by Airborne Medical Transport to Virginia City, Nevada in 2002

Close-up of Multi-color Dial of the 5656A

Restoration - I performed a complete restoration on this 5656A about 2004. I had to have four knobs cast from one of the good condition originals. At the time, Larry Bordonaro, now "Antique Radio Knobs.com", was just starting out and did custom knob casting in plastic material that looked just like bakelite. His knob castings were first-class and indistinguishable from the originals (Larry is still in business - 2024.) I had to rebuild the tuning system that relied on a pinch-wheel working against a phenolic plastic drive disk. Internally, the original paper capacitors were restuffed with polyfilms, the electrolytic capacitors were replaced using the original cans and the waxed cardboard box that held the two cathode bypass electrolytic caps. The resistors were checked and all were close to spec. I did a full alignment and the 5656A was working to its likely specifications. The indications were that this 5656A was used very little when it was new and was probably stored away somewhere that preserved almost everything about the receiver (like the proverbial "kept in a closet" example?)

The chassis of the 5656A showing the separate BFO chassis. The green material is felt padding for the lid. The chassis is mounted to the bottom plate of the cabinet and rests on rubber cushions. The front panel isn't mounted to the chassis directly. When the receiver is taken out of the cabinet the front panel is only held in place by the control shafts and the knobs. The loudspeaker is an eight-inch Jensen electrodynamic field-coil type. There are three dial lamps that can be either #40 (155ma) or #46 (250ma) types. The steel cabinet adds quite a bit of weight to the 5656A. NOTE: The 6V6GT is still installed when this photo was taken. It has been replaced with the proper 6F6G tube.

More Circuit and Operational Details - The 5656A, at first glance, seems to be a typical AM-BC-SW radio that would have been for the consumer-entertainment market and probably offered as a console model radio. The circuit does have a TRF amplifier stage and a single IF amplifier stage. The chassis was professionally engineered by Howard Radio to be built with parts and components that were the same as those used in their consumer radios but the end product would have all of the basic necessities of a ham receiver. The BFO chassis is an obvious addition but it's well-designed and functions quite well. The Variable Coupled IF works from under the chassis in a manner that's similar to how the Hammarlund Super-Pro Variable-Coupled IF works but using a rotating pin and slider crank rather than the Super-Pro's cam-driven levers with spring-loaded plungers. The Sensitivity control, AVC switch and the Phone jack are also integrated quite well with either shielded cable or twisted wire cables to the front panel controls. Tuning is a little crude being a small phenolic plastic disk that works with a pinch-wheel type of drive. Without any band spread provided, tuning is difficult especially for SSB signals (you have to fine-tune with the BFO.) Of course, in 1937, there weren't any ham SSB signals around and most hams were on CW.

Under the receiver chassis showing its pristine condition

Performance - Years ago in Virginia City, I had thought about using the 5656A as a station receiver with a 1937 Utah UAT-1 transmitter I had but, at the time, the receiver didn't seem to perform with sufficient selectivity for operation on 75M phone. It was a fine SWL receiver and the dial illumination for "lights out" listening was impressive. Well,...that's what I thought back then when I was in Virginia City,...one of the worst radio locations I've ever lived in. Nowadays, in Dayton Valley, Nevada, the QTH is one of the best radio reception locations I've ever lived in. At some time past I had installed a 6V6GT tube in place of a proper 6F6G tube, so before testing began, I installed a NOS 6F6G tube. Although the 6V6GT and the 6F6G can be interchanged and function okay, the 6V6 is a beam-power pentode while the 6F6 is just a pentode. They do sound slightly different with the 6V6's reproduction in a 6F6 circuit sounding a bit shrill. The 6F6 will sound mellow and reveal a bit more bass response (geese,...I sound like an audiophile.) When using the large Collinear Array antenna, the reception using the 5656A is impressive for such a "basic receiver." I was surprised at the signals I was receiving on 20M.  

To demodulate SSB signals, the Selectivity control can be used as an attenuator along with reducing the Sensitivity control in order to have the BFO injection to signal level ratio correct (AVC off, too.) On 40M or 80M, SSB signal levels can be extremely strong when using a large array antenna and even with reducing the Sensitivity and increasing the Selectivity, still some distortion will be experienced. It doesn't prevent demodulation but the distortion is noticeable. When looking at the schematic it can be seen that the Sensitivity only adjusts the IF gain and with the AVC turned off, the RF amplifier is running at maximum gain and that's not adjustable. In the thirties, it was common for the ham to have a separate "receive-only" antenna coupler that not only allowed matching the antenna for best receive response on weak signals but would also allow the coupler to act as an attenuator to prevent "overloading" when tuned to strong signals. One has to remember that most hams in the thirties were on CW and the receiver would have the BFO on and the AVC off so overloading could happen. The ability to attenuate an extremely strong incoming signal was an easy method to prevent overloading. 

The Variable Coupled IF Control showing the rotating pin and slider crank type of mechanism for changing the physical distance between the IF coils to control Selectivity. The grommet is a bit dried-out but a new grommet might increase the "drag" and make adjustment difficult. It works fine now.

As to disadvantages when using the 5656A,...the lack of band spread is foremost and makes tuning in CW or SSB signals relatively difficult. Using the BFO for fine tuning helps. The tuning dial's lack of resolution and very small ham band spans doesn't help the tuning situation. This again relates to the lack of band spread. Using the split-second hand to concentrate your vision on that scale will give the impression that the ham bands have a little more span. It helps quite a bit but still covering all of 20M only takes from 50 to 56 on the split-second dial. Overloading on strong signals can be a problem but it depends on the antenna size and gain. Using the Selectivity as an attenuator works most of the time and more selectivity doesn't hurt on CW and SSB. All in all, the 5656A is a decent performer with some limitations that should be expected for the time period in which it was produced and the design limitations due to its relatively low selling price (it's difficult to imagine but $49.50 in 1937 is equivalent to $1075.00 today in 2024,...WOW!) Though it certainly could be used as a vintage ham station receiver on 75M, its success would depend on the user's motivation and experience with using pre-WWII receivers "on the air." 

Sears-Roebuck - Silvertone  STANDBY - Model 5752 - built by Howard Radio Co. - 1940
 

The Model 5752 is a nine tube, single preselection superhet with two IF amplifiers, crystal filter, electrical bandspread, R-meter, switchable AVC, BFO with front panel frequency adjustment and a single-ended 6V6 audio output all built onto a copper plated chassis. Tuning range is in five bands covering .55mc up to 40mc. An external speaker was required. Dials were illuminated but the R-meter wasn't. Dates from 1939 to 1940. The Model 5752 was built for Sears by Howard Radio Company.

