balancing the bias of quads on SF twin

[pontiacpete]

sorry to bring this up, I know this is probably basic, and I've actually done this procedure a couple of times at a shop I worked at with guidance but have forgotten. I searched the forum and couldn't find anything that explains it either., anyway, this twin reverb has the balance bias pot on the back(can't find a schematic of the exact model I have here, but it has a master volume) and it's a tapped pot i.e. it has the added lug on the opposite side of the 3 usual ones. The wiper has a 33K to ground and one pair of 6l6s has a 68k and the other pair a 47K. The back lug goes under the chassis to the cap box to the + end of a pair paralleled 70uf/100v caps. The other 2 outer lugs are connected to the 47k and 68k.
So right now with a matched quad of 6l6s I'm reading 16mA on one side and the other something like 26mA and when I move the balance pot the imbalance continues in this fashion.
I was thinking of rewiring the bias set up to a BF ab763 and that would probably do it. But I know there is a way of balancing the four tubes while keeping the basic wiring the same. Is there someone here that could help me with this?
Much thanks,
pete

[pontiacpete]

mistaken it's 135 twin

sorry, it's the 135 ultra linear twin, here's the schematic

[txstrat]

Have you checked (or cleaned) the pot?

[pdf64]

Try measuring the voltages at the 6L6 grids, at various pot settings.
It would be best to get