[RGK20m3]

Estate sale find, was told it was “restored “, but obviously not, all there with cabinet in fair original finish. Pulled the chassis and found that the filter caps had been replaced (extent of the restoration) probably in the 70-80s (Sprauge orange and blue TVA label). Also the power transformer had been replaced with a Philco part that has the rectifier on the end bell. Brought up slowly on a variac, had sound at about 90v, but at very limited volume even when brought up to line voltage.
Tested tubes on my emissions tester, all ok. Measured antenna coils and both had primary and secondary continuity; oscillator coils seem ok as dial tracking is good. Replaced all the paper and wax caps, and restuffed some period electrolytic cans for the power supply. Out of spec resistors replaced, most were within 10% so left alone. Speaker field coil measures right on spec, and output trans and voice coil measure good.
Fired up and has good volume and selectivity initially; after a minute or two the volume drops in half. Sometimes probing the b+ will bring volume back for another minute or two. The b+ and bias voltages do not change appreciably when this occurs. The AVC voltage does not seem to change, either when going from a station to off station, or when the volume drop occurs. The screen voltages do not change, either.  I tried a new 2nd detector tube, along with a known good IF tube.
Volume increases when grid cap of the detector/oscillator is touched, both when first turned on and after the volume drop. I have another 6A8 coming to try….
Any ideas? I did recheck the antenna coils and they are ok.

[MrFixr55]

Hi RGK,

Consider adding an Aux input not only for extra enjoyment but also for "divide and conquer troubleshooting. Does the issue also occur with an external audio signal?

Is the AVC voltage a negative voltage? does it change at all from station to station? The bias for the convertor and IF come from 2 sources. There is a fixed bias from the voltage divider from chassis ground to the center tap of the HV winding on the power supply (common way of getting bias in pre-war radios. The variable bias (AVC) comes from the plate of the 6J5. How they use the 6J5 is quite clever. They are not quite using it as a triode. They use the grid as the plate for the detector diode. They use the plate as the plate of the AVC diode. The 100pF cap #23 supplies the output of the IF to the AVC plate. The 0.05uF cap #3 filters the DC voltage created by AVC. Check these caps and the 1Meg resistors #27 and #15. Consider just replacing these components. The 100 pF caps are mica. Most folks believe that these don't fail but sometimes they do. Not as commonly as a paper-foil cap but once in a while they fail.

In addition, consider heating all the solder connections, cleaning and tightening tube socket pins. Based on the statement that the volume changes when you attempt to make voltage measurements, you may have a cold joint. You may also consider replacing resistors 10, 12, 15 and 27, and Caps 11, and 14. these caps and resistors supply B+ to several parts of the radio's RF and IF section including screen grids. If the screen grid voltage changes, output changes. Resistors may have checked OK and solder joints may look good, but the radio IS 88 years old, and these parts may check OK at static low voltages but fail when current is drawn.

Hope this helps.

[RGK20m3]

That is the direction that I was going to try; injecting an audio signal to see if that stage is normal. I also am going to try determining current draw through the circuits in the audio stage, as that obviously affects the bias voltages.
AVC voltage is -2.8v, and only varies by a few tenths. Decoupling cap and the bias resistors (27,33) have been replaced. Screen voltages on the mixer and IF tubes is rock steady; decoupling cap has been replaced. I did measure if there was any voltage present on the ground points to the chassis, none. I did disassemble the volume potentiometer and cleaned it. I did clean the tube sockets and tension the terminals. Also poked every wire, coil,and solder joint with no voltage fluctuation or popping from the speaker. Hopefully the 6A8 will fix it, but hope is a bad strategy…..

[DaleHCook]

Note that cleaning a potentiometer requires the correct chemical. Chemicals such as DeoxIT D5 are contact cleaners, designed for switch contacts. To clean and lubricate potentiometers use DeoxIT F5 Faderlube.

Chemicals not designed for electronic use, such as WD-40, should never be used on electronic parts, only on mechanical fasteners. Their ingredients may adversely affect electronic parts.

[RGK20m3]

I actually removed it, disassembled, and cleaned the brass wipers with Deoxit; cleaned the carbon track with alcohol.

[DaleHCook]

Faderlube cleans and lubricates, reducing future wear of the carbon track.

[MrFixr55]

Hi RGK,

A couple of things:

When you stated that you checked the chassis ground points and found them at 0V, what did you check them against? There should be about 18V or more between chassis ground and the center tap of the transformer.

