Radio Boulevard
Western Historic Radio Museum

 

Howard Radio Company (1926 - 1949)
Howard Manufacturing Company (1922 - 1925)
Electric Specialty Company (1921)


Profiling Six Ham Receivers built by Howard Radio Company

Sears-Silvertone - Model 5656A - 1937 - Restoration and Detailed Profile
Sears-Silvertone - Model 5752 "STANDBY" - 1940
Howard Radio Co. - Model 437A - 1941
Howard Radio Co. - Model 450A - 1938
- Restoration and Detailed Profile
Hallicrafters SX-9 - 1935 - Howard Radio Co. Contractor
Hallicrafters 5-T SKY-BUDDY - 1935 - Howard Radio Co. Contractor

 

 By: Henry Rogers WA7YBS


Austin A. Howard

photo from Alan Douglas/Don Patterson

 

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 Mfg. Co. - 1925 Model C Neutrodyne
The Model C is typical of the Howard radios produced in the 1920s. This was a very large and very expensive five-tube Neutrodyne receiver
that sold for $180 during early 1925 (introduced in March 1925.)      photo: eBay

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.


Sears-Roebuck Co. - Silvertone 5656A - 1937
Sears had Howard Radio Company build these eight-tube superheterodyne receivers in 1937. Despite the low selling price of just $49.50, the 5656A was a surprisingly good performer that offered a RF amplifier stage, variable-coupled IF amplifier stage, IF gain control (Sensitivity,) a Standby switch, headphone jack and a built-in 8" Jensen field-coil loudspeaker.

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.


1938 Howard Model 450A Communication Receiver
This was one of Howard's largest and most elaborate communication receivers featuring 12 tubes and tuning from .54mc up to 60mc. Other features included dual IF frequencies, a Crystal Filter, Dual Frequency BFOs, 9 watts of Push-Pull Audio output and a spectacular tuning dial system

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.

WWII and Post-WWII - Howard's ham gear line-up ended when WWII began and Howard Radio Company then built for the war effort. Howard Radio's most famous WWII contribution was as an alternate contractor for the production of BC-779 Hammarlund Super Pro receivers. These BC-779 receivers will have a data plate on the front panel indicating that the receiver was built (or assembled) by Howard Radio Company.

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

Unusual Acquisition - In 2002, back when I had the Western Historic Radio Museum as a real "brick and mortar" museum in Virginia City, Nevada, I was contacted by a ham collector that wanted to sell me a Silvertone ham receiver. He described the receiver as looking something like a Hallicrafters Super-7. The price was very reasonable and he offered to deliver the receiver in a most unusual way. It happened that Ray (the seller) was a private medical transport charter pilot for the most of the Western USA. He would fly people that chartered his service from Western rural areas into larger metropolitan areas for non-emergency hospital surgeries, treatments or appointments. The radio transaction was to take place a few weeks later when he was going to be flying from his SoCal home-location up to Nevada to do a "pick up" and transport that person from a rural location in Nevada to the Reno Airport for land transport to one of the Reno hospitals. He wanted to come up to Virginia City afterwards since the medical transport was a "drop off" and he only had his return trip to SoCal left. He wanted to visit the WHRM and he also would bring along the 5656A for me to purchase. This all came off without a hitch and the receiver was delivered as expected. The 5656A was in very nice condition with only a couple of exceptions. First, the grille screen was painted black and that allowed the speaker cone to be easily seen. Then there were four knobs that were obviously not original. Everything else looked original and later, when I started the rebuild and could look under the chassis, I discovered that the 5656A was all original parts. It was really hard to believe that nothing had been changed under the chassis,...no repairs, no modifications,...very unusual for a relatively inexpensive receiver.

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


Under the BFO Chassis

The 5656A does a good job on CW signals although, like modern SSB, tuning is a little difficult but the BFO can be used for fine tuning. Actually, the tuning dial's sweep-second hand pointer is very handy when tuning through a ham band as it allows the operator to "visually expand" what is really a very narrow section of the main tuning dial. The 8" Jensen loudspeaker is a good-quality field coil electrodynamic speaker that sounds very nice on AM signals. The Tone control does mellow the reproduced sound considerably.

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.