One certainly notices the similarity of the Silvertone STANDBY's appearance to that of the Radio Manufacturing Engineers RME-69. It had to be a deliberate design "copy" of the RME-69 but what would have been the reason? A clue might be in the name "STANDBY." At the time, many hams believed they should never be without receiving capability. If the main receiver was down for repairs, then the "standby" receiver could be used. The active and prepared ham would always have a "standby" receiver. If the station receiver was a RME-69 or RME-70, then this Sears "look-alike" would blend in and appear to be another RME product. It seems unlikely that Sears' "RME-69 look-alike" would have been styled to fool someone into thinking this Howard-built receiver came from RME, especially at the price Sears was selling it for. It's more likely that the designers imagined that a ham looking for an inexpensive "standby" receiver would find the Model 5257 appealing because of its familiar appearance.

No restoration has happened on this "STANDBY" receiver,...at least, not yet. One can note that to the left of the tuning dial escutcheon there is a small red pilot lamp jewel. This superfluous addition will be removed as soon as work starts on this receiver. It's obvious that some mechanical rework is going to be required since the dials are obviously tilted somewhat. More to come,...

Silvertone STANDBY Model 5752 built by Howard Radio Co.   ca: 1940

RME-69 look-alike, minus the RME price,...and performance

Howard Radio Company - Model 437A - 1940-42

The Model 437A was the last evolution of the 435 and 436 models of nine tube receivers that were mid-level communication receivers. The 437A covered .54mc up to 43mc in four tuning ranges with band spread capability. The IF operated at 465kc and there was an optional Crystal Filter and an optional S-meter. This 437A has the Crystal Filter option installed but not the S-meter. As the 437A with Crystal Filter, the selling price of the receiver shown would have been $69.75 in 1941.

Many of these smaller models of Howard receivers had a matte silver-based chemical treatment for the dial finish that nowadays is almost always is found in terribly corroded condition. This particular 437A's dial is in excellent condition (it may have had a protective coating applied when manufactured but more likely the pristine condition of the dial is due to the dry climate here on the East Slope of the Sierra.) In fact, the entire receiver was cosmetically in very good condition and the chassis was essentially original. Unfortunately, although I took the time to clean this 437A and photograph it, I didn't restore it to operational condition. In fact, I sold this receiver quite some time ago (and I don't even remember who bought it since I sold it "out of the museum") but at least I did keep the photograph.

Howard Model 437A with Crystal Filter option   ca: 1941

A transient relic now in other hands

Howard Radio Company - Communication Receiver Model 450A - 1938-39

I was tipped off that this Howard 450A was going to be at the May 11, 2024 Spring SNARS (Sierra Nevada Amateur Radio Society) "Cabela's Swap Meet" - so called because the ham swap meet is held in Cabela's parking lot in Verdi, Nevada (10 miles west of Reno.) The 450A was part of a very large selection of some really nice "collector" ham gear that was brought to the swap in a 14' moving van by Ham and Hi Fi of Sparks, Nevada. Of course, at a swap meet you have to "take your chances" since a detailed inspection isn't really very practical - "as-is" or "as found" is the expected and typical condition. However, the 450A looked very complete and the dial was magnificent, so the purchase was made. This write-up has a lot more information than the proceeding restoration write-up on the Silvertone 5656A since this write-up was composed as I was restoring the 450A receiver. So, expect the typical OCD-type of restoration write-up presenting every bit of minutia I find (with many photos.) How the multitude of minor problems were corrected will be covered and eventually how well (or poorly) this Howard communication receiver performs. 

The Model 450A was one of Howard's largest communication receivers at 12 tubes and a tuning range that spanned .54mc up to 65mc in six tuning ranges (only the Model 490 had more tubes at 14 total.) The Model 450A was one of a few receivers available at the time that tuned the 160M through 10M ham bands and included 5M coverage. The smaller Model 440 was nine tube receiver with a similar appearance but without 5M coverage, the dual IF, push-pull audio, the "stop watch-like" band spread logging dial, the band-in-use indicator and, like many of Howard's larger receivers, the crystal filter was an optional installation. The Model 440 sold for $84 in 1938. The Model 450, also from early-1938, featured additional tuning ranges increasing the coverage up to 65mc however when going to bands E and F a different IF of 1560kc was used and the RF amplifier was bypassed on the highest frequency band since the 6K7 RF amplifier tube didn't have very much gain at VHF. The Model 450 sold for $87 and $10 extra if the crystal filter was installed. The Model 450A was introduced within a very short time as the Model 450's replacement (still in 1938.)

The Model 450A had basically the same features as its predecessor but now the band F antenna input actually was a TRF transformer to the Mixer grid where before it had just been a LC input. The VHF antenna terminal had been A-G two-terminal and was now A-D-G three terminals. The dual IF of the 450 was basically unchanged for the 450A. The ABCD IF was 465kc (frequency coverage from .54 to 16mc) and the EF IF used the same 465kc IF amplifier tubes but had entirely separate IF transformers that were tuned to 1560kc and switched into the IF circuitry the Crystal Filter/IF Shift switch. Frequency coverage from 16 to 65mc was available on bands E and F. The 1560kc IF helped reduce the rampant images that would be encountered on bands E and F. With single preselection (one TRF amplifier,) images on A,B,C,D bands probably wouldn't be encountered until around 15mc, depending on the antenna used and the strength of the signal producing the image. With the 1560kc IF use (beginning at 16mc,) images would appear 3.12mc below the transmitted frequency which tended to put the image response out of the passband of the receiver's front end selectivity, especially with band E also having a preselector stage. Two separate sets of antenna terminals are provided on the back of the chassis. Both are three screw terminal types marked A-D-G. One antenna input is just for Band F (5 meter band-VHF.) A vertical whip antenna 59" tall is recommended (with an 8" wire connection to the antenna terminal) for 5 meter reception. The antenna input for Bands A-E is for any type of standard balanced or unbalanced antenna (with resonant antennas providing the best performance.)

