V376 motor-boating at lower gain settings

[Ike Zimbel]

Hi All,
I have a TAB V376 on the bench that's turning into what I call a "bench bomb", that is; a piece that starts taking up so much time it becomes my problem instead of my client's problem. This one came in with two others for a re-cap and check-over. So far, I've corrected at least three faults: Intermittent- broken trace on input transformer, no gain trim - bad pot, no 140 Hz LPF - wire broken off at switch (although the wire was so stiff from age that it was just hovering over the switch contact like it was still soldered ). I also had to replace at least one transistor. So, now it works perfectly, but only at the higher gain (really lower attenuation) settings, 40 to 70 dB. Below that, 0 to 34 dB, it's motor-boating. These are the first 376's I've worked on, so if you've had experience with these, I welcome your input. One other note, on these cards, all of the active components are on sub-boards, and I've been able to rule those out by trying them in the other two modules (both 100% now), so it would seem like the issue is on the main board...
Thanks!

 

[JohnRoberts]

I am not familiar with that unit but in general, lower gain results in less stability margin. Is it possible that it is breaking into oscillation at low gain?

JR

 

[Ike Zimbel]

I am not familiar with that unit but in general, lower gain results in less stability margin. Is it possible that it is breaking into oscillation at low gain?

JR

From the looks of these fine ocean sunfish swimming across my scope, I would say "Yes, totally!".

 

[Ike Zimbel]

I am not familiar with that unit but in general, lower gain results in less stability margin. Is it possible that it is breaking into oscillation at low gain?

JR

I haven't dealt with this phenomenon before. Can you suggest any approaches to dealing with it? I have already tried increasing power supply decoupling on the module, but that didn't help.

 

[JohnRoberts]

Most NF amplifiers use dominant pole compensation, effectively a one pole LPF in the open loop gain path. The intent is to reduce total gain open loop and NF combined to below unity, by the time the delay equals 180' of phase rotation making negative feedback positive. Higher nominal closed loop gain, actually attenuates the NF signal further increasing stability.

A brute force solution is to increase this compensation capacitor.... BUT ASSuming this circuit is a working design without oscillation we can search for other suspects. With amplifiers PCB layout can matter. I have seen ground current from speaker returns corrupting the clean NF network grounds. (Sorry I didn't open up the schematic again so I only have general advice).

A cheat to make the amplifier think it is operating at higher closed loop gain than it is, is an RC from - input to ground,,, but this should not be necessary for a proven stable design.

JR

 

[Ike Zimbel]

Thanks John. I've had to back burner this one for a few days, I'll report back when I get back into it.

 

[jokeramik]

Hello,

which transistor you replaced? What for a type you used?

In which amplifierstage the motorboating ist detectable at first?

Best regards!

jokeramik

 

[Ike Zimbel]

Hi, I haven't had a chance to narrow down where it's happening, but my notes say " Repl. BC-149 on H2a (lower card) with BC184L."
I used the BC184L because I had some in stock and it measured similarly to a good BC-149. The pin-out was a bit different but I just had to put the leads in the right places.
On another front, the amplifier cards from this one work properly in another module...
Thanks!
iz

 

[Ike Zimbel]

Update! I finally took another look at this today and spotted the problem almost right away. It was another case of a wire breaking off its contact in the rotary switch, and then staying in exactly the same spot, so it looked like it was still soldered. 2nd one of those on this unit, although the other one was on the HPF switch.