This intense level of signal wouldn't happen when using a smaller antenna but I'm testing using my large Collinear Array that has some gain, so it's the "worst case scenario" for trying to successfully demodulate extremely strong SSB signals on 80M and 40M. SW-BC signals are easy to tune in and sound very nice although it does depend on the broadcast material. When in Virginia City, I used a 60' EFW with this receiver and the combination of poor location with an even poorer antenna didn't provide very impressive results. But, a RF-friendly QTH and a large antenna system will help any receiver, so the 5656A went from unusable to a receiver that could easily and successfully be used on 75M AM. For SWLing, it seems to perform quite well and will easily receive stations up to the 19M band and even 16M SW stations are easy to receive with the upper limit of tuning being 18mc. As would be expected by looking at the dial, the tuned frequency accuracy is vague at best and resolution is extremely limited. Most hams at the time would have had a separate Frequency Standard (a 1000kc, 100kc and 10kc oscillator) that could provide a check on band edges. Many hams, especially entry-level hams, would have used a crystal-controlled transmitter and would know their operating frequency, although on-the-air QSOs were usually somewhat "split-frequency" because of the multitude of crystal-controlled transmitters. There isn't a remote standby provided but there is a front panel standby switch. 

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

This Silvertone STANDBY receiver had belonged to KB6BKN Walt Schivo (in fact, Walt currently has a photo of this receiver on his QRZ page.) For a while, there were some Internet postings about this very same receiver with speculation that it was a RME product. Walt sold a lot of his gear to Ham & HiFi about five years ago and that's where I found the receiver in July of 2019. Once I had the receiver and could thoroughly examine it didn't take too long to discover that it had been built by Howard Radio Company.
 

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.  >>>


Howard Communication Receiver Model 450A   ca: 1938    "as found condition - May 11, 2024"
The industrial-scientific instrument-type dials appealed to techno-minded hams but note that the glass dial cover has slipped out of its mounting clips and the top edge of the glass is showing.

>>>   Selling price for the 450A was $95.45 and with the optional crystal filter installed the price was $105.45 in 1938 and 1939. The selling price of $105.45 put the Howard 450A at about the same price as the larger Breting receivers (the 14AX or the 40) and the Patterson PR-15 that was then being built by Pierson-DeLane. The larger communication receivers from the big three, National, Hammarlund and Hallicrafters, or even RME, were typically at least 50% higher in cost, so price was certainly an important factor in the Howard 450A sales. But, even $105 in 1938 was quite expensive being equivalent to about $2200 today (2024.) Besides the comparatively low price, there was the attraction and the visual appeal of the 450A with its impressive dials looking like whirling stop watches, pressure gauges and barometer needles. The 45kljhlkjh0A's industrial-scientific instrument-type appearance must have prodded prospective techno-minded ham buyers with their purchasing decision.
The Dual IF Function - The IF on Bands E and F, or 16mc up to 65mc, operates on 1560kc. This 1560kc IF uses two dedicated IF transformers that are specifically for 1560kc and function with the two 6K7 IF amplifier tubes. When selecting to either Band E or Band F, the IF isn't switched along with the Band Switch function. The operator must also place the Crystal Filter switch to "IF SHIFT BAND E-F" and the BFO SHIFT toggle switch has to be placed in the E-F position, if the BFO is to be used. When using Band A, B, C or D the 465kc IF must be used. The 465kc IF has two dedicated IF transformers specifically for 465kc and they will function with the two 6K7 IF amplifier tubes. For these bands (.54 to 15mc,) again, the band switch doesn't change to the 465kc IF amplifier circuitry. The operator must set the Crystal Filter switch to either BROAD or SHARP for no crystal filter or to XTAL IN for crystal filter operation. The BFO SHIFT toggle switch has to be placed in A-B-C-D for 465kc operation.

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.

