I am trying to de-bug an AWA B15 Radiola (6BE6-6BA6-6AV6-6AQ5).
I get distortion on strong signals, so I suspect the AVC.
The AVC line (just below the if coil to the 6BA6 grid) swings nicely from +0.6 to -6.0 V, but I don't really seem to get any AVC action...

Any thoughts? I don't have a circuit diagram for the B15, but it seems pretty easy...

Derek.

Did you change the paper caps & check resistors?

The fact that the grid swings positive rings alarm bells. Normally valves do not like positive grids.

The most common cause of too much positive is leaking caps connected from B+ to the AVC. Likely a Mica.

Marc

Distortion on strong signals or distortion at high volume?
It is important to understand the difference between the two or you could be chasing the fault in the wrong section of the set.

The B15 uses a circuit common to many AWA sets from the late 50's onwards. Different cabinet, same circuit, essentially sums up AWA radio designs of the era.

The figures of AVC voltage you quote do sound OK, given the B15 has only a ferrite loopstick to pickup the signal, and it's not a set with copious amounts of gain anyway. A slight amount of positive bias (under a volt) is nothing to worry about with no signal. It can come from grid emission of the IF amplifier, or with TV sets, a high value resistor between AGC and B+.
However, this is only acceptable where the IF amplifier has cathode bias. A positive grid is fine provided the cathode is also positive by the minimum required amount.
In the TV world, a 10M or 22M resistor is connected from B+ to the AGC line to offset the AGC generated from internal noise; in otherwords, a method of achieving AGC delay.

A quick way to see if the AVC/AGC is inadequate is to feed a variable power supply into the AVC line (positive to earth of course) and force the gain to be reduced - it should cut out altogether by no more than -30V.

If you confirm the AVC is indeed inadequate, then the causes will be leaky paper bypass caps on the AVC line, leakage between primary and secondary of the last IF transformer (very rare), or leaky mica cap between 6BA6 plate and the AVC diode inside the 6AV6.
Mica caps are now old enough to be a common failure, especially where they have DC present across them.

Assuming the AVC is OK, then the audio section will need to be looked at. Distortion comes from a high 6AV6 plate resistor or leaky grid coupler for the 6AQ5, both of which are common faults. An annoying scratching noise will be the result of the 6AV6 plate bypass failing, or the speaker transformer about to go open circuit. Distortion can also come from a faulty speaker, but this is easily tested.

Having said all that, all the paper capacitors should be replaced as a matter of course, and resistors checked before the fault finding process begins.

Thanks for the thoughts. Just a couple of clarifications; The voltages I quoted (+0.6 to minus 6.0) are measured from the chassis, and the cathode has a resistor and is more positive than 0.6. So that looks OK.
Yes, the distortion is on strong signal not high volume, I can increase the signal by connecting an external antenna to the ferrite rod (the B15 has external Ant & Gnd connections) this causes a weak undistorted signal to become a loud distorted signal.
I will continue to look at all the caps around the AVC line tonight.

One other possibility has just come to mind and that is the 6BA6 is very weak. In this situation it would clip the IF signal prior to detection, if it was very strong. A weak IF amplifier will also result in less AVC being generated. A CRO on the 6BA6 plate would confirm this theory, or of course you can simply try another 6BA6.

Unless the circuit says otherwise the control grid voltage is with respect to cathode.

Circuit would be of benefit. I did have a set recently where only failed paper caps had been replaced. All of its' issues were due to replacing a cap with a bigger one to cover up the problem caused by the 6AV6 plate resistor, going very high.

I normally find that getting rid of old electrolytics, paper caps & checking the resistors as I go; fixes over 90% of the issues, including the faults you never knew were there.

Australian Mica's are normally reliable, however, some do fail and I had to replace every one in an Astor recently.

I find an insulation tester very good for non polarised caps & Micas, but you need one that can produce a range of voltages.

Always lift one end of the cap to do this. Don't waste time on papers & check any resistor & Mica that has one end lifted while changing caps. Saves re-work.

Marc

Delete double post

Hi guys ...a very interesting Thread ...thanks!

One clarification .......distortion caused by an inadequate AVC/AGC??

Does this mean that AVC is designed into circuits so that it prevents overdriving of RF/IF stages when strong signals are present?? Until now, I had only thought of AVC as 'balancing out' volumes between strong and weak signals.

I will try for simplified:

AVC / AGC Voltage is developed normally on the diode plate or plates of the detector.

The strenght of the signal, governs the amount of voltage developed.

This voltage is applied to the controlled grid of the Pentagrid and often to the grid of the 1st RF amp. as well.

A strong signal will naturally develop a greater voltage which when applied to the controlled grids, reduces the gain of the valve.

The valves used will either have a super control or variable mu grid.

If there is a failure of a component in the AVC / AGC line this can cause the timing constant to go out of it's design parameters and that can cause distortion.

Alternatively: If due to high resistors, leaking caps or faults at the detector , insuficent control voltage is developed, the amplifiers can overload & start signal clipping for one.

Marc

I have one of these radios, the AWA Radiola B15 with a number of issues that went away after replacing all the caps marked "UCC" packaged in plastic barrels, and also replaced a "Simplex MS" cap in the 6AV6 plate circuit (was making noise). One of the UCCs was visibly leaking wax, so I figured that they were wax paper caps, and replaced all of them. They had date codes like 3462, 1463 which look to mean made in the 34th week of 1962 and 14th week of 1963. Replace all these caps, you'll save time and fuss.

A leaky AVC filter bypass cap can drag "down" the AVC voltage (in this case "down" meaning making it less negative and thus less effective) , and cause distortion on strong radio stations. Also check the resistor, as megohm range resistors can change resistance value, usually going higher.

I am not at all surprised with the UCC caps failing. The SM caps are fairly reliable, but as I pointed out above, not infalable. This is why I always check them especially if they have a lot of B+ on them.

Often the signal in delayed AGC / AVC is taken from the plate of the !st RF after the pentagrid (super het) or mixer (autodyne). That one always feeds B+ where it's not wanted when it fails.

A lot of UCC's are paper but I rate them with the same contempt as Micamold and those are normally replace on sighting.

It is interesting that a lot of the open construction laminated Mica wafer types are more reliable and the 1936 Radio on the bench has them.

None have failed & some I tested at 500V

Some radio's do have voltage dividers on the AGC of several Meg & I have seen plenty of them fail. Do check the grid leak on the 6AV6. Often that is around 10M and I have also seen several of them around 15+ meg.

Marc

Finally I can report back, now that I am confident I have fixed the fault.
Turns out the problem was resistors rather than capacitors or valves. Not before I sourced a full set and tried them one by one though
With a good antenna and a strong local station I was getting -12v AVC, but this would drift down towards -20v and after a good heat soak occasionally flick down to -70v at which point the audio signal was clipping on one side and of course very distorted. I couln't see any way a faulty capacitor could do this, it looked like the converter or IF gain was jumping way up, like it would if AVC was failing to control them (or if gain went up for any other reason). So I took a punt just replaced all the resistors, as being a feeedback loop I can't see how you can pinpoint the problem without a comparison set. Some were looking a bit distressed (browned) anyway.
Result is a good stable 12-14v AVC on strong signals, no distortion, job done. Thanks to everyone for the discussion.