Het-Duo-Gen/Brush

     C. E. Brush described the theory of operation and the construction of his “Het-Duo-Gen” in a series of Radio Digest articles (published weekly, starting with the December 27, 1924, edition). Mr. Brush didn’t follow the rest of the pack and he came up with some new ideas and techniques with his new circuit.
     A first glance of the chassis shows little evidence of a superheterodyne. It only has six tubes, and three of those are associated with audio frequency transformers. There are no series of IF transformers, which are the normal indication of a super. The electronics looks more like a TRF or some type of reflex circuit. But it is definitely a superheterodyne; it has an oscillator and a mixer (first detector)! The first detector is a regenerative type, controlled by the large vernier knob in the center of the front panel. This is the same regeneration control that is on many early superhets, but most of the time this control takes a minor position on the panel (perhaps giving the indication of a minor effect on the circuit). It looks like the first detector regeneration control is definitely a major player with the Het-Duo-Gen!
     There is nothing new with oscillators and regenerative first detectors. The revolutionary idea incorporated with the Het-Duo-Gen is a regenerative filter transformer stage (in this case the second detector) running at a frequency of 500 KC! That IF frequency is ten times higher than what 99.9% of the contemporary experimenters were working with. Here we have a pleasant surprise, also, as Mr. Brush points out. The regeneration control on the first detector has to be continually adjusted for peak operation because the frequency changes. Because the frequency is constant on the filter transformer, the regeneration control on this stage can theoretically be fixed and it does not need to be adjusted.
     Mr. Brush believes the extra performance of the second detector made an additional IF stage unnecessary. Instead, the Het-Duo-Gen uses three audio stages to make up the difference. I suspect the radio would probably work better with one more IF stage and just two audio stages. However, I did not experiment with many different combinations of circuits over a period of months like Mr. Brush did. Perhaps tuning the circuit properly with an additional 500 KC IF stage, using a triode tube, was just too difficult for the benefits gained.

     Het-Duo-Gen, as the name implies, is a superheterodyne using two regenerative stages. Here are a few particulars of my set:

1. The front panel is 26” long by 7” high by 3.16” thick.
2. Home-brewed oscillator coupler, filter transformer, and antenna coupler.
3. Audio transformers are unknown, but look like good quality.
4. Tube sockets are not marked and are unknown.
5. Pilot main tuning capacitors and dials (not original to the radio).
6. Pacent rheostats.
7. Signal cabinet.
8. Six tubes.

     Base board layout of the Het-Duo-Gen, from the January 3, 1925, Radio Digest.

     Top chassis view of my set. The radio is very grimy and in bad need of some restoration work. The first detector tube socket is missing. The bad news? The main tuning capacitors and tuning dials were “upgraded” somewhere along the line with Pilot brand parts. The good news? The three audio transformers measure good and the four remaining tubes have good filaments. I believe the tubes have been with the radio for ages and all four of these tubes have a paper label on them with a 1931 date. This radio was probably still being used in the early 1930s. Some remote locations in the 1930s were probably not “electrified” yet with AC power and this battery operated superhet could have still been useful.

     The basic wiring diagram of the Het-Duo-Gen, as copied from the December 27, 1924, Radio Digest magazine. The set is actually wired a little different than this diagram (there are two tube rheostats, two adjustable grid leaks - one on each side of the base panel, and two audio output jacks).

     I normally don’t like messing with the original cosmetics of a set, removing old patina, changing things, etc. However, this is one set that I don’t mind doing restoration work on because it is not a clean original one-owner set. My thoughts right now are to clean this set up, fill in all the unused holes in the base panel, and do other things to make it look nicer. I would like to get this set working again and verify the IF frequency and listen to some distant stations. Put this Het-Duo-Gen through its paces!

April 24, 2014, update:

     I have completed the restoration of the Het-Duo-Gen. One of my original intentions was to fill in all the unnecessary holes in the baseboard and make it look nicer. To do that I needed to disassemble the entire radio. This was kind of a learning experience for me. Wood filler does not take stain very well, so I filled in the holes in the baseboard with small wooden dowels. After a lot of work and sanding I tried giving the baseboard a wood stain to hide my repairs but I could never get the baseboard to look good. The wood dowels stained different than the original wood. I had a feeling that it would do this before I ever started but I tried it anyway. After several attempts at staining I gave up and I ended up just painting the baseboard with a rust-colored flat spray paint. I assembled the entire radio with the painted base panel but was not entirely happy with the results. What to do? Start from scratch all over again with a new base panel! I never should have tried to salvage the old base panel.

