Sansui BA-5000 Oscillation

[megajocke]

I'd use short wires to jury rig your fresh capacitors in parallell with the original ones. If the problem goes away then you know the original capacitors are bad and need to be replaced.

[EchoWars]

I used alligator clips & jumper wires to temporarily connect the 35,000µf 80V caps in parallel to the original caps. Didn't seem to make any difference.

I also removed C5 and since I didn't have any 150pf caps, I stuck in a 220pf cap (10x the original value). The frequency of the oscillation seems to have gone up (to about 700K) and the amplitude down (from about 10V P-P to about 6V P-P), but it is still there.

Man...

[gbyleveldt]

Eish.

Last suggestion, with the 220p cap remaining in the Vas position, try removing C6 (the lag comp 5pf in series with the 47k resistor) and/or C301 (also 5pf I think).

Let us know the result? But it is pretty tough to try and diagnose these things remotely.

[jacco vermeulen]

Both channels oscillating.
=>Two channels -Two main powersupplies-One high voltage regulator board

Originally posted by EchoWars
but recapping the small +/-65V regulated supply (and installing new transistors & zener diodes) cured that.

[BFNY]

Here's an idea to also consider - maybe off the wall, but this appears like it's a tough one.
This unit has an output transformer separated by a fuse from the output stage. Try removing the fuse and measuring there (i.e. output transformer and wiring out of the circuit)
Look for a misguided "tidy up" effort where someone may have bundled the amp input wires together with the internal speaker outputs. Doing this would allow capacitor coupling the input and output together via wiring. Taking out the fuse would remove the effect.
You may also want to check the Classic Sansui yahoo group, I have recall of others reporting oscillation problems on serviced Sansui amps sold on ebay. If it was not the BA5000, it was a model very similar.

[EchoWars]

Originally posted by gbyleveldt
Eish.

Last suggestion, with the 220p cap remaining in the Vas position, try removing C6 (the lag comp 5pf in series with the 47k resistor) and/or C301 (also 5pf I think).

Let us know the result? But it is pretty tough to try and diagnose these things remotely.

Yes, I know. But I appreciate you all throwing out ideas. I have removed the amp from the bench to get some other things done, but I'll put it back to try this and perhaps a few other things.

Originally posted by jacco vermeulen
Both channels oscillating.
=>Two channels -Two main powersupplies-One high voltage regulator board

OK...it isn't like I haven't thought of the fact that both driver boards are running off of the same regulated supply, but without a load the supply looks completely clean. The input to the regulated supply is about +/-78VDC with about 1VP-P sawtooth ripple. The output is about +/-62V, and ripple-free, with only trace noise and certainly no oscillation when the driver boards are out of the amp. With either driver board reinstalled and oscillating, the oscillation can be traced from the output back to this regulated supply.

Anyway, I assume you're trying to feed me a hint here. I'll take the hint, but I'm at a loss to know what to do with it. FWIW, TR02 and TR03 are replaced with On-Semi MJE15032, and TR01 & TR05 with Zetex ZTX694B and TR04 & TR06 replaced with Zetex ZTX795A. I did try to replace C04 and C05 with a .001µf ceramic, which made no difference.

Originally posted by BFNY
Here's an idea to also consider - maybe off the wall, but this appears like it's a tough one.
This unit has an output transformer separated by a fuse from the output stage. Try removing the fuse and measuring there (i.e. output transformer and wiring out of the circuit)
Look for a misguided "tidy up" effort where someone may have bundled the amp input wires together with the internal speaker outputs. Doing this would allow capacitor coupling the input and output together via wiring. Taking out the fuse would remove the effect.
You may also want to check the Classic Sansui yahoo group, I have recall of others reporting oscillation problems on serviced Sansui amps sold on ebay. If it was not the BA5000, it was a model very similar.

I have not popped the fuse out to see what happens, but I'll try it and see what happens.

Oscillations on many older Sansui's can be traced back to a type of ceramic cap used on some high-speed units (AU-919, BA-F1, and others with super slew-rates). These caps are absent in the BA-5000, which is a fairly slow amplifier in general...

[paul79]

Are you sure the transformer was the cause for the noise and not the soft start relay chattering? Might take a look at the contacts on the soft start relay. Had a similar problem with a spec-4 pioneer.

[EchoWars]

I know what a chattering relay sounds like (as a former robotics tech, I had to deal with hundreds of them at a time). It was from the transformer (sounded like an uneven load on the windings).

[djoffe]

Both channels are affected...it had to be something common to both channels...earlier in the thread, you said,

"I figured there were damaged semiconductor junctions with a few of the outputs, so all 16 of them, and all four drivers, were replaced with new MJ21193 & MJ21194 devices (all emitter resistors and base resistors were checked for proper values). That made no difference."

Well...emitter follower configured transistors, like the output devices, and perhaps TR701 and TR703 should be included in this list...can oscillate all by themselves if they see too low a driving point resistance...looking into the emitters, it looks like an inductor and a negative resistance...add a bit of capacitance, and you have an oscillator.

The new transistors may have enough high freq gain to let this happen.

We could test the theory...increase R705, 707, 709, 711 to 22 ohms...if the oscillations stop, or change...then maybe we're on to something, and we can back them down a bit...

We might also have to add a bit of series R in the bases of TR701 and TR703, perhaps 22 Ohms again.

(P.S. I also assume you've checked C714 (0.1 uF) and R735 (4.7 Ohms) they could also have some effect on stability, by reflecting a bit lower z at high frequencies across the auto-xfmr.

I know it may be painful to try this, but so many loop gain things have been ruled out, that it sounds like a local thing.

[EchoWars]

You have no idea how difficult it would be to change those base resistors...the entire output assembly is point-to-point wired directly to the TO-3 transistor sockets. On top of that, the whole output heatsink has to be removed to do it, and that's a chore in and of itself...

Here's what I've done in the last couple of hours...Jacco's post made me look back at the regulated supply. Now, I've used Zetex transistors for TR01 and TR04 before, but not for TR05 and TR06. So, I replaced those last two with 2SC1845 and 2SA992 and reinstalled the stock value for C04 and C05. No change.

So I decided to try BFNY's idea of pulling the output fuse (the 15A cartridge fuse between the amp output and the transformer). The oscillation goes away when this is done. I can even reinstall the fuse and it only returns about 1 time in 10. Oh, and installing a new Bussman fuse makes no difference...

I think I need to take a look at C714 & R736. I have a nice metal poly cap for replacement, but I'm not sure about the resistor. I gotta go dig through my parts bin and see what I got.