The crystal filter was switched out of the circuit when IF SHIFT BAND E-F was selected (that eliminated the need to have two crystals, 465kc and 1560kc, in the Crystal Filter circuit.) The BFO had a variable injection level adjustment in addition to a frequency control (pitch.) The 1st AF amplifier used a bias cell for the grid (these were popular in the late-thirties and were very small batteries that fit into a special holder, some were solder-in types, usually providing about -1.5vdc bias voltage.) A phase inverter type of audio circuitry was used. The audio output used push-pull 6V6 tubes and 9.5 watts output was claimed with impedance outputs of 5Z and 500Z from the internal audio output transformer. The entire chassis and shielded box for the IF amplifier/Xtal Filter are both copper plated. A 10" PM loudspeaker was available in a matching housing.  >>>

Documentation - Fortunately, the Howard 450A is documented fairly well in Rider's Perpetual Troubleshooters Manual Vol. X, pages "HOWARD PAGE 10-28 to 10-31." The pages include the schematic, the component layout for both the top and bottom of the chassis, color codes for the transformers, receiver specifications and the alignment procedure. There are many online sources to view any of the Riders PTSMs. The easiest to use is at www.worldradiohistory.com with all of the Riders manuals (selectable by Volume number) in PDF format. Unfortunately, BAMA edebris doesn't have an online manual or any documentation on the Howard 450A.

450A Tube Line-up - RF Amplifer - 6K7, Mixer (called a "TRANS" on the schematic - Translator, a Howard moniker for Mixer) - 6L7, HF Oscillator - 6J5G, First IF Amplifier - 6K7, Second IF Amplifier - 6K7, Detector/AVC/First Audio Amplifier - 6Q7G, BFO - 6J7G, Voltage Amplifier (R-meter) - 6J7G, Audio Phase Inverter - 6J5G, Push-Pull Audio Output - 6V6G (2), Rectifier - 80. It's interesting to note that many of the tubes that could use metal octal types are specified as "G" or glass envelope tubes operated without shields (pin 1 is grounded so metal tubes could be used and shielding would then be provided.) Some receiver designs would use glass tubes for oscillators because the glass envelope without a shield would run cooler than a equivalent metal octal version and might possibly have less thermal drift. The Detector/AVC/First Audio Amplifier would operate at a fairly high level signal and shouldn't pick up any noise so the use of a glass tube might have been for thermal reasons. In almost any vacuum receiver, the audio tubes and the rectifier will almost always be glass tubes because of the heat involved with their operation.

450A Glass Dial - A Multitude of Indicators

I checked the power transformer and it's a physically smaller replacement that doesn't quite fit correctly. Two pieces of 1/2" steel channel are used for mounting adaptors making the replacement obvious. This transformer does have close to the standard voltage outputs but it appears too small for a 12 tube radio set (should be HV 375-0-375vac ~.120ma, 6.3vac~4.25A, 5.0vac~2.0A.) This power transformer looks like it was for an eight-tube set. The original power transformer was probably a Stancor since they, like Howard, were located in Chicago. It shouldn't be too difficult to find a correct spec Stancor power transformer. The filter choke tested okay.

The audio output transformer has an open plate to CT winding. It's the dual secondary of 5Z and 500Z with a rating of about 10 watts that makes an exact replacement difficult to find. The closest that I can find (in my stock) would be out of a National HRO-60. This audio output transformer is for P-P 6V6 tubes and has a secondary for 8Z and 500Z. It's a frame-type transformer and should be a fairly close fit. Another "Howard issue" is that the original audio output transformer and the filter choke are both mounted using rivets compounding the replacement task. 

Typical of some receivers designed at this time, the tuning condensers are mounted on rubber cushions which have compressed and dried out. The rubber bushing mounting of the tuning condensers involves the entire front dial backing plate with the front of the MT condenser and the BS condenser mounted to the plate, then the plate is cushion-mounted to the front chassis apron and the rear of the MT and BS condensers are mounted to the top of the chassis with cushions. This arrangement allows replacing the cushions without unwiring the MT and BS condensers.

The RF Gain control shaft is extremely worn and wobbles quite a bit. The AF Gain control is not worn. The BFO air variable has a strange "feel" to it,...it's like it's rubber mounted but it isn't. Further investigation will be necessary to see what's going on with the BFO.

The Main Tuning uses dial cord for the drive but the Band Spread uses a tuning belt. Belts are always difficult to replace and even if the correct belt is known (JFD #21 in this case) it may not fit correctly because of wear. The 450A has the MT/BS condensers mounted to the backing plate and both tuning shaft bearings are in the same backing plate so the dimensions can't change, therefore, a belt with an inside diameter of 3.5" should work. Using square rubber extrusion works fairly well for custom rubber belts that are cut to size and then glued with super glue. However, they don't last indefinitely. Likewise, O-rings will only work for a short while before drying out and breaking (O-rings aren't designed as "belts" so it's not surprising that they don't last very long in that function.) The original belts lasted a long time because they were a fabric sandwich with impregnated rubber. The trouble is, these belts are at least 70 years old, so they might be non-pliable nowadays.

The glass dial is in perfect condition. The champagne-silver backing plate is excellent. The glass dial cover needed some padding where the clamps contact the glass. The pads allow tightening the clamps to slightly compress the padding and secure the glass without breaking it. The receiver front panel is in very good condition and the cabinet is in good shape. The bottom cover is missing but would be easy to replicate.

The Plan - The Howard 450A is going to require a complete "strip out" of the old sloppy repairs, the incorrect parts, the modifications and the defective parts. This is usually what has to be done with ham radio "kits" but so much of the 450A has been tampered with, it almost has to be treated like it was a "kit" that was assembled by someone lacking any experience. The original sections of wiring will remain installed. Any original components that can be used (tested in spec) will be used but all replacement components will be of the original style and vintage that test as usable. Then the chassis will be rebuilt exactly to the schematic with component placement as shown in the layout drawing.  >>>

Under the chassis of the Howard 450A with the under-size power transformer removed and the bad audio output transformer removed. Note the many non-original capacitors and resistors besides the terrible wiring lead dress. A lot of that "messy looking wiring" is original. Howard routed some wiring in a harness but most wiring is "point to point."

Cleaning Begins - For now, I've cleaned the chassis with WD-40 followed by Glass Plus. It was actually quite dirty and responded well to the cleaning. I had to replace the two front rubber bushings that mount the tuning dial backing plate. The original rubber bushings had disintegrated so the backing plate would move around. The top chassis rubber bushings for the rear mounting of the MT/BS condensers were in good condition. I ordered a JFD #29 tuning belt. If that works, I can reassemble the tuning dial.

Belt Update - May 25, 2024 - The JFD #29 tuning belt arrived today and, while it's not what's called for in the JFD index (#21 is,) I went by what I measured. The JFD #29 was 11" inside circumference or 3.5" ID. The fit was perfect but the belt kept derailing while tuning. However, when I mounted the front dial plate that provides the front bearing for the tuning shafts, then the shaft was perfectly horizontal to the vertical pulley on the BS condenser drive and the belt then functioned correctly and didn't derail.