     This is how the set looks now. This radio needed more work than I counted on. Somewhere along the line one of the owners replaced the original tuning caps with Pilot tuning caps. After he did this he found out that the support brackets for the front panel/base board would no longer fit (the pilot tuning capacitors stick out more on one side than the other side). The tuning capacitors had to be moved to clear the brackets and several new holes were subsequently drilled in the front panel. I think a couple of different combinations were tried because there were three different large holes and a number of smaller holes for each cap. A couple of the smaller holes could be seen after the Pilot vernier dials were installed. To repair the damage I cut out a section (about 1.5” by 1.5”) of the front panel where each of the main tuning caps were installed and epoxied in new panel material. Then I positioned the tuning capacitors where they were originally installed. Now the Pilot vernier dials cover up the old holes that were visible on the front panel. By the way, this panel was made out of a hard rubber substance that had oxidized. I was able to sand the front panel with 2400-grit sandpaper and then give it a new polish. The original panel looks great!

     This is an inside view of the set. It’s not perfect but I think it looks very nice. I was able to salvage much of the original wiring. I simply cleaned and tightened up the wires on the coils. I replaced some of the stranded wire with red insulation (not original to the radio) with original buss bar.

     This is a view of the oscillator section. I put a piece of circular sheet brass behind the front panel where the main tuning dials are located to hide the sections of the front panel that I replaced.

     This is a view of the filter transformer section. I had to make new baseboard/front panel brass brackets, formed in a partial circle, to make clearance for the Pilot antenna tuning capacitor. The clearance problem can be seen in this photo. Those marks on the end-cap of the filter transformer are the remains of where an earlier type of regeneration trimming capacitor was mounted. The trimming capacitor was replaced a long time ago with a new unit that is controlled with a small knob on the front panel. I have only seen two Het-Duo-Gen radios and they have both been modified with the new style of trimming capacitor. I thought that I would leave the remains of where the earlier type of capacitor was located; that capacitor was in the original plans for the set.

     Initial testing of the set involved putting in a single tube in the oscillator position and making a log of the oscillator frequency at every 10 counts on the dial. The original plans for the set called out a 250 mmfd oscillator tuning capacitor, but the Pilot oscillator tuning capacitor that was upgraded was 350 mmfd. This resulted in lower frequencies than what was originally designed.
     A single tube was installed in the first detector position and a similar log was made of the antenna tuning capacitor. With a log of each dial I was hoping it would aid in tuning in a particular station.
     I measured the peak frequency of the filter transformer by removing the wire on the grid of the first detector and injecting a signal at this point. Both the first and second detector tubes were used for this test. The peak frequency was measured to be about 265 KC on the plate of the second detector. This frequency was supposed to be around 500 KC. I don’t know what the problem is here. Perhaps there was too much wire wound on the secondary of the filter transformer?
     Below is a chart of the oscillator and antenna frequency logs.

     I tried the set with a short antenna about 20 feet long. I received nothing but whistles while turning either of the main dials! The set had internal oscillations. I adjusted the regeneration controls and the rheostat controls to no avail; I could not get the set to stop oscillating. I played around with the set for about one hour before giving up.
     Several days later I was reading some of the Het-Duo-Gen articles and I came across a description for operating the set in the January 31, 1925, Radio-Digest. If all else fails, read the instructions! The instructions detailed the proper way to set the regeneration in the first and second detectors to prevent oscillations. I adjusted the set following the instructions and it worked! I must say though that it is a difficult set to operate. The set seems to be struggling for gain, and both regeneration controls must always be adjusted to try and get the maximum gain (just below the oscillation point). That must be why the regeneration control for the second detector was moved to the front panel.
     I suspect that the original designer was able to get better performance out of his set than what I’ve been able to get out of my set. I think that the set would be very difficult for some 1920s homebrewer to get working correctly. There’s a reason why this set did not get very popular and remains obscure to this very day.
     I like sets that are different though, and I’m glad to have the set in my collection.

Haynes-Griffin Square IF Transformers<     HOME     >Infradyne

.