Dial Details - The dials are glass with multi-color silk screened nomenclature on the back side. With the matte silver back plate, the glass dial is spaced slightly above the silver plate giving the appearance of depth. The illumination is from the top edge of the glass using two #47 lamps. The main tuning dial scale has color coded indicators on the linear scales with red for amateur, blue for aviation, yellow for police and green for shortwave relay broadcasts. Some 450A receivers had black back plates and were fitted with the glass dials that had white nomenclature along with the red, blue, yellow and green dial scale indications. The convex dial cover is also glass. The stop-watch logging pointer is recessed slightly making it appear as a separate indicator scale. The "band-in-use" indicator is an white-on-black plastic scale mounted to the band switch shaft and shows through a hole in the silver back plate. The "band-in-use" indicator also shows which IF should be used for the particular band selected. The "R" meter is also mounted behind the silver metal plate and shows through a hole. A close-up of the dial is shown to the right.

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

Preliminary Inspection - May 18, 2024 - The 450A chassis shows signs of endless repairs that were accomplished at a "worse than amateur-level workmanship" both in technique and ability. It's obvious that the Candohm resistor was open and has been repaired poorly with sloppily installed resistors replacing the open sections. The 2.5K 0.5W section of the Candohm was replaced with a 10 watt vitreous-enamel coated Ohmite resistor, certainly "over-kill" and showing the lack of comprehending the basics of Ohm's Law on the part of the repair technician. Most of the capacitors have been replaced with a variety that is either because of repairs using "junk box parts" or that the repairs took place at different time periods. Fairly modern Orange Drops are installed here and there implying that some repair work was performed in the last couple of decades. The few vintage capacitors remaining are Aerovox brand but even they aren't original to the 450A. The electrolytic "metal can" capacitors appear original but have a USN "anchor" stamp on top, so they are more than likely very old replacements since they are both dual 8uf cans. The schematic shows a single 8uf can and a dual 8uf can being used originally.

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. 

Mechanical Inspection - The entire chassis is copper plated and the copper is in pretty good condition with just a few small corrosion spots. The entire IF shielded box is also copper plated and in fairly good condition. There's a non-original punched hole in the rear apron of the chassis that may have been for an auxiliary socket. There's also a 1/2" hole on top of the chassis behind the tuning condensers. I doesn't appear to have any function and doesn't look original. Both holes are very neatly done (unusual.)

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.

Past Repair Work - Often times "worse-then" amateur-level repairs do more damage then actually repairing anything. When dealing with low-cost, high tube-count receivers, such as the Breting 12 or 14 and the Patterson receivers, it seems that these receivers required many more repairs than comparable receivers from the Big Three (National, Hammarlund and Hallicrafters) or RME. I'm not sure why this is happens but the quality of components doesn't seem to be a factor. The passive components are name-brand parts,...same as used by the Big Three. The exception might be that the power transformers used were rated for intermittent service, which is common for consumer-entertainment radios, and then the ham-owner exceeded that operational limitation by having the receiver powered-up for hours-on-end. Also, often problems discovered seem to have been caused by carelessness or inexperience resulting in circuit component damage on an extensive level. Most of the problems seem to have been with the power supply section and the audio output section, which is where the components are "pushed" to their maximum ratings and where most failures would naturally occur. Although one does run into these types of problems on National, Hammarlund, Hallicrafters or RME receivers once in a while, the low-cost, high tube-count receivers from Breting, Patterson and, apparently Howard, all seem to suffer from what seems like a litany of endless repairs that were accomplished by amateurs with little or no experience in troubleshooting or in the proper techniques of repair work. In almost all cases, these receivers will not function correctly or don't function at all.

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."

>>>   All paper dielectric capacitors will be polyfilms installed into vintage Aerovox shells to preserve the under-chassis appearance as much as possible (this is just a "plan." UPDATE: The Aerovox caps are replacements, the original Howard capacitors of which only two have survived just have a value and part number, no manufacturer shown.) For the audio output transformer,...the transformer from the HRO-60 is perfect. It's the same physical size as the original Howard transformer and has nearly the same specs. The primary is for P-P 6V6 tubes and the secondary is 8Z and 500Z, so it's a very close match. The original power transformer dimensions were 3.5" x 4.0" and it was a tub-mount. So far I've found one power transformer that is very close for all of the necessary windings but it has one extra winding that produces about 25vac. This transformer is the correct size but uses four bolts for mounting where the original power transformer used two bolts. The replacement xmfr is ink-stamp dated July 29, 1938 and has a hand-written note indicating it came out of a large Wards (Wells-Gardner) radio. Some of the 13 and 15 tube W-G radios had a 25vac winding for four series wired 6 volt tubes. This W-G transformer will work and would look stock but I still have many other power transformers to go through and I might find one that's an exact match.