Next is to remove all of the modern parts installed. To make reassembly easier, I'll take a close-up photo for reference. Usually, when doing a "strip-out" on a ham-kit receiver, the documentation will have a detailed drawing of all of the components and wiring as part of the assembly instructions. The 450A wasn't a kit, so the only drawing shows just the placement of the capacitors and resistors without any detail on wire routing. The photograph with the schematic and Howard's component location drawing will all be used when doing the reassembly.

Any modern resistors that are replaced with correct vintage resistors will have to have the vintage resistor values checked. I have a large quantity of late-thirties "dog bone" resistors that are NOS. Over half of the resistors checked have drifted in value by 50% and some are 100% or more higher in value than marked. Many of the original resistors are still present under the chassis but those will have to be checked for drift. The original resistors seem to be all "dog bone" types except for three straight-end BED resistors but with a color band replacing the dot (not the normal color-band configuration.) Those three straight-end types appear to be originals and are (2) 50K and (1) 500K. 

Showing the type of sloppy repairs performed on the voltage divider sometime in the past. The obvious "over-kill" 10W Ohmite resistor with a ring of glopped-on solder is the B+ input to the Candohm. Black electrician's tape, hook splices and tack soldering show that this is worse-than-amateur-level work. Due to a multitude of errors, the receiver never could have worked afterwards (but almost certainly the perpetrator blamed Howard for building a such a "crummy" radio.)

It's really quite easy to calculate the entire Candohm divider network and the dissipation requirements. First, total all of the Candohm resistances (17.6K.) Observe what the B+ level is at the input to the Candohm (+245vdc under load,...+305vdc at the rectifier, IR=60vdc across 365Ω choke=~160mA total B+ current draw.) Then I=E/R will give the total current through the Candohm (245/17,600=14mA.) The I²R=P will give the dissipation of each resistance in the Candohm. The TOTAL dissipation across the entire Candohm is about 3.5 watts (so 10W for one section was definitely "over-kill.") The greatest dissipation is the 10K section at 2 watts (and that section is okay.) Generally, dissipation (watt rating) is doubled as a safety factor since the calculations are based on the receiver operating with normal current draw but at "turn on" much higher voltages will be experienced (although less current) until the tubes warm-up and start to draw current off the divider taps. Since the resistance values are already known from the schematic, only the dissipation of each section is in question. The current draw of the operating circuit off of each Candohm tap also affects the total current through the divider to a certain extent but for dissipation requirements, the static voltage drops will be close enough. The affected resistor in the 450A Candohm is the open input resistor that can be replaced with a 2.5K 6W (.5W actual dissipation.) The 1.5 section is okay (.3W,) the 3.6K section is okay (.72W) and the 10K section is okay (2W.) The dissipation levels shown in parentheses are calculated with +245vdc B+ (from schematic) and are probably slightly lower than if the receiver was operated on 120vac line voltage (the +245vdc was probably specified with 110vac to 115vac line voltage.)  >>>

Howard 450A Top of the Chassis   (W-G replacement Power Transformer Installed)
Note the Band Spread condenser only has one rotor plate and one stator plate per section. This "Howard" modification is shown in the Rider's PTM information and indicates that one rotor plate and stop per section can be removed to reduce vari-C and thereby increase the "spread" for better fine tuning response. Unfortunately, this mods RUINS the band spread action entirely.

W-G Power Transformer Installed - May 28, 2024 - Just a very slight adjustment of the square hole was necessary for this replacement to fit. About .090" was removed from the front-most short side and that was enough for the transformer to "drop in" the tub hole. I marked the four holes necessary for mounting, drilled them and then mounted the W-G transformer. I tested the transformer several times just to be sure that the removal of the 25vac leads (cutting, insulating and placing the leads inside the bottom cover) didn't cause any problem and to verify the operation of the transformer was as expected. I didn't wire in the W-G transformer yet since I still need to "strip out" the modern components and sloppy repairs first.

The Audio Output Transformer - This transformer is from a parted-out HRO-60 receiver with a P-P 6V6 audio output stage. The specifications for this transformer are almost perfect as a replacement and, physically, it's exactly the same size as the defective original. Installation was straight-forward with no problems. Since some of the HRO-60 transformer's leads are shorter than what was required, I had to splice the original transformer's wire leads to get the length needed. Although one would first think that shrink tubing should be used to insulate the splices, shrink tubing looks "too new" (because it is.) In the late-thirties splices were insulated either with friction tape or with lacquered tubing. I used the lacquered tubing if possible and friction tape as a second choice.

More Stripping - May 30, 2024 - I removed all wires going to the Candohm. Several were not connected to the correct voltage taps. I marked all wires (wrapped tape ID) after verifying were the wire came from so, when reconnected, the circuit will have the correct voltages. I found a wire going to the LO tube socket that was a broken solder joint. This was under the coils and would have been almost impossible to spot except I was tracing wires to the LO and happened to see the broken joint when I moved the wire slightly. Luckily, there was sufficient spacing between the LO vertical shield and the LO coils to access the tube socket and the wire was soldered back onto the socket terminal. Removed all components and wiring to the rectifier socket and both 6V6 sockets. Only one resistor was original. 6V6 cathode resistor was a replacement that wasn't the correct value.

Lead Dress - The Howard 450A component drawing shows the proper placement of the capacitors and resistors. Unfortunately, it's obvious that Howard's assemblers didn't install ALL of the components exactly as shown on the assembly drawing, so there will be a few capacitors placed differently but they are "as originally installed." Some of the wiring in the chassis uses a harness. The other wires look randomly installed. Comparing the original wiring in the Silvertone 5656A to the 450A, it appears that Howard wired the tube heaters and some of the B+ in the harness while the signal wiring was "point to point" wiring. At the time, "point to point" was a lead dress style that believed that the shortest, most-direct wiring resulted in less capacitive coupling in the signal circuit so less losses and better performance resulted. I'll be rewiring the 450A in this manner to adhere to "Howard's style" of wiring.

Capacitors - Since so much of the under-chassis isn't original I don't think there's really any point in restuffing the few remaining original caps and adding non-original shells into the mix. So, I'm going to use yellow-jacket polyfilms but, as I've done in the past, I'm going to paint the capacitors to hide the hideous yellow color of the modern capacitors. Resistors needed will be as close as I can find to the original style. The effect this creates is at first glance the under the chassis appearance is that of an original receiver. Close examination would easily show that the painted capacitors were obvious replacements.

