VHT Special 12/20RT low preamp voltages

[audiopete]

Schematic is on page 10 http://www.vhtamp.com/images/Manuals...t%20manual.pdf

I got this amp in on warranty service. It was brand new and the purchaser returned it to the store because the trem wasn't working well. The store discussed with VHT and they decided to send it to me to fix, rather than ship back to them. I have already discussed this problem with VHT designer Terry Buddingh, and gone through his checklist after I did all my own, with no success. VHT wrote off the amp to the store and I bought it from the store.

At first, I went through all the quick basic checks and discovered that the preamp voltages of B+ 3,4 and 5 were way down, maybe 60% average (but varying at each node). B+, B+1 and B+2 all are at the specified voltages. I cannot find or figure out what is going on with this amp. I have spent hours going through and checking tubes, components, wiring, sockets, solders, etc etc. I don't get why B+3, 4, 5 are down but not B+, B+1, B+2.

So far: (6V6s installed)

- determined V3 has no effect so its been removed for all tests
- swapped all 12AX7s for some good testers I have and no change
- pulled all 12AX7s and the voltages come up to specified
-- pulled V1 with all other tubes in - minimal change in B+5
- pulled V2 with all other tubes in - Voltages rise to about 85% of specified
- pulled V4 with all other tubes in - minimal change in B+5
- pulled V5 with all other tubes in - voltages rise back up to specified

- measured every cathode resistor and all measure within spec
- lifted every cathode bypass cap with no change
- checked all coupling caps and everything looks to be good - with lower voltages showing at schematic test points due to the low B+ I assume
- measured all plate resistors and all measure good
- Measured all cathode voltages and they appear to correspond with the low B+s
- measured current of preamp board by breaking the ground path and measured approx 4.2mA - seems fine with the low voltage condition

- lifted each power supply filter cap for B+3-5 -no change.
- lifted each bypass film cap in PSU - no change
- changed R69, 70, 71 - no change

- changed zener diode ZD1 - no change

I don't think the Watts circuit with the MOSFET could be affecting the amp, and it appears to be functioning properly, but perhaps I am mistaken there?

It is a most bizarre problem made worse because it is a basic tube amp like any other. If anyone has any ideas or suggestions, I would appreciate hearing them.

Cheers.

[g1]

What voltage do you measure on each side of R69 ?

[audiopete]

I have 333V/209V - please note i swapped R69 to 27K (didn't have 33K in stock)

I also took more measurements:

B+3 - 209V
B+4 - 174V
b+5 - 160V

V1B plate - 106V
V1A plate - 159V
V2B plate - 75V
V2A plate - 109V
V4A plate - 110V
V4B plate - 173V
V5A plate - 167V
V5B plate - fluctuating (trem)

[Mick Bailey]

I've had a few of these amps where the Mosfet has blown - they go short but the amp works fine with it shorted - just the power level is stuck full on.

I'm wondering if the trem circuit is pulling the voltage down. I'd lift R60 and re-check the voltages. Especially as the original owner reported the trem wasn't working properly.

[audiopete]

Okay thanks Mick, i'll try that in the morning. Right now C40 is lifted on the VR8A pot side, not that it is causing the issue, but I was checking the cap for leakage.

I get a lot of Traynors that use a MOSFET to switch B+ on and off with the standby switch. The MOSFET blows and there is either full B+, no B+ or somewhere in between.

[g1]

Don't get too hung up on those voltages shown on the drawing, there are obviously some mistakes there. They show 0V drop across R71, even if B5 is only supplying 1mA, there would be at least 10V drop across R71 .
Across R69, approx. 1mA would account for the difference between their voltage drop across R69 and what you have measured.
Is there any fault aside from the trem ?

[J M Fahey]

Same here, I find more "normal" the voltages you measured rather than the schematic ones, which to boot have errors.
1) B+4 and B+5 are *both* labelled as TP18 .
Obviously a drawing error and of course voltage indicated will be the same.
2) drop across R69 is 333V-209V=124V which means current through it is 124V/27Kr=4.6 mA .
Given that we have 8 (eight !!!) triodes after it that averages 4.6/8=0.57 mA each.
A VERY reasonable load.

If anything, the design problem is being very optimistic.
Donīt think itīs the root of your problem, but if you wish, simply lower R69 to, say, 22K and I guess youīll be much closer to schematic voltages.

As Enzo often says, Ohmīs Law still rules.

