question for SVT mavens

[Prairie Dawg]

I've got an SVT here that will throw itself out of gear every once in a while, sometimes it'll happen on startup and sometimes after a few hours. I have replaced the screen resistors, three of which were defective.

Since I had this amp in the shop last, it was part of the reason I built a tube matcher that could deliver some serious voltage. It's a variation on the Jack Price tube matcher. I've had it around here long enough to know that it's relatively reliable.

At a vP of 500 and a vB of -60v I'm getting the following cathode current readings: 24.3 ma, 30.2 ma, 30.8 ma, 30.4 ma, 32.4 ma and 40.1 ma.

Is that enough spread to trip the fault indicator and throw the amp off line?

This is a SVT Classic made by St. Louis Music before they went overseas.

Thanks

[beans-on-toast]

I would have changed all the screen resistors so that a matched set was in place. If three were blown, maybe excessive stress was put on the other three. Not a big deal.

Here is a pointer to a partial set of schematics for your amp. There are some notes concerning changing six diodes and a cap in the bias circuit. It might help with your problem. The addition of the cap shown on the schematic will make the amp less likely to generate a line fault.

People that sell tube sets typically claim to match plate currents to within 5% or better. Matching transconductance is also a good idea. The 24mA and 40mA tubes are 20% and 33% off. A closer match would be better.

If you are able to set the bias using the LED's (two green lights on) with the tube set that you have, the variation in cathode current is within the tolerance of the circuit. If you can't get two green lights, the schematic says bias them so that both LED's are red. In such a case, I would set the bias manually by taking measurements inside the amp with a dummy load installed.

[Prairie Dawg]

I did change all the screen resistors and I've just opened it up and they look fine. I'm about to open the rest of the chassis and check everything out. I did have one a while ago that did the same thing and what cured it was redoing all the solder joints in the fault detector circuit-for lack of a better name.

Thanks for the tip on the diodes-and I will check all of that out before I tell the guy to buy another set of tubes. These are Winged C's, supposed to have been matched, came from Tube Depot, but if they were matched a few months ago they aren't now.

What about switching over to KT88s? Is there any advantage to be gained?

Realistically what needs to happen is for Ampeg to derate these amps by lowering the plate voltage to reasonable levels.

Film at 11.

[beans-on-toast]

Originally Posted by Prairie Dawg

What about switching over to KT88s? Is there any advantage to be gained?

Although it makes sense to switch to a KT88 because of the higher plate and screen voltage specs, 6550's have a long standing track record when installed in an SVT.

The coke bottle shape of the glass envelope of the KT88 puts the tubes in even closer proximity which can affect heat dissipation. People run them in these amps without any problems. Sound wise, a good KT88 can match the tone of the best 6550A.

The Winged C tubes are very good. Unfortunately, even when you buy a match set, you never really know what you are getting. You have to deal with a reputable vendor and put your trust in them.

If you are having power faults, I'm thinking that adding the feedback cap on the op-amp will help.

[okcrum]

Those circuit changes should do the trick for spurious faults. What do the tubes idle at in the amp? That's at 160V higher plate voltage.

[Prairie Dawg]

I haven't checked the tubes for idle current in the amp yet. My guess is the split will be pretty much similar.