Side-Tracked - June 1, 2024 - The N7RCA Minden Swap Meet had a few of impacts on this project. First, I had to get a SCR-511 "Pogo Stick" WWII transceiver packed and shipped to North Carolina. Then I had to do a "quickie repair" for KB6SCO of a National NC-183D receiver that was going to be given to a newbie SWL. Then I made a purchase at the swap of a Linear Amplifier that had been built by W6MIT (this was a case of "mental side-tracking.") These events delayed the Howard 450A project for about a week (and the MIT-linear is intermittently continuing with its interruptions.)

Jun 8, 2024 - Completed wiring of the audio output transformer. Extensions had to be added to the plate connections to the 6V6 tube sockets. The splices used helically-wound 24ga TC coils that were used for "soldered butt joints." That way the sleeving could be installed over the joint and it would appear original.

Jun 9-11, 2024 - Continued removing non-original parts and installing vintage dog bone resistors that are checked for correct value. Correcting errors (this 450A would NEVER have worked after its last "hamstering" job) and removing solder bridges (due wrapping component leads around terminals and glopping solder everywhere,...example,...the 2nd IF amplifier screen was shorted to ground with a solder bridge on the bypass capacitor. Had power ever been applied, the screen load resistor would have "smoked." ) There's something about the solder too. Either it's some sort of "no lead" solder or else all of the non-original solder joints are "cold solder joints." The solder just "clumps off" when heated. It wasn't adhering to the terminals at all. Also, found several joints where the solder wasn't flowed properly and there wasn't any contact with the lead that was just wrapped around old cut leads from the old removed parts. The rework quality on this 450A is way beyond dismal. Very close to the worst workmanship I've seen in a very long time.

Bias Cell Installed in Holder

I identified all of the polyfilms I'll need for the capacitor replacements. Seventeen capacitor are needed. I'm short five .05uf caps but I'll look in the shop tomorrow and should be able to find the amount needed. These will be painted and then the value written with a paint pen. While they won't look "original," they won't look like modern polyfilms either.

The Bias Cell - The 1st AF Amplifier grid is biased using a "plug-in" cell. The cell holder isn't wired (disconnected sometime in the past) but the old cell is still installed. A modern Energizer #357 hearing aid battery or some similar type of small cell will fit into the cell holder. When wired into the circuit, then about -1.5vdc would be provided as the 1st AF Amplifier grid bias. Bias requirements are extremely low-current so cells will last for its shelf life. It's also possible to change the 1st AF Amp to a cathode bias set-up using a resistor and bypass capacitor to provide the -1.5 grid bias by elevating the cathode to +1.5vdc and then grounding the 1meg grid resistor in the circuit,...but for testing performance, I'll keep the battery-cell grid bias at this time.

Using an Energizer #357 happened to be pretty easy since there was a cell holder that was still present. Of course, the modern cells are much smaller but by using .25" diameter brass stand off that was about .312" tall as a conductive spacer, the #357 was secure in the holder and did apply -1.5vdc to the grid of the 6Q7 triode 1st AF Amplifier. I'm going to look through my stand offs and come up with a brass one that's .375" diameter that will fit into the cup better and I'll use a height that allows the #357 cell to set "square." See next photo down under the chassis and note the better fit of the bias cell with a proper standoff.

Another Delay - June 13, 2024 - I can't believe that I couldn't come up with five .05uf polyfilm capacitors around here. Well, I did find lots of .05uf caps but they were either Orange Drops, NOS molded paper caps or weird looking types - no yellow polyfilms with axial leads. So, I went ahead and ordered all of the values needed for this Howard (and also for the Hallicrafters SX-9 and 5-T that are up next for a rebuild.)

June 17, 2024 - The polycaps arrived today. These are MET brand that I've never used before. They are much smaller than the IC brand (Illinois Capacitors) or other yellow jacket polycaps. Although I have to use the .05uf and .002uf METs, I'm using the IC brand for the .1uf, the .25uf and the .02uf caps. I WON'T be ordering METs next time but will pay more for the IC brand polycaps. It's a "size" thing,... the METs function fine. On painting the polyfilms,...I tried a few colors, like dark red, dark gray, dark brown but I settled on Raw Sienna which is a light brown color. Using Artist's Acrylic paint will result in a flat finish to the capacitor that looks a lot like paper. Also, this color looks very close to the original capacitor shell color and, except for physical size, results in a "yellow jacket" almost looking convincing like an original paper capacitor. As to capacitor value markings,...through experimentation, I've found that the least amount of writing looks the most original. More markings, like "outside foil" bands or voltage ratings usually lead to the writing looking quite bad. Simple is best because it results in the best consistency of number appearance. I just marked the value and that was all. At least the painted caps shouldn't standout too much as replacements. I installed seven of the painted polycaps starting on the IF side of the chassis. The acrylic paint doesn't change the capacitor value or affect its performance,...it's all about cosmetics.

June 18, 2024 - All capacitors installed in the IF, Det/AVC and Audio stages. Five caps left,...two in the RF front end, one for the BFO, one on the Candohm and one on the AC line. Double checking the resistor installation and the wiring as I progress through the capacitor replacement job.

Howard 450A chassis after rebuild. The color I painted the polyfilm caps is Raw Sienna - a very good color that results in the caps looking sort of vintage. The electrolytic capacitors were painted maroon so they wouldn't be too obvious either. As expected, the 2.5K 6W resistor installed on the Candohm is barely visible.

A close-up of the power supply section of the 450A showing the bottom of the W-G replacement power transformer and the HRO-60 audio output transformer. I don't know why the HRO-60 audio transformer is MFP coated, normally they aren't. This transformer was in a box of HRO-60 spare parts I was given and had been marked "HRO-60" and it does check-out as that. Works great!

June 20, 2024 - Installed the two remaining electrolytic capacitors. Installed the vintage AC power cord. Tested the tubes and found that one of the 6V6 tubes was much lower gm that the other 6V6. The low 6V6 was just barely above minimum acceptable, so it was replaced. One of the 6J7 tubes tested at minimum acceptable so it was replaced. All other tubes tested good. Found some replacement tubes for the 6V6 and the 6J7(replaced with a 6J7GT.) Cleaned tube sockets with DeOxit and installed the tubes.

AC Applied - To power up the 450A, I monitored the input voltage to the Candohm resistor which is also the 6V6 plate voltage. I slowly ramped up the Variac and stopped at 90vac input. The receiver took a while warming up (at that low of an AC input voltage) but I finally heard some noise in the loudspeaker. The 6V6 plate voltage was +210vdc. I increased the AC input voltage to 115vac and the 6V6 plate voltage was at +255vdc which was perfect. I had the receiver on the AM-BC band with a ten foot long test antenna. I adjusted the RF Gain and AF Gain controls and tuned (no dial mounted yet) by looking at the tuning condenser mesh. Several of the Reno AM-BC stations "pounded in" quite strong. KKOH 780kc had the R-meter about mid-scale. Lots audio power available. Tried a few other stations and all sounded fine. I switched the AVC on and off and some distortion could be heard with the AVC off which isn't unexpected. So far, the 450A seems to be functioning fine but I'll have to install the glass dial and points before I do any alignment. Also, I have to do further testing on shortwave with a large antenna array to see how the higher frequencies are operating. Also, the BFO needs to be aligned. But, for a first test, this was a good one. I'm sure this is the first time that this Howard 450A has been functional in several decades. 