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.

Stripping the Chassis of Undesired Parts - May 26, 2024 - I removed the defective audio output transformer by drilling out the rivets. The wire leads were clipped at the bad transformer because I will use them to extend the leads on the HRO-60 transformer. I unsoldered all of the wire leads that were from the undersize power transformer. The transformer was then dismounted and removed. When it was placed next to the W-G replacement transformer, the size difference was surprising with the W-G being about 50% larger (yet it fits into the chassis opening and aligns with the "shadow outline" from the old "long gone" original.) The only problem is that the W-G requires four corner holes for mounting where the original used two center-side mounting holes (and there aren't any center holes in the W-G xmfr.) I still have a few more transformers to look at but so far the W-G is the best replacement other than requiring drilling four mounting holes. (No exact replacement was found. The transformer I'm going to use is probably from one of the 13 or 15 tube W-G radios. These radios had a 25vac winding for four 6.3vac tubes wired in series. The 25vac winding isn't necessary for this application and the regular 6.3vac winding can supply the filament current needed for the 450A. The +HV winding is typical for P-P audio applications rated at 8 watts being about 370-0-370vac (with 120vac on the primary.) With the rectifier voltage drop and the IR drops involved, the resulting B+ at the 6V6 plates should be around +250vdc but this is also dependent on the AC line operating voltage which probably should be 115vac for this receiver.)

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.)

The Candohm Voltage Divider - Candohm dividers always were very prone to opening up on the various sections. Usually, an associated bypass capacitor would fail and the increased current would exceed the wire wound sections and one or more sections would open up. Most repairs would parallel a new resistor connected to the terminals of the open section to make the repair. This approach would probably result in the neatest appearance since finding an original Candohm would be impossible and repairing the bad Candohm would require rewinding the burnt-out sections with new ni-chrome wire and that could only be acccomplished if the original could even be taken apart (they weren't designed for repair,...just replacement,...and maybe not even that since it's riveted to the chassis.)

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.

>>>  But, even if the B+ actually was +330vdc as the receiver was "warming up" before the tubes start to draw current, the resulting total Candohm static dissipation only increases from 14ma to 18.75ma and the 2.5K or the 1.5K resistors' dissipation only increases a few tenths of a watt (total Candohm dissipation at +330vdc would be 5 watts but it would only be for about 15 to 25 seconds.) The result is a lot of variables that aren't going to affect the overall function of the voltage divider. The replacement resistor can have a much higher dissipation rating and, with modern resistors, their physical size is quite small. So I could use the exact resistance value that was required, I ordered an older NOS replacement wire wound resistor. Even though the replacement resistor has a 6W dissipation, physically, this type of resistor is only 1" long and .250" diameter, so mounting it on the Candohm terminals won't be too obvious of a repair (conveniently, the resistor is black and that also helps to hide its installation.)
Push-Pull 6V6 Specs - For around 10 watts output power, the 6V6 plate voltage should be around +250vdc. The screen voltage will also be around +250vdc. The grid bias is determined by the value of the cathode resistor (250Ω) resulting in about -15vdc bias that shows up as +15vdc on the cathode.

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.

Wire Harvesting - I'm sure I'll be needing some vintage hook-up wire as I rebuild the 450A. The hook-up wire is solid conductor 22ga. TC with "push-back" cloth insulation. Almost identical wire was used in Hallicrafters receivers about the same time. I have several SX-28 "junk chassis" that can provide this type of wire in the correct color insulation. Having ample wire supplies will allow for proper routing when doing this rebuild.

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.

Power Transformer and Audio Output Transformer - Jun 7, 2024 - Wired the power transformer into the circuit. Mounted the HRO-60 audio output transformer and wired the secondary into the circuit. I had to recondition the Phone Jack wiring and reconditioned the wiring for the Standby switch. Routed all of the B+ wires to the proper terminals on the Candohm resistor. Installed the 2.5K 6W WW resistor to the open terminals of the Candohm. This resistor is very small for its dissipation rating and even that 6W rating is over ten times what's actually needed (it was the smallest dissipation resistor that was convenient to order.)

