Recently I acquired a Philco model 39-55 from eBay. This was the "el cheapo" version of the 39-116 Mystery Control (wireless remote) Radio. The obvious difference between the two models is that the 39-55 just has the standard broadcast band while the 39-116 has BC plus two shortwave bands. But, it turns out that there are a few other differences as well, implemented, I presume, to make the 39-55 a less expensive radio to manufacture. These are:
(1) No tracking RF amp stage
(2) One AGC tube instead of two in the Mystery Control receiver portion of the set
(3) No phase inverter triode driving the type 42 output tubes.
Item #3 is what I found interesting. Looking at the schematic and not seeing a phase inverter tube (the 39-116 model uses a type 37 triode as an inverter/driver), I concluded that the two 42 tubes were being driven in parallel rather than push-pull configuration, to boost the power output over that of a single tube. I had repaired the set to the point where I was picking up stations, but the sound out of the speaker was strained and distorted. I removed each 42 tube individually to isolate the problem and found that the "bottom" tube on the schematic was not only weak, it was totally inoperative. The tube tested good, had normal heater, screen, and plate voltages, but just didn't work. Very strange.

At this point I re-consulted the schematic and discovered that my initial judgement that the tubes were operating in parallel was wrong. What Philco did was insert a dropping resistor between the screen bias pin of the upper tube and B+, then ran a 0.01 cap from the pin to the grid of the lower tube. This basically caused the screen voltage at the pin to fluctuate along with the plate-cathode current, 180 degrees out of phase with the signal at the grid of the top tube. So, the tube stack was indeed working as a push-pull circuit, albeit with a few nanoseconds of phase delay between the two tubes.

This explained why the circuit functioned with the "top" tube in place and didn't work with just the bottom tube installed- the top tube is the master and must be in place and working in order to invert the signal going to the bottom tube. However, this didn't explain why the sound was so low-quality.

Checking the radio against the schematic, I discovered that a previous owner (the set had been totally recapped) had omitted the coupling cap between the top tube's screen pin and the bottom tubes grid. I installed this cap and the audio became classic big-box Philco: full (but not thumping) bass with crisp trebles. It sounded great and I honestly can not discern any qualitative differences between it and my 39-116 set which I restored several years ago.

So, my question is this: since the circuit seems to work well, and it accomplishes its job with one less tube, why wasn't this audio output stage design used extensively? In 40 years of repairing tube radios, this is the first set that I've encountered that uses this circuit.

Your description was used in quite a few cheaper sets and in some battery sets. In the battery sets I assume not only to cut down production costs but to also cut down on the current draw from the filament and B+ of the set's battery.

Clifton

:Recently I acquired a Philco model 39-55 from eBay. This was the "el cheapo" version of the 39-116 Mystery Control (wireless remote) Radio. The obvious difference between the two models is that the 39-55 just has the standard broadcast band while the 39-116 has BC plus two shortwave bands. But, it turns out that there are a few other differences as well, implemented, I presume, to make the 39-55 a less expensive radio to manufacture. These are:
: (1) No tracking RF amp stage
: (2) One AGC tube instead of two in the Mystery Control receiver portion of the set
: (3) No phase inverter triode driving the type 42 output tubes.
: Item #3 is what I found interesting. Looking at the schematic and not seeing a phase inverter tube (the 39-116 model uses a type 37 triode as an inverter/driver), I concluded that the two 42 tubes were being driven in parallel rather than push-pull configuration, to boost the power output over that of a single tube. I had repaired the set to the point where I was picking up stations, but the sound out of the speaker was strained and distorted. I removed each 42 tube individually to isolate the problem and found that the "bottom" tube on the schematic was not only weak, it was totally inoperative. The tube tested good, had normal heater, screen, and plate voltages, but just didn't work. Very strange.
:
:At this point I re-consulted the schematic and discovered that my initial judgement that the tubes were operating in parallel was wrong. What Philco did was insert a dropping resistor between the screen bias pin of the upper tube and B+, then ran a 0.01 cap from the pin to the grid of the lower tube. This basically caused the screen voltage at the pin to fluctuate along with the plate-cathode current, 180 degrees out of phase with the signal at the grid of the top tube. So, the tube stack was indeed working as a push-pull circuit, albeit with a few nanoseconds of phase delay between the two tubes.
:
:This explained why the circuit functioned with the "top" tube in place and didn't work with just the bottom tube installed- the top tube is the master and must be in place and working in order to invert the signal going to the bottom tube. However, this didn't explain why the sound was so low-quality.
:
:Checking the radio against the schematic, I discovered that a previous owner (the set had been totally recapped) had omitted the coupling cap between the top tube's screen pin and the bottom tubes grid. I installed this cap and the audio became classic big-box Philco: full (but not thumping) bass with crisp trebles. It sounded great and I honestly can not discern any qualitative differences between it and my 39-116 set which I restored several years ago.
:
:So, my question is this: since the circuit seems to work well, and it accomplishes its job with one less tube, why wasn't this audio output stage design used extensively? In 40 years of repairing tube radios, this is the first set that I've encountered that uses this circuit.
:

Hi CV,
Interesting. Your description is well written, too.
I can think of two reasons. The first could be a patent problem. The inventor of that may have been asking too much money to patent his invention.
Another possibility is that it was hard on the tubes and that is why you have a bad one.
The second idea isn't too likely, though. Bad 42's that test good are common. I suspect a patent problem.

All the Best,

Bill

Bill, it turned out that the two 42s were both good. In fact, they tested within a few points of each other. The "bottom" tube was inop because whoever recapped the set before I acquired it had neglected to install the coupling cap from the top tube's screen to the bottom tube's grid- so there was no inverted audio signal reaching it. When I installed the cap the previously "dead" tube worked fine, at least to my ears.

I suppose that there is distortion inherent in this scheme due to the audio signal being delayed by the propagation time through the tube (nanoseconds) and the 90 degree signal phase shift of the coupling capacitor; but the actual result sounds very good. I'm pretty sure that Philco wouldn't have risked damaging their reputation on a flaky audio output stage design; but maybe this Mystery set was considered by them to be a low-sales-volume novelty which didn't matter much one way or the other, so they were open to some experimentation with it.