Complete IF and RF alignment was required and allowed this 450A to now receive signals on all bands.

Reassembly - June 23, 2024 - The front panel is secured to the chassis vertical rails by three screws on each side. Control that are mounted to the chassis aren't mounted to the front panel but the Phone Jack, the AVC toggle switch, the BFO frequency selector toggle switch, the Send-Receive switch, the Tone control and the BFO-on/injection control are all mounted to the front panel. The knobs were then cleaned and installed. The receiver was now ready to install into the cabinet. The front panel is screw-mounted to the cabinet. The chassis mounting screws were missing so replacements were installed. There's no lid on the cabinet but there is a "flip-up" panel on the back of the cabinet that allow access to the tubes,...barely. The bottom cover over the RF coils was missing. I'm not sure of the material. Some receivers had heavy cardboard covers but if it was metal it would probably have been mild steel and probably thin, like 22 gauge. I'll have to make a cover. 

Overall, for a little over $100 in 1938, the Howard 450A was probably a bargain and undoubtedly performed much, much better back then when there were 85 less years of wear and tear along with endless "diddling and heavy-handed repairs." I tried to put the 450A back to as close to original as I could but there's just too much that's been ruined by careless rework in the past. I certainly believe that when it was new, the Model 450A was a spectacular receiver that performed admirably. If I had a 450A "parts set" that could provide replacements for the many "ruined assemblies" I think this 450A would become quite a nice receiver. As it is now, it's a very good receiver with a few minor issues.

More on the DOWNSIDE - The Howard Suggested Band Spread Mod - Listed in the notes for alignment in the Rider's PTM information (from Howard) is the suggestion to remove one rotor plate and the stops from each section in the band spread condenser to increase the spread, that is, to have very little change in frequency per BS knob revolution. It's really,...and I mean REALLY TOO BAD that someone in the past believed the Howard hint and did remove one plate and stop on each the band spread rotors. This has made the band spread virtually useless with endless knob turning for very little frequency change. I've found what works okay is to leave the BS on 500 and just tune in stations using the Main Tuning and then only use the BS for a vernier F control. It really doesn't affect the calibration much since from BS tuning from 1000 down to 0 must only change the frequency about 30kc if on 40M and maybe 70kc on 20M band (BS has very little effect on 80M.) The Howard mod RUINS the BS function, in my opinion. No quality repair is possible unless an original, unmodified BS condenser could be salvaged from a "parts set" (since the rotor plates and the stops are long gone on mine.) Finding such a "parts set" would take miracle.

1938 Howard Model 450A Receiver

Hopefully the stunning bench presence with the dazzling array of whirling needles and tell-tale indicators made up for the more-or-less average performance. Certainly, the new owner/operator was, with a little practice, able to soon discover that his new (but slightly idiocentric) receiver really was capable of providing decent performance.

UPDATE: July 1, 2024 - Listening on the 20M band about 1400hrs PDT. Using the Band Spread as a freq. vernier works pretty well for tuning in SSB signals. Went up to just over 15mc WWV and thought I heard a SSB signal,...and I did. It was Trenton Military on 15.035mc USB with the Aviation Weather out of Trenton, Ontario, Canada. This is a pretty easy station during the winter but fairly difficult during the summer. Trenton was Q5 and S6. The Howard 450A takes a little getting used to and I've noticed that the more I use it,...the better it performs (or the more I get used to its idiosyncrasies.)

NOTE: July 5, 2024 - The Howard 450A is going to return to the workbench next month. I've ordered a special "long shaft" 5K 2W potentiometer with switch to replace the worn-out RF Gain control. Also, I'm planning on repairing the Band Spread condenser by adding the missing rotor plates back to the air variable. I'm going to search through the junk boxes and find an air variable with about the right size rotor plates and see if I can remove them. Early style (from the 1920s) air variables could be disassembled to remove the rotor plates and that might be how I accomplish getting three rotor plates. I'm sure I'll have to trim and size the rotor plates to match as closely as possible. How to successfully mount them so they are secure and still making contact with the grounded rotor shaft will have to be something I'll experiment with. Other work planned is to revisit the 1560kc IF alignment and the 1560kc BFO adjustment. Neither seems to work very well although they seemed to align fine. Also, I want to recheck the Tone control functionality. The 450A is a pretty good receiver and I think with a little more rework, I can get it to approach a "very good performer."

Aug 1, 2024: 450A Back on the Bench - The first bit of work was to install the new 5K 2W pot with AC power switch. I got this part from Greece because I needed a "long shaft" and the AC power switch and these options combined with a rather low resistance of 5K seemed to be difficult to find from a USA supplier. I never thought about the part actually being "made in Europe" until I started the installation and discovered that the standard USA mounting nuts didn't fit. Luckily, the pot was supplied with one mating nut, so I just found a larger diameter nut to act as a spacer. I had to just slightly enlarge the mounting hole in the chassis. A few passes with a fine file allowed the pot hub to fit through the mounting hole. The standard USA lock washer did fit but just barely. Also, the pot shaft had to be cut down from 4" long to about 2" long and that was easily accomplished. Then the pot was mounted and wired. Testing went fine with no problems. Now the RF Gain shaft doesn't wobble around and feels nice and tight. Operation is as it should be.

Next is the Tone control that doesn't seem to do anything. There's nothing wrong with the pot. The problem is the very low value of the C in the RC. It's only .01uf so the change in tonal response is barely noticeable. I clip-lead connected an additional .033uf and then the Tone control made a nice change to the audio response. I replaced the .01uf with a .05uf (the test had .043uf so .05uf is close enough.) Lots of range in the TONE adjustment now.

The 1560kc IF,...IF was way off. I thought I had aligned it to 1560kc but apparently not. The IF was so far off, I realigned the front end on Bands E and F just to make sure everything was on frequency. Although the noise floor came up significantly, a substantial antenna is required for much in way of signal reception on Bands E and F.

