Ampeg SVT-1000 Power Amp Mod / Fix?

[tekhedd]

I have an SVT-1000 that I purchased new a couple of years ago. It distorts (with a very ugly waveform that looks like there is a bite out of it) well before the output limiter kicks in. I am trying to see if there is a fix for the output stage distortion.

I hear rumors of a mod for the output stage, possibly related to the octo-limiter circuit. (huge problem with Ampeg SVT1000 [Archive] - Music & Musician Forums)

Does anybody have a clue about this? I can't find a repair tech who has a 4 Ohm dummy load rated to handle 1000W continuous, much less a reactive one, so the amp is essentially a boat anchor at the moment.

The amp checks out fine into an 8 ohm dummy load, so two (2) official repair centers have declared it "in spec".

I realize that the power supply may be under-capable, but considering that it's basically a B4R in a vintage case, I find it hard to believe that this is the entire problem.

SLM has been totally helpful in the sense that they're more than glad to pay for me to have it fixed...except that no authorized repair tech has reproduced the problem and they tell me it's "in spec". So, I'm on my own here I guess. Ideas? Anybody?

[Jerrold Tiers]

That unit is a 2 channel amp which is permanently connected mono bridge.....

So, the odds are that there is an issue with the cross-connection for bridge. What happens is that if the "follower" channel does not get a signal, its output stays at zero, but the other channel is pumping current into it through the load.

That makes the protection work, accounting potentially for your "large bite" out of the signal.

Why that (or any OTHER problem) does not show up at the tech?........... the 8 ohm load probably does not draw enough current to cause the problem. The output stage is stout enough that the protection is fairly loose.

As for that other thread..... there seems to have been massive clue-lessness over there.... back in 2005......

[tekhedd]

Fascinating. Where's my voltmeter? (My scope, sadly, seems to be dead.)

[tekhedd]

With no load, I am getting a nice big AC signal on both + and -, and it appears to be differential. So... apparently I'm still reaching the rated voltage out.

I believe the limiting circuit works, as I've watched it limit the output voltage waveform in the shop, and it seems to be fine.

Thanks for the idea... I'm not giving up yet!

[Jerrold Tiers]

There COULD be a protection problem.

The "limiter" (opto-coupler limiter) is NOT FAST ENOUGH to make a notch in the wave except at low frequencies...... and there is just adds some THD, generally. If it is working you will see the amp limit, usually with just a little clipping.

A notch below full power is something other than the limiter, nmost likely.

if you get full output NO LOAD, then everything is working EXCEPT maybe the protection.....

[tekhedd]

Originally Posted by Jerrold Tiers

There COULD be a protection problem.

The "limiter" (opto-coupler limiter) is NOT FAST ENOUGH to make a notch in the wave except at low frequencies...... and there is just adds some THD, generally. If it is working you will see the amp limit, usually with just a little clipping.

A notch below full power is something other than the limiter, nmost likely.

if you get full output NO LOAD, then everything is working EXCEPT maybe the protection.....

As far as I can tell from the schema, the only real protection is the limiter and the turn-on relay... I wonder if it's as simple as one channel overbiased or something.

I'll bet I can check each channel separately, into a 2 Ohm load. This solves several problems:

1) 500W and 2 ohms (I really don't have a 4 ohm setup that can handle 1000W continuous without eventually melting)

2) test power supply - discover whether the problem occurs with half the current draw

3) see whether it's just one amp, or both

After that I suppose it's time to locate a scope and start checking voltages. Hm.

tom

[Jerrold Tiers]

Originally Posted by tekhedd

As far as I can tell from the schema, the only real protection is the limiter and the turn-on relay... I wonder if it's as simple as one channel overbiased or something.

tom

It has a form of V-I protection in addition to limiter and relay..... look again.

There is a transistor for each bank of mosfets, with sense resistors to the mosfets, and a de-sense resistor to ground...... and an R-C circuit as a time delay

If the resistor to ground is loose, or ditto for the delay parts, it can protect wrongly.... usually being over-aggressive.

[tekhedd]

Originally Posted by Jerrold Tiers

It has a form of V-I protection in addition to limiter and relay..... look again.

There is a transistor for each bank of mosfets, with sense resistors to the mosfets, and a de-sense resistor to ground...... and an R-C circuit as a time delay

If the resistor to ground is loose, or ditto for the delay parts, it can protect wrongly.... usually being over-aggressive.

Thanks. I think I get it now. In fact, it looks like protection/current detection *is* most of the circuitry in the output stage.

[tekhedd]

I did the old "just solder everything" fix. Annoying but usually effective...

The clipping has reduced (hmm), but the amp is still next to useless into a 4 ohm load. Low frequency transients from this amp blow past like the winds of change--which would be cool but unfortunately it appears to trigger current limiting at the peaks. I imagine there's something wrong with the time constant--maybe even a wrong component (we're talking about a St Louis era amp after all) or possibly my 4 ohm speaker is really 2.5 at those frequencies--and now I really need a scope to go further. *sigh* Or I could just buy an 8 ohm refrigerator cabinet... or sell it as-is. Option C is calling to me.

Thanks way much for the help.

[edit] For future reference, I've tried the amp with different 4 Ohm rigs and it seems to work. Perhaps the 322 is less than 4 ohms at low frequencies? I'm going to declare it "fixed but doesn't work with my cab"...