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 couldn't find a suitable dog-bone 1K 1W resistor in the style needed so I took a JAN 1K 1W CC and painted it in the BED code but with a band in place of the dot. There are a few of this style original resistor in the receiver so it does match.

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 19, 2024 - Three electrolytic filter capacitors are required to replace the single 8uf and the dual 8uf capacitors. I'm not going to remove the can filters but they will be disconnected. Under the chassis I'll install three 10uf 450vdc for the 8uf electrolytic capacitors. Modern electrolytics are very familiar-looking so I've painted the replacements maroon so they won't "stand-out" as much. Finished all of the polyfilm capacitor installation. I installed one of the 10uf electrolytics. One lead on the filter choke needed an extension added. All that remains will be to install to two remaining 10uf electrolytics and then install a suitable AC power cord (I found a vintage cord with a molded plug that should look original.)

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. 

Further Testing and 465kc and 1560kc IF Alignment - June 22, 2024 - Although the 450A pulled in signals quite well on AM-BC, that's not particularly difficult for almost any receiver,...even a crystal set. Further testing at higher frequencies resulted in no reception of anything on shortwave. I connected up an outdoor wire antenna and I did manage to hear 15mc WWV but nothing else. Since 15mc WWV was tuned at about 16.5mc it was apparent that the first thing required would be to get the IF in alignment. The 465kc IF trimmers are located on the top of the IF shielded box. The trimmers were so far out of adjustment that I had to inject a 465kc signal into the grid of the second IF and peak that transformer. Then move the signal injection to the grid of first IF and peak that transformer. Finally, a signal was injected at the Mixer grid and the remaining IF transformers peaked. The trimmers were way, way, way off. Nothing was getting through unless the signal was tremendously strong. Also, when adjusting the 465kc IF, the switch has to be in SHARP so that all of the transformers are being used. Of course, the crystal was given the sweep treatment to find out exactly its resonant frequency for the final IF adjustment. The 1560kc IF is adjusted using the trimmers that are on the lower side of the IF shielded box. The 1560kc IF was moderately off, not anything like how far off the 465kc IF was, but adjusting the 1560kc trimmers did improve that IF section's response. When switching back to the lower frequencies and the 465kc IF, the improvement was unbelievable. Signals on 20M were pouring in (well, it was Field Day) and after a little adjusting of the BFO, the SSB signals demodulated pretty well. RF gain had to be reduced a lot which not unexpected. WWV 15mc was now very close to the 15mc index mark (that's how far off the IF was!) WWV 10mc was very close. Nothing heard on 40M. I had to DeOxit the AVC switch to get it to work and that got the R Meter also working correctly on AM signals. RF Tracking Alignment - A few unusual things in the 450A RF section. The component assembly drawing shows tubular trimmer capacitors but this receiver has vertical sliding plate trimmers. Each LO band has a trimmer and a padder. This allows adjusting both the upper end of the dial and the lower end of the dial for very accurate tracking (the exception is a padder isn't provided on Band F.) Access to the padders requires a right-angle blade tool. A blade screwdriver can be used with some difficulty. Band A (AM-BC) was quite a bit off. Band B was also off. Band C and Band D were very close to tracking correctly and only needed a slight adjustment. Band E requires the 1560kc IF but the RF tracking adjusted fine. Band F (up to 60mc) had to use a harmonic of 30mc but alignment was okay.

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. 

Performance - I first tested the 450A using a 50' end-fed wire antenna. Performance was okay but due to wide impedance variations, 40M reception was way down while 20M was pretty good. I moved the receiver into the ham shack so I hook up the Collinear Array antenna. 40M perked right up with a matched antenna. 20M was very active with signals. This was about 3:30PM in June, so 80M was dead. I tried Band E and switched the IF to 1560kc but nothing was heard. I'll try some test signals from the HP606B RF Signal Generator. When doing the alignment, all bands received signals. The 450A sensitivity is decent and typical of late-thirties design. The dual IF concept might have looked good on paper but it complicates the circuitry and quantity of components with little benefit. The Crystal Filter works quite well. The entire receiver seems a bit "flimsy" in construction although it might only be in the RF front end that's using lead-supported coils and those leads are going directly to the band switch tabs (asking for stability problems?) There's a lot of wear in the RF gain control but if adjusted carefully it's okay. The TONE control pot might be defective since it really doesn't do anything (typical RC Tone circuit.) The R-meter works fine.