In looking closely at the band spread condenser I was surprised to find that the rotor plates are "glued" in place on the rotor shaft. At first, the glue looked like old Duco cement but it's as hard as epoxy and it doesn't want to chip off. The epoxy would have to be removed because it was smeared over and into the rotor shaft grooves that the rotor plates would mount to. Whatever this glue is, it's permanently part of the rotor shaft now. This glue has all but eliminated the possibility of replacing the missing rotor plates. At first I thought about adding a second stator plate but further thought and I knew that would only increase the static C value and there would still only be one rotor plate varying the capacitance, so the end result would be about the same band spread effect. Unless an original 450A (or 450 or 440 models) band spread condenser turns up, I guess I'll be stuck with this destructive mod.

That completes the rebuild on the 450A. It's a nice performing receiver with commanding bench presence but I wish the band spread hadn't been ruined.

 The Hallicrafters, Inc.  -  SX-9  -  Contractor: Howard Radio Company

Bill Halligan made several deals with Silver-Marshall Mfg. to start The Hallicrafters, Inc. and began offering receivers in 1933 with the first receiver designs using TRF circuits. Halligan didn't have the necessary licenses to build the radios and sold his early radios by mail order to avoid potential RCA lawsuits. S-M also didn't have the RCA license for Superheterodynes (or TRFs.) Since neither S-M nor Halligan had the necessary RCA licenses all legal radio building had to be performed by Howard Radio Company (who did have all of the radio licenses.) Howard had moved into the Silver plant after he purchased his company back from Everett Piano Company in 1934 and moved his company back to Chicago. RCA had sued McMurdo Silver for building radios without the RCA license so the move of Howard into the Silver plant solved several problems for all involved. Howard Radio is the most likely contractor for the SX-9 since the early Hallicrafters receivers involved Silver-Marshall and McMurdo Silver. Howard Radio built the Hallicrafters receivers up through the 5-T and the SX-9. By August 1936, Halligan had made his arrangements with Case and had purchased Echophone. Since all of the early Hallicrafters receivers seem to have been built at the Silver-Marshall plant and Howard Radio was the actual RCA-licensed company building radios there, the SX-9 was almost certainly built by Howard Radio. In fact, on this SX-9, the paper tag is still on the chassis that has "Manufacturer No. 1731" indicating Howard Radio Company as the builder. The SX-9 was offered in late-1935 through early-1936 and featured a built-in speaker, nine metal-type tubes and Aladdin iron-core IF transformers in a superheterodyne circuit with bandspread. The first Hallicrafters receivers built without a contractor were the SX-10 and SX-11. The SX-9 was a significant improvement over earlier TRF models that Hallicrafters had been selling. Its performance is usually rated quite good, even though it still retained a somewhat crude "amateur" appearance.

The Hallicrafters SX-9 was their last contractor-built receiver although early 5-T Sky-Buddy radios were also contractor-built.

I didn't do anything to the SX-9. I didn't even test the tubes. I put it on display in the museum and it stayed there until we closed in 2012. Since then, the SX-9 has been stored in the shop on a top shelf in my climate-controlled room (well, the temp only varies from 75F down to maybe 20F and there's absolutely no light unless the overhead lights are turned on.) I did "touch-up" the bottom raised platform where the wrinkle finish paint had flaked off before I put it on display but, other than that, nothing has been done to this receiver in the 25 years I've owned it. So,...I guess it's about time I find out just how good of a receiver the Hallicrafters SX-9 receiver is and, additionally, a close examination maybe provide some clues as to the contractor's identity (Howard Radio Company is the suspected contractor.)

The "Mostly Original Parts" Challenge - In checking all of the major components and finding that everything is okay. First problem was with the power cord. It had one side open so no AC to the power transformer. I clip-lead connected a test AC cord in place and then slowly ramped up the AC input voltage. Just the band spread dial lamp illuminated. Nothing heard, no even a slight hum. Well, not unexpected. I'll have to do the usual inspection and test to see what I come up with. On closer examination, like looking at the schematic, I saw the the audio is completely cut-off if the bias cell is gone (or open like this one.) To avoid damaging the cell (since it wasn't necessary) I made up a -1.5vdc cell from a Energizer #357 and soldered it to the terminals. Now I had some hum in the speaker but no signals. I connected up the HP606B and input a signal at 2.0mc and I could just hear and could tune in and out of resonance the signal, indicated that it is possible to "ram a signal through" but that's all. I checked all B+ voltages, plates and screens and everything looked okay (probably those replacement filter caps were a major factor.)

Weak Tubes and Weird Bias Cell Hook-up - I tested all of the tubes just to see where that would lead. The 6L7 Mixer tube barely moved the TV-7 meter. The 6C5 LO was about 50% below minimum acceptable. The 6F5 1st AF Amp barely moved the meter. The three 6K7 tubes all measured exactly at minimum acceptable (okay to use but not like new.) Only the 6F6, the 5Z4 and the 6H6 tested as good quality tubes.

With all tested-good tubes installed, the SX-9 still didn't function much. The bias cell is in series with the audio signal which seems like an odd way to achieve the -1.5vdc grid bias, but it does work. Audio injected at the 6F6 grid sounds fine, not so at the 6F5 grid. If 465kc is injected at the IF grid or the Mixer grid, only a pulse of a signal is heard and then it drops off. A 2.0mc signal at the antenna will produce a weak signal.

Repair versus Restoration? - It seems obvious that this SX-9 has several defective capacitors. But, repairing a set to just "get it to work" is really only a temporary fix. This SX-9 is going to have to be restored. There are other issues that can then be taken care of, such as the severely bent front panel that's going to require panel removal in order to straighten it properly. The panel seems to be tempered and removing the bending is going to be difficult. The band indicator pointer is a crudely-made repair piece that needs to be redone or replaced. The wire index behind the band spread dial is totally bent and will require dismounting to fix. So, lots of mechanical problems with the SX-9 besides the numerous defective components. Restoration is the only practical solution.

Inspection - After disassembly, I discovered that the panel finish was stained and in very poor condition. The more I cleaned it, the worse it looked. I'm not sure what type of matte finish was applied but since the silk-screened nomenclature is still present, I'd assume it's the original finish. There are problems with some sort of black tar that was used to mount or seal the loudspeaker to the panel and that has gotten everywhere. It can be removed with denatured alcohol but it appears to have stained the panel finish. The plastic Band Spread dial has been damaged by the black tar and the dial itself is severely warped. Additionally, all of the small knobs are not original but are WWII versions that look similar (have hex-socket dual set screws.) The loudspeaker is original and has a special tapped field coil and a special output transformer. The cone was damaged in the past and was glued with Duco cement. The speaker can't be replaced unless another original could be found. I don't think it can be reconed either since it has a metal upper spider support for the VC-cone with no lower flex support for the VC-cone.