As of tremolo problems, check that you have nice and strong LF signal at V5b plate and follow it along the path until it reaches the net R37-38

I suspect VR8A which reduces tremolo signal when the amp is "Power Padded" down so at least for testing Iīd short across it, then you should have strong tremolo signal where itīs needed.

[Mick Bailey]

If just pulling V5 with the other tubes in place restores the voltage up to spec, it may indicate that the current is being hogged by V5 or it's associated circuitry and it isn't pulling an average current. Is the plate load resistor on V5b the correct value?

[audiopete]

Don't get too hung up on those voltages shown on the drawing, there are obviously some mistakes there. They show 0V drop across R71, even if B5 is only supplying 1mA, there would be at least 10V drop across R71 .
Across R69, approx. 1mA would account for the difference between their voltage drop across R69 and what you have measured.
Is there any fault aside from the trem ?

No fault and the amp plays, but I think it could be louder and stronger and when you dial down the Watts control, the output is very distorted and low.

[audiopete]

If just pulling V5 with the other tubes in place restores the voltage up to spec, it may indicate that the current is being hogged by V5 or it's associated circuitry and it isn't pulling an average current. Is the plate load resistor on V5b the correct value?

The plate resistor measures fine, but I think I will concentrate on the trem circuit a little more today. I was also concentrating on the PI, since V5 also provides the cathodyne splitter. It seems to check out.

[audiopete]

Same here, I find more "normal" the voltages you measured rather than the schematic ones, which to boot have errors.
1) B+4 and B+5 are *both* labelled as TP18 .
Obviously a drawing error and of course voltage indicated will be the same.
2) drop across R69 is 333V-209V=124V which means current through it is 124V/27Kr=4.6 mA .
Given that we have 8 (eight !!!) triodes after it that averages 4.6/8=0.57 mA each.
A VERY reasonable load.

If anything, the design problem is being very optimistic.
Donīt think itīs the root of your problem, but if you wish, simply lower R69 to, say, 22K and I guess youīll be much closer to schematic voltages.

As Enzo often says, Ohmīs Law still rules.

As of tremolo problems, check that you have nice and strong LF signal at V5b plate and follow it along the path until it reaches the net R37-38

I suspect VR8A which reduces tremolo signal when the amp is "Power Padded" down so at least for testing Iīd short across it, then you should have strong tremolo signal where itīs needed.

Yes there are errors in the drawing...just for fun all the triodes are reversed from the drawing so when tracing you have to mentally swap all the pin numbers. This board is a second version so I doubt there is an updated schem. Between the standby switch and R69 is a 1N4007 diode not shown (I changed it to no effect), as well as a 1N4007 in the watts control, but I can't recall off the top of my head this morning exactly where it sits in the circuit. Also not sure about the TP10 voltage at V5A higher than the B+. I'll check when I get back to it later today.

[Mick Bailey]

when you dial down the Watts control, the output is very distorted and low.

That's a characteristic of the control - it causes the output tubes to shift towards increasing non-linearity and hence distorting badly at low levels. There's also no gate stopper on the Mosfet and oscillation will kill it in the end. It needs a 220 Ohm resistor right on the gate pin.

[audiopete]

That's a characteristic of the control - it causes the output tubes to shift towards increasing non-linearity and hence distorting badly at low levels. There's also no gate stopper on the Mosfet and oscillation will kill it in the end. It needs a 220 Ohm resistor right on the gate pin.

OK. Put me in the camp of not liking a MOSFET in my tube amp power section....we'll see if I get it back one day for this problem.

[J M Fahey]

Well, it ain't exactly in the power section but in the power supply.

If anything, what I donīt like is they only pad down screen voltage, which when way too far away from plate voltage makes the tube work in a weird way and to boot, kills some compression because screen voltage becomes sort of regulated.

I much prefer padding down plate voltage and letting screen voltage just follow that, being derived from it through a resistor (or choke, although it becomes meaningless here).
Also a little sag can be added, to behave even closer to a conventional PSU.

[audiopete]

Well, it ain't exactly in the power section but in the power supply.

If anything, what I donīt like is they only pad down screen voltage, which when way too far away from plate voltage makes the tube work in a weird way and to boot, kills some compression because screen voltage becomes sort of regulated.

I much prefer padding down plate voltage and letting screen voltage just follow that, being derived from it through a resistor (or choke, although it becomes meaningless here).
Also a little sag can be added, to behave even closer to a conventional PSU.

Yes sorry that's what I meant - power supply.

The plates are on B+1 and the screens are on B+2 so I don't see what you are saying about padding down the screen voltage. The MOSFET is before both of those nodes.