[RodB]

I'm wondering if it might have something to do with the trimmers in the IF transformers?

Maybe not. They work OK either way.

[dconant]

Didn't have any trouble doing the IF alignment. In fact it was pretty close as it was.

[morzh]

What bothers me is that statement I mentioned, about touching the probe and all works after that.

Could you repeat that experiment?

When the radio craps out, touch that Sig Gen probe anywhere, and kep touching until (or maybe it is any place, the very first you touch) the radio plays normal.
Notice where you touched.
Also notice if this started with the actual probe touch or before that, when you connected GNDs together, or....

The more details the better.

Rod's suggestion is also valid. An IF can could crap out. Not the winding but a trimmer. But then, if the IF still works, how's that possible? Or maybe we think it works and it does not.


I thought of trying to run the signal from the sig gen WHEN the radio craps out, but if touching ANY probe to the radio makes it work again, that is gonna be futile. But if not, if you can possibly run the signal, starting through the Antenna terminal, from the Generator and see that it is weakened, you could possibly try to see exactly where it is attenuated.

[dconant]

Mike it was the signal tracer not generator. I will see if touching the RF probe to anything brings it back.

[morzh]

I know it was the tracer; I mean, if touching anything else would also do it, like any probe that has some capacitance to it (and actually tthe gen probe has signal), taht would complicate things.

[dconant]

Radio is playing normally now waiting for crap out. Then will do sig tracer test. While playing normal I did some tracing. Plate of 6K7 plays nice. Grid of 6K7 has distortion and plate of 6A8 is distored badly. Shouldn't all those points be clear. Speaker sounds good. A friend says sounds like FM (yeah right).

I have a question about C5 and C14 which parts list says are 1mf. I put in .1mf as that is what I took out. I assume the parts list is wrong.

[morzh]

Put in what was before, PROVIDED it was a factory-looking soldering / parts.
if it looked like someone changed then before you, then follow the sch / BOM values.
Again, this is probably not related.

The plate of 6A8 and the grid of 6K7 should sound the same. They can be a bit distorted, as you are listening to a modulated IF.
BTW does your tracer have modes/probes for Audio and RF probing, or all done in the same mode with the same probe?

I am asking because I never got to use a tracer; I do things my own way.

[dconant]

Probes are separate. Eico 147A. I don't really trust it. I'd like to get a modern one. Caps were original looking same as rest (Zenith).

Radio did its thing and now touching RF probe to anything did not bring back signal. Must have been a fluke (not meter).

[dconant]

I will be away the rest of the day so now take a break for a while.

[Eliot Ness]

I might be beating a dead horse here, and you might have tried all of these suggestions, and if so, just ignore.

The most useful initial test is to break the radio down into sections and completely rule out sections. It sounds like you've done that (at least with the audio section) if the Bluetooth never cuts out.

I have seen bad solder joints that looked good on the surface but were just bad or cold joints that somehow got covered with shiny solder. I think you've done enough tube substituting to rule the tubes out, but the tube sockets and all solder joints could be suspect so you might want to systematically start resoldering all joints in the suspect section/s. At this point I'd replace ANY original resistor or capacitor that remains (even if they initially measured OK), especially before the audio section. I'd also seriously consider replacing the sockets of suspected tubes.

I’d also replace that complete candohm if you haven’t already. They are known to intermittently ground to chassis. It looks that it’s just a three-section resistor easily replaced with a terminal strip.

And on my schematic of the 6S254AT (turntable version) C5 and 14 are listed as .1, so you’re OK with those values… I have a parts chassis for the 9 tube ‘38 but not the 6 tube. Fortunately it's a nice roomy chassis to work on, although by now I’m sure you’re sick of looking at the underside!

[morzh]

OK, when back to it:

when the radio works fine, establish the setup with a sig gen, pass the signal, make sure the AVC reacts. Variate the strength, try to make it react, if it does not.
Measure the amplitudes (if you have scope, use scope, if not, use tracer; or use a meter, AC Volts) of the signal along the way:

1. Cap of 6A8
2. Plate of the same
3. Cap of 6K7
4. Plate of the same
5. Plate (any) 6H6.
6. Connection between R5 and R7 (Detected voltage)
7. AVC voltage (top pin C2 or left pin of the R5)
8. All the 3 taps of the antenna secondary (which you measured before together with C2 and the cap of 6A8)

Now, when the radio craps out, with the same exact signal in, re-measure/re-trace the above.

If possible, do both measuring and tracing in each of the two situations. Tracing requires subjective assessment: write down what you hear.

[dconant]

You can say that again Eliot. I think the candohm is a good idea. I have reflowed and soldered all connections and replaced all caps and resistors.

[morzh]

John

Yes, Candohm was and still is my suspect also, but I think it was measured for the voltages (well, for the one that biases the Cathodes, 2V or something, which is the one that matters) and the voltage stayed put.

[MrFixr55]

OK, from the crazy idea department:

If you have a signal generator that can be externally modulated, hook up a pleasant audio source to the generator, pull the 6A8 and connect the generator to the IF stage. Listen to the radio for a long time to determine if the issue occurs. If you play the radio for a very long time (days) and the issue does not recur, you can possibly rule out the IF, Detector, AF stages and the power supply and concentrate on the convertor stage. Set the radio for the highest broadcast band frequency (Tuning cap fully open), couple the signal generator to the antenna and tune the signal generator toll the signal is heard. Play for several hours (days?) If the issue does not occur, check the tuning cap. The aluminum plates of a cap corrode and create "whiskers" similar to the stalactites and stalagmites in a cave. These can short out the RF stage section of the tuning cap, attenuating the signal.

You can use a set of feeler gauges to clean the gaps between the plates of the tuning cap.

[dconant]

I'd like to try MrFixr's idea(s). Can you tell me exactly where to connect the generator to the IF? Are you giving me two things to try? First to the IF stage and second to the antenna term? Sorry if I seem thick headed but I'm not clear on this.

Tuning cap has been clean with contact cleaner and a wire brush similar to a pipe cleaner only a little more coarse.