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.

Ant/RF Coil Impedance on Bands B and C - Although during alignment Bands B and C seemed to work, when connected to an antenna, the gain is noticeably down. I did use the HP606B to provide a marker signal to confirm this lack of gain. Since I was receiving the 606B via the antenna, I thought that maybe there was a burned primary on the antenna coil (common problem.) I used a clip-lead to connect the antenna directly to the RF Amplifier grid and bypass the primary. I retuned the antenna and signals jumped out. I disconnected the clip-lead and went back with the antenna connected through the primary. I noticed signals were still present and strong. The antenna tuner tuned radically different than normal for the frequency and retuning makes the difference between signal reception and not receiving much of anything. I used the standard dummy antenna loads when doing the alignment but not the Howard dummy antenna,...maybe that might account for the big change in matching impedance. The other approach might be to align the Antenna coils/RF Amp trimmers with the Antenna Tuner already matched and then tune the 450A trimmers for maximum background noise on the desired frequency. The Howard dummy antenna was the standard using a choke, a capacitor and a resistor with an additional .05uf coupling cap for IF alignments. I just used a 200pf on Band A and B and then a 300Ω for Bands C-F. With the HP606B pumping a signal into a ten foot long test wire located in the next room, I had a signal at the 450A that was being received through the antenna. I found that on Band C the Ant/RF coil was quite a bit off. I set the 606B to 6.0mc, did the same with the 450A and then adjusted the coil trimmer for max signal. I went to 4.0mc on both units and then found the the Antenna Tuner did resonate very closely to the normal settings for 4mc. On Band B, I only used the center conductor since the tuner won't go any lower than 3.5mc. I did the set up at 2.8mc and found that while the Ant/RF trimmer was fairly close, for some reason I had the Mixer trimmer pretty far off. Test signals received okay on Band B. So, just minor adjustment problems. It's hard to say whether it all would have been avoided if I had used the Howard Dummy Antenna. Maybe. The only other place I've seen the RLC Dummy Antenna is in military manuals on some equipment. Most thirties and forties commercial gear alignments use 200pf for MW and 300Ω for SW. It assumes that the antenna to be used will generally be short for MW use but will be good for SW and the receiver's antenna input Z is 300Ω.

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.

Acquisition - Many years ago, in the late-1990s, the Sierra Nevada Amateur Radio Society (SNARS) held their then annual August swap meets at the Stead Airport north of Reno. Old friends NU6AM, K6DGH and KØEOO used to come up for the swap meet on Friday afternoon and would stay overnight at the Gold Hill Hotel just over the Divide from Virginia City. Usually a Friday afternoon visit to the WHRM was part of their trip. On one of their "pre-swap meet" visits, Peter K6DGH saw a Meissner Signal Shifter that I had. Now this wasn't just any old common WWII Signal Shifter. It was a pre-WWII very early all black wrinkle panel with glass covered airplane dial Signal Shifter and it had a separate power supply. It was the only early Signal Shifter like that I'd ever seen (then and since.) Peter "had to have it" to go with his Hallicrafters HT-4 transmitter and offered a particularly generous trade,...a nice condition Hallicrafters SX-9. After seeing the SX-9 that Peter had brought along to sell at the Stead Swap Meet, I agreed to the trade.

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.)