The AVC voltage seems a little "low" to me, high being a more negative voltage. for example, to cut the transconductance down to 2 micromhos (minimum stage gain), the tube requires a grid bias of -38V, so a strong signal from a local station should generate a rather negative voltage, which can be read with a DVM (or VTVM if you have one of those). Do not use an analog VOM like a Simpson 250; the input impedance will be too low. The AGC circuit can generate a rather negative voltage, but the current is absolutely miniscule. So, if you have low volume and low AGC voltage, you have a gain issue in the RF and IF sections or an issue with the detector. If the RF or IF stage is overdriven to clipping and AGC is only around -2V, then check the 100pF Cap # 23. Consider changing it. This is what sends IF signal to the AGC diode, so if this cap is bad, no AGC. A strong station may overload the RF and IF stages to the point where there is little signal amplified. However, I suspect that there would also be distortion. However, your symptoms of varying volume may point to this as being an issue.

[RGK20m3]

Good point about the loading bringing the AVG voltage down, my DMM is a Fluke- I will try my VTVM. The only antenna that I have tried is a 10’ length of wire, so maybe even a strong station isn’t of sufficient strength to generate a AVC voltage (or the DMM is loading it and I am chasing my tail…).
What I was referring to was there is no voltage potential between the various solder joints and the metal chassis, as would be present if the was corrosion or oxidation. There is approximately-16v on the power transformer negative, which is the output tube bias.
I will inject a signal and monitor the AVC voltage with my VTVM. How critical is the value of that cap between the plate and grid of the 2nd detector tube (#23, 100pf)? I probably have a 120 or 150….

[RGK20m3]

Well, I probably have been chasing my tail on the AVC circuit; my Fluke DMM has a loading of 10meg.
As the circuit resistance in 2meg, that would cause a 20% increase in load. Bias of the candohm tap is -3.25v, lower that by 20% and that is approximately what I am seeing. I will try the VTVM.

[DaleHCook]

I doubt the VTVM will help. They often have DC input impedances of about 10 meg.

[MrFixr55]

Hi RGK,

I think that the 120 pF may work. It is not exactly a tuned circuit; it is cap coupling to a diode for the purpose of generating a DC voltage. Pulsating, just like that audio, but smoothed by the RC circuit (C#23, R#15). If a VTVM would load it down, try an O'Scope. I seem to remember (maybe correctly, maybe not...) that an RCA schematic stated to use a Channelyst or Voltohmyst (RCA brands of VTVM). I am a little surprised that a DVM or VTVM would load it down. If a DVM is 10 MOhm/V, at 10 V, isn't the impedance 100MOhm? Alternatively, install a "tuning eye" tube. They work of the AVC line, I believe.

Definitely try a longer antenna. If possible, stretch a line outside your house and use a water pipe or made ground.

[RGK20m3]

I am a little confused on the Fluke DMM terminology “internal resistance of 10M or less”, my VTVM is a RCA WV-98 Voltohmyst, so we will see.  I can also try the oscilloscope method. 
Also more than a little confused as to why probing anywhere on the B+ that causes a click or pop will restore the audio to normal volume for a brief period, almost seems that one of the grids is developing a space charge that affects its bias. When it loses its volume , it also seems the hum increases- but audio is still full when the volume wiper is touched.
I am going to get back to it today hopefully…..

[radiorich]

Hello RGK20m3 ,
Far,as the fluke goes what is meant by the internal resistance means is that when you have it connected to a circuit for test the circuit sees a the meter asa 10 megohm resistor .
Now, if you were using the RCA WV-98 Voltohmyst, the internal resistance is 11 megohm resistor .

Hope this answers your question ?

Sincerely Richard

[MrFixr55]

the issue of probing anywhere on B+ causing the volume to restore to normal is strange indeed, as a 10 MOhm load on a power supply that has to deliver about 40mA at 250 VDC is not going to affect much in the line of disturbance.

Try the "Divide and Conquer" method of injecting Audio from an MP3, CD, etc. in at the volume pot. Not at the wiper but at the input to the pot. If the radio already has a "phono' jack, use that. Otherwise, consider putting one in. If this issue happens on injected audio, hen the issue is with the Audio and Power Supply. These circuits are simpler to troubleshoot than the RF / Detector / AVC section.

One possible issue is the speaker. It is not uncommon for the flexible leads from the voice coil to the speaker terminals. Although it is more likely to cause static especially on low notes or loud low voice talk, cut in / cut out is not uncommon. The "spike" from touching B+ may cause this. If this is the issue, the fix is simple. the wire is like a braid and flexes most at the terminals. add solder, let it go a little further down the braid. This may solve the issue.

[RGK20m3]

Speaker is not the issue, as I have substituted another known good pm speaker, using a separate output transformer and a 1200 resistor in place of the field coil, although that was a good thought. B+ doesn’t change when the volume reduction occurs, nor does the screen supply voltage to the 6A8 or 6K7. Nothing changes in the audio voltages.
I am suspecting that there might be some voltage leakage between terminals either on a tube socket or terminal strip; there is a terminal strip adjacent to the 2nd detector that has B+ along with the grid bias reference resistors….