SVT II oscillation/distortion

[drewl]

Getting a high frequency oscillation/distortion on the bottom half of the some wave when you start driving the amp.

Probing the 12AU7's causes some nasty oscillations, but it present on them and the 12AX7 driving them.
Preamp is fine and power supplies are good with little ripple.

Swapped out the tubes, they are fine.

Think a coupling cap is flaky?


With output tubes removed signal on grids is perfectly clean and I can probe the 12AU7's and it doesn't go nuts oscillating.

[drewl]

Hmmm, as soon as I put any output tube into any socket on the one side I get the oscillation.

The other side is clean.
The negative feedback is throwing the distortion back to the phase inverter.

So, something funny with one side of output trans or circuit.

[Enzo]

Two 12AU7s and aone 12AX7. Make SURE they are in the proper sockets. I often found the tubes swapped in order, and sometimes I found three 12AX7s.

[drewl]

Yes, yes, quite.

Removed negative feedback and no more nasty nasty.

Nice clean amplification from both sides.

I don't think OT wires were switched, but this was a basket case another "tech" couldn't fix.
Wires aren't long enough to have mixed them up.

Let me search for pix of board wiring.

[drewl]

Replaced feedback resistor, amp still working great.
Jumper 120pf feedback cap in and -UH OH!

Nasty back.
Try a different 120pf....still good.
Wait, 120pf is actually a 1nf.
Would that make a difference?

Let's try a different 1nf.....oscillations!

It does make a difference.

[Chuck H]

Replaced feedback resistor, amp still working great.
Jumper 120pf feedback cap in and -UH OH!

Nasty back.
Try a different 120pf....still good.
Wait, 120pf is actually a 1nf.
Would that make a difference?

Let's try a different 1nf.....oscillations!

It does make a difference.

Wow! It probably wasn't as quick as the posts make it seem, but that's some good and fast troubleshooting. Finding the source and omitting any associated components, then introducing them one at a time. And almost no side tracking! I'm humbled

[drewl]

Don't be.
So, all proud of myself for being master of all, after buttoning up said amp it now goes into a crazy square wave as soon as signal gets to the output tubes.

After burning the F- out of my thumb removing output tubes, signal is fine on the output tube grid sockets.

That's it for tonight.
Try again tomorrow.

I sure hope it isn't the OT.

[Chuck H]

Don't be.
So, all proud of myself for being master of all, after buttoning up said amp it now goes into a crazy square wave as soon as signal gets to the output tubes.

After burning the F- out of my thumb removing output tubes, signal is fine on the output tube grid sockets.

That's it for tonight.
Try again tomorrow.

I sure hope it isn't the OT.

Need a "don't like" button

Still... You've narrowed it to an area of the circuit. It worked for a time. So something you were dinking with made a difference. It's possible it was coincidental timing though. I know I repeat this somewhat, but make sure the power tube sockets have good traces and contacts and the pin sockets are clean just in case this problem is an intermittent fault, since it seems now it might be.

[bob p]

I don't think OT wires were switched, but this was a basket case another "tech" couldn't fix.
Wires aren't long enough to have mixed them up.

If you're getting instant howling as soon as you plug in a tube, why not check the results with the OT primary wires reversed?

I wouldn't put too much faith in lead length ... you never know, someone else may have clipped one of the wires and now the amp looks like the wrong wiring topology is the only way that will work.

[bob p]

Replaced feedback resistor, amp still working great.
Jumper 120pf feedback cap in and -UH OH!

Wait, 120pf is actually a 1nf.

120 pF is really 0.12 nF, not 1.2 nf. That's an easy mistake to make.

Are you talking about the 120pF cap that's in parallel with the 47k feedback resistor? That' cap is commonly known as the feedback "speed" cap.

One of the evils that comes with the tube amp/output transformer relationship is that a phase shift occurs at high frequencies and open loop gain falls off at high frequencies. The phase shift in HF content can cause HF closed loop gain to increase with NFB. The textbook approach to correct for this is by adding capacitance across the feedback resistor to "regain" some of the phase shift. This sort of "speed up" capacitor isn't all that common in guitar amps because flat extended HF response doesn't matter in a guitar amp. It matters even less in a bass amp. but as you know, the SVT has a lot of hifi engineering features that aren't present in the garden variety guitar amp, flat extended HF response in the feedback loop being one of them.

Here's what I would do:

first I would try reversing the OT primary leads, just because. then I would try driving the amp with a 10k square wave signal into a dummy load, and look for ringing on the rising edge of the square wave. i would try using different values of the speed cap until the ringing is gone.

one might ask why we have need for a speed cap in the SVT circuit, when MI amps are bandpass limited and don't normally require them. i think that the reason is that the original SVT design involved some parasitic capacitances and 120pF was chosen through trial and error. when the SVT II came along maybe that value was just carried forward.

[drewl]

Point is moot now, looks like the OT died.

[g1]

For next time, when looking on the scope, it should be visible whether connecting the NFB loop increases or decreases the output signal.
In a situation where the OT leads are 2 short, you can also flip the wires coming from the PI coupling caps.

[Jazz P Bass]

Ouch!

[drewl]

It definitely looks like the OT was messed with.
It has an older style trans with not only orange, red and brown wires, but also blue, green and yellow secondary wires that were not connected but still in a square plug where you can see the orange, red and brown wires were attached to the plug.

So it may have been a left over trans from the 70's or it was replaced at some point.
Also on the 1970 schematic all six secondary wires were used.
This was wired with only three.

sorry I can't attatch the schems from phone.
I'll add them Monday at work for posterity.

The friend got it for free so he may not mind spending for a new ot.

[Enzo]

Look at part numbers on the transformer to identify it.