Power tubes red plating

[navdave]

At idle all 4 tubes read fine but as soon as I start to turn the volume up the current starts to spike and they start to red plate.
I thought maybe the tubes were going bad so I swapped em out but I got the same results with new tubes.
Now I'm stumped any suggestions?

[jeff]

 When you say turn the volume up do you mean just by turning the volume knob up with no signal or as you play through it?

 Do you have a schematic?

You may be biased too hot. What's your voltage and how much current are you drawing at idle?

[navdave]

Yeah nothings plugged in if I turn up the master volume which it right before the phase inverter they start to red plate.
I got -52 on all the grids at idle that's running 30mA through these 6L6's so they are running cool with 493VDC on the plates.

[navdave]

Well I went in and re-checked my bias voltage with the amp on standby. On one side of the quad I got -52 on the grids the other two tubes read -23?
I guess one or both of these two tubes had a short at the grid where it was drawing the negative bias down.
I don't know what happened gonna keep a close eye on it.

[sluckey]

Pull the output tubes and fix the bias circuit. Pin 5 of all 4 tubes should read the same negative voltage unless you have separate bias pots. If that's the case, be sure you can set all grids to the same negative voltage. The grid voltage should be independant of tube operation unless you have a bad tube. Don't put the tubes back in until the bias circuit is fixed.

Post a schematic or link to a schematic. Maybe some hi-rez pics.

[navdave]

One or two of those tubes went bad on me if all the tubes where out it read -52 when I put all of em back in with thats when the negative voltage on those two tubes when way low.

Pull the output tubes and fix the bias circuit. Pin 5 of all 4 tubes should read the same negative voltage unless you have separate bias pots. If that's the case, be sure you can set all grids to the same negative voltage. The grid voltage should be independant of tube operation unless you have a bad tube. Don't put the tubes back in until the bias circuit is fixed.

Post a schematic or link to a schematic. Maybe some hi-rez pics.

[phsyconoodler]

Determine if the redplating follows the tube first.That will tell you if the tubes are bad.A bad tube will not alter the bias voltage.It will just draw more current.
  Is this an older amp?If so,just replace the bias supply components and try again.

[navdave]

I built this amp in 2007 so it is newer I've inspected the bias capacitors and they don't appear to be bulging or leaking been letting it burn with the new tubes in and all looks good so far.
I use a modified Sunn style bias scheme this allows me to run el34's or 6550's.

Determine if the redplating follows the tube first.That will tell you if the tubes are bad.A bad tube will not alter the bias voltage.It will just draw more current.
  Is this an older amp?If so,just replace the bias supply components and try again.

[HotBluePlates]

One or two of those tubes went bad on me if all the tubes where out it read -52 when I put all of em back in with thats when the negative voltage on those two tubes when way low.

Gassy tubes could result in grid current, which then drops the bias at the grid pin. You would *ass*ume a leaky coupling cap would booger bias voltage whether the tubes were in the socket or not. Oscillation could cause an (otherwise) undriven tube to redplate. A tube which was passing way too much current could start to exhibit gassy behavior (excess current strikes the control grid and liberates gas ions), and redplate. Too-big grid reference resistors could make "marginally gassy" tubes act up, where they would operate without issue in another circuit.

Those are the primary causes I can think of which would result in the situation you're describing.

[navdave]

Thank's for the info HotBlues I did swap the grid reference from 220k to 100k. I also swapped out the coupling caps before I put the new toobs in. Hopefully this cured my problem.

[HotBluePlates]

I didn't look to see what's common in 6550 amps, but the old Tung Sol data sheet specifices 50k as the max grid resistance for fixed-bias operation. If you have a pair of tubes fed from a single resistor, that really is supposed to be lowered to 25k.

But that really saps the gain from the phase inverter, because it's a heavy load (ideally, the plate load resistor at the PI would be half or one-quarter that of the grid reference resistor).

And you probably already know that if you halve the grid reference resistor, you'll need to double your coupling cap value to maintain the same frequency response.

Hopefully 100k works for you. If it doesn't, then we might have rediscovered why the big 6550 amps used 12AU7's (or 12BH7's) as phase inverters; they have the low internal plate resistance to drive lower loads like 25-50k.

[Willabe]

Hopefully 100k works for you. If it doesn't, then we might have rediscovered why the big 6550 amps used 12AU7's (or 12BH7's) as phase inverters; they have the low internal plate resistance to drive lower loads like 25-50k.

Ahhh!  That is very interesting.I bet your right.


               Brad