Sunn 200S clone in pentode mode screen voltage too high

[g1]

Here's that thread, it covers both the red-plating and PI issues:
http://music-electronics-forum.com/t37473/

[Wes]

Thanks for posting those. Killer info there for sure!

[Wes]

Just because you are using the 8ohm tap, and it is labelled as 8 ohm, does not necessarily mean the power tubes are seeing the optimum primary Z, it could still be too low an impedance load.

So to find the actual pri Z of the OT would you use the Vin/Vout squared multiplied by the tap that you're using?

Doing that gives me 5k on all taps of the Hammond 1650R which is what it says it is.

I have the Triode Electronics OT which says it's 4.3K. Measuring it using the same method gives me 4.16k on the 16 and 4 ohm, but 4.76 on the 8 ohm tap.

I need to understand better all of this plate to plate, primary impedance business. I'm going to research this. Any quick suggestions where to look or book recommendations?



Thank you guys.

[g1]

You know what the primary Z of the OT is (5K), what you need to know is what primary Z the amp needs (plate to plate Z).
This depends on the type of power tubes and their operating conditions.
There are methods for figuring this out, nickb did here, although he did not explain how he figured it out:
http://music-electronics-forum.com/t40692/#post401972
I'd imagine there may even be an on-line calculator where you plug in your values and it tells you the impedance.

[pdf64]

When in pentode mode, are the screen grids being fed from the same HT node as the OT CT?
If so, as you've found, that can maintain too high a screen grid voltage at high signal levels, resulting in excessive plate current.
I suggest adding another HT node with a 470 ohm to 1k high wattage >10W dropper, and a smoothing cap.
Then feed the screen grids via their individual screen grid resistors from that node.
High value individual screen grid resistors can result in excessive local feedback in the tube, that causes compression of high signal levels.

[Wes]

I had them on a separate node. OT CT was at the first node and the screens were after the choke.

I'll try some stuff... thank you.

[g1]

Here's the online calculator from nick's site : Interactive Valve Data Sheets

[Enzo]

Not feeling well, I haven't logged in for a while.

But if you are concerned that the screen voltages are too high, why not reduce the voltage at the screen node? The screen resistors limit current, but if you increase the dropping resistor in the B+ string, wouldn't that work? You may have to reduce another resistor to keep preamp voltages intact.

[soundmasterg]

Another thing you could try is to run the amp UL like Sunn did and see how it behaves. It should be able to run just fine with the Triode/Sunn UL OT, and then if it behaves in UL, try to adjust from there to get it to behave in pentode. If you have already taxed the power tubes a bit with red plating and what not, you may want to try a new set of tubes as if you red plate tubes long enough they won't ever go back to being stable.

Greg

[Wes]

Ok-

I have the Triode/Sunn UL OT installed and set up UL. New Kt88s.

I am concerned, now, with the way that the PI is operating. The voltages there are all jacked up
and the cathode side is driving the piss out of one of the power tubes. This is the side that is red
plating when driven hard. I haven't forced it to red plate, but watching current when pushed a bit-
that side jumps way high real quick; I think it will end up red plating if allowed.

I did swap the feed from the PI and that heavy drive follows the cathode side of the PI.

The voltages recommended by the schematic for the 6AN8 and what I'm ending up with are

Pentode

schem..........my reading

A 75................126
S 35................29.2
K .75................. .61

Triode

schem..........my reading

A 340...........272.5
G 75..............126
K 90.............134.5

The resistors to the tube are all correct except a 2M on the screen of the pentode where 1.5 is called for.

Power supply B node calls for 425 and I have 416

Power supply C node calls for 320 and I have 320

I did change the 15k resistor between the A and B nodes to 22k.

Is it possible that the 6AN8 is just worn out or bad?



Another thing- The Hammond 278CX PT has an additional primary lead. It is not on the spec sheet.
In this build it is being used for the bias supply. I haven't seen something like this before. Is this safe?!

p-t278.pdf

[Wes]

Wowee zow wow! What an effect changing the screen resistor on the pentode half of the 6AN8 has...

I dropped that resistor from 2M down to 1.5M and the voltages went in the right direction.

Pentode

..........schem..........measured

A...........75................94

S...........35................35

K......... .75.............. .75

Triode

A..........340..............318

G...........75................94

K...........90...............106

Now the PI output is closer to even. I'm going to double check my bias on the 88's and go from there.....

[Wes]

The power tubes are idling at about 12W each. I just drove the hell out of the amp and it did not red plate.

Man, it's mean....

[g1]

Another thing- The Hammond 278CX PT has an additional primary lead. It is not on the spec sheet.
In this build it is being used for the bias supply. I haven't seen something like this before. Is this safe?!

The newer hammonds have a 120V or 125V tap for the primary side.
The extra wire should not be attached to anything at all.
Without seeing how it's wired up, it certainly does not sound safe to run anything off the primary side AC.

Edit: here is the hammond update note: http://www.hammondmfg.com/263.htm#263newpri

[Wes]

I figured that's what it was. I will have that taken care of. I thought that was kinda crazy....

[soundmasterg]

For a pentode, the screen is all important as it has much more effect on what happens than the plate does. You can pre-figure stuff out with the equations but in the end there is always empirical testing to see how a circuit will work with a pentode and what values you need. If you haven't read Merlin's preamp book yet, you should. His website is also very good. http://www.valvewizard.co.uk/

Greg