Besides the usual restoration problems, the band spead dial should be replaced, the loudspeaker really needs rebuilding and I'm sure there are even more problems lurking. The more I look at this SX-9,...I'm thinking "shelf queen." 

The Hallicrafters, Inc. -   5 -T "SKY-BUDDY"

(1935 "Boy Dial" versions built by Howard Radio Co.)

Bill Halligan believed there should be a market for an inexpensive shortwave receiver that would perform well enough to inspire young enthusiasts to choose "RADIO" as a hobby or as a career. First, the youngster would become an SWL (Short Wave Listener,) then go on to become a ham and finally would make RADIO a career. The idea certainly appealed to the parents of technically talented kids and it was those parents that were the target market for "entry level receivers." Halligan's goal was to be able to produce the receiver at a cost low enough for the Depression-era parents of 1935 to afford the set for one of their "radio-minded" youngsters. At $29, the 5-T Sky Buddy certainly wasn't low priced (equivalent to $660 in 2024 dollars.) But with 5 tubes the performance was adequate to inspire hoards of young users to become hams. Many hams remember that their first listening experience to shortwave reception was on an "entry level receiver." (Of course, today most older hams remember the Sky Buddy's post-WWII successor, the S-38, as providing their first SW reception. As a kid, my first successful shortwave listening was on a second-hand "borrowed" S-38B.)

To ensure that the purchasers knew who the intended users of the 5-T Sky Buddy were supposed to be, the first production receivers included a picture of a "young ham sending with a bug" on the dial. There was a story that Bill Halligan had a neighbor that couldn't afford an expensive receiver for his radio-enthusiast son who happened to have been named Buddy. Supposedly, Halligan built (or had built) a small receiver for the boy and called it a "Sky Buddy" - an apocryphal tale? - perhaps - but this was the inspiration for Halligan and Hallicrafters to offer an inexpensive, entry-level receiver. 

The serial number stamped on this "Boy Dial" 5-T chassis is 024919.

The 5-T "Boy Dial"

Even though the 5-T Sky-Buddy is thought of as a product of Hallicrafters, the first versions built in 1935 (with the "Boy Dial") were built at the Silver-Marshall-McMurdo Silver building in Chicago by Howard Radio Company. Many of the assembly and component characteristics of the "Boy Dial" 5-T are recognizable as a Howard Radio product. The external-to-the-chassis BFO coil, the belt-drive tuning along with the dial and speaker bezels and other construction techniques are very recognizable as characteristics found in other receivers built by Howard Radio Company. Another interesting bit of speculation is that McMurdo Silver designed the 5-T. It's certainly a possibility and certain design characteristics seem to have basic similarities to McMurdo's designs (the tapped field coil speaker and the elaborate belt-drive for such a small radio, for instance.)

The next run of 5-T radios were produced after Halligan had set-up at the Echophone building and was producing his own products. The dial mechanism was changed to eliminate the belt-drive. The "Boy Dial" was also gone. No doubt, the 5-T was successful and popular. So popular, in fact, that many adults were buying the receiver for their own use. When the next run of Sky Buddy receivers were produced, they were actually built by Hallicrafters and the new dial had SKY BUDDY in a banner across the center of the dial. Maybe this was to make the receiver look less like a kid's toy and more like a serious ham receiver.

The Sky-Buddy continued on with various changes made but the basic concept of an inexpensive entry-level receiver remained. Sky-Buddy production was halted as WWII began. When Hallicrafters entry-level receiver returned after WWII, the circuit was changed to an AC-DC radio and the receiver was designated as Model S-38.

June 29, 2024 - This 5-T project was placed on the "back burner" after I had found the necessary parts to do the restoration. I know I was trying to figure a way to build the waxed-cardboard box that held the electrolytic filter capacitors. The metal multi-section is pretty obvious as a "hack-job repair attempt." The non-original WW resistors also need replacements. There weren't too many paper caps and all were originals, so I had planned on restuffing the paper dielectric capacitor shells. For some reason, I assembled the receiver and never did the work. I think I had put the 5-T back together before I had found all of the correct hex knobs for the receiver. What needs to be done now isn't really that much to complete the job. First will be a complete recheck and assessment of the receiver and then proceed from there.
 

The photo to the right was taken when I initially got the 5-T. Obvious is the missing speaker. Besides the high DCR of the field coil, the replacement speaker also needs to have the audio output transformer mounted on the speaker. There's no plug on the speaker cable and it's connected to the chassis by soldering the wires.

Under the chassis showing careless repairs. The multi-section electrolytic isn't original. It should be a cardboard box type of  electrolytic. The ceramic resistor and the two green wire wound resistors aren't original.

Inspection - July 2, 2024 - Power transformer tests okay. IF transformers test okay. The Candohm tests okay. The loudspeaker isn't the original so it doesn't have the bucking winding. The field coil is part of the negative B- and is part of the grid bias used on the 42. This wiring was missing since someone had tried to use a PM speaker as a replacement years ago. The speaker I have is a single field coil winding and that will work as far as the B-. There was a popular trend in the thirties that believed the field coil presence near the voice coil caused hum in the VC of the speaker so a bucking winding was added to the VC connection to "buck" the hum. It's a very slight hum at the most and the bucking connection wasn't ever used on all field coil speakers. The audio output transformer on the speaker is the correct vintage for working with a 42 (it was salvaged from a Play-Mor radio.) There are 11 paper dielectric capacitors and two electrolytic filter capacitors used in the 5-T circuit. Moderate wear, not excessive.

The Plan - In the past I had thought about restuffing the original Sprague paper dielectric capacitors but then I'm looking at the very non-original loudspeaker that's going to require some modification to the circuit to function correctly. Also, there's the filter capacitor situation. The 5-T will never be original or even look original upon close inspection. The dial drive system was missing parts when I got the 5-T but a rubber belt was able to function without the missing parts. So, I'm thinking about doing the same thing with the 5-T that I did with the Howard 450A and paint modern polyfilm capacitors Raw Sienna which makes them look vintage. The electrolytics will be painted maroon to look vintage. However,... 

Reconsidering the Project - Like before, it's difficult to become inspired by this particular 5-T example. It would have to be modified from original because of the missing loudspeaker and the dial tuning using a rubber belt doesn't work like the original. To have this 5-T operational wouldn't really reveal anything important. It's a five-tube receiver of limited performance capabilities. Like the SX-9, the 5-T looks like it's going to remain a "shelf queen." I'm going to reassemble it and consider that it's a decent example of the 5-T but not an original one.