Circuit - The SX-9 uses all metal-octal tubes, specifically a 6K7 RF amplifier, a 6L7 Mixer, a 6C5 HFO, a 6K7 IF amplifier, a 6H6 detector and AVC, a 6K7 BFO, a 6F5 1st audio amplifier, a 6F6 audio output and a 5Z4 rectifier.  Eight tubes are used in a single preselection, single conversion superheterodyne circuit using one stage of IF amplification. "SX" denoted that a Crystal Filter was installed in the IF that operated at 465kc. The rectifier accounts for the ninth tube. The SX-9 was advertised as using ALL metal-octal tubes and was one of the first receivers actually sold with all metal-octal tubes (although the Tobe Deutschmann "TOBE Special" receiver, also from 1935-36, used all metal-octal tubes but it was a ham receiver kit.) The use of the 6H6 duplex-diode and the 6F5 triode as "separate tubes" rather than using the duplex diode-triode 6Q7 was an unusual choice, something like what Zenith Radio was doing in some of their consumer radios to "increase the tube count." But, designs for Hallicrafters were certainly in a state of constant and rapid change at the time since Howard Radio was building the radios at Silver-Marshall/McMurdo Silver. By the SX-10 and SX-11, with Hallicrafters now doing their own assembly, the designs become more conventional and follow the typical single preselection with two IF amplifiers, duplex diode-triode detector/AVC/1st Audio and then Push-Pull audio output with external loudspeaker receivers that had become the "standard ham receiver" of the time. 
Getting Started - As I mentioned, in the 25 years that I've owned this SX-9 I've never even taken the chassis out of the cabinet to have a look underneath. What I find there will be "all new" to me. June 25, 2024 - Brought the SX-9 in from the shop. I pulled it out of the cabinet and found that, except for the two filter capacitors and one obvious Black Beauty cap, it was all original underneath. The filter capacitor were probably replaced decades ago (date code from 1949.) These were top quality bakelite sealed caps so I tested them and they seemed to be okay. I thought I'd try the SX-9 "as is."

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"

Certainly, the initial "Boy Dial" was meant to inspire young enthusiasts. A close-up of the dial is shown in the photo to the left. Note in the close-up of the "boy dial" that the boy is depicted using a semi-automatic telegraph key (bug) and is listening using "phones" just like a "pro." Of course, the vest and tie help for that future job interview to get into professional RADIO!

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.

Acquisition and Mechanical Rebuild - This 5-T Sky Buddy was in the collection of Dave Walker (Walker Electronics) of Reno, Nevada. Walker had been a Reno radio/electronics repairman for decades operating in a small converted grocery store on Wilcox St. in downtown Reno. When Walker retired in about 2016 (at age 84!) he sold his parts, equipment, benches and junk radios to a collector friend of mine. The collector friend practically gave the 5-T to me ($25) because it was in need of so much electronic and mechanical help. Walker had never done anything to it, which isn't surprising. Since his job was fixing up radios, it wouldn't also be his hobby. The 5-T was certainly in the same condition as he had found it when I obtained it. It was missing the speaker entirely. The knobs were gone. The tuning didn't work. The cabinet was bent. On and on. I did the mechanical rebuild and found a suitable field coil speaker to replace the missing speaker (the original field coil had a very high DCR of 3500 ohms.) I took apart the dial to install a tuning belt. It took a few years but eventually I found all of the correct hexagonal knobs. The 5-T still is awaiting an electronic restoration in order to become functional once again.

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.  

 
References:

Hard Copy:

1. Radio Manufacturers of the 1920s by Alan Douglas, Volume 2 of 3 - History of Howard Radio Company during the 1920s.

2. Riders Perpetual Troubleshooter's Manual, Volume X, Howard Radio pages 10-28 to 10-31 - covers the Howard Model 450A, schematic, component layout, top chassis layout, specifications and alignment procedure.

3. Communications Receivers - The Vacuum Tube Era - 1932-1981 4th Edition by Ray Moore - covers many Howard communication receivers from the 1930s, general information.

4. Los Angeles Radio Manufacturing by Floyd Paul - Covers RCA Superheterodyne Licensing policy.

Online:

1. Alan Douglas on Hallicrafters - Alan Douglas also wrote a short paper on the origins of Hallicrafters that is online. It provides a lot of information on Howard Radio Company's involvement with McMurdo Silver and Silver-Marshall Company along with Halligan's initial deals with Silver-Marshall.

2. N7RK Dave Hollander's website - has a section on Howard communication receivers and has photos of many of the smaller receivers. Dave does have quite a bit of info on the nine-tube Howard Model 440.

3. AWA Old Timer's Bulletin - Hallicrafters 5-T "Boy Dial" Receiver - mentions that the "Boy Dial" versions were built by Howard Radio Company.

 

 

Return to 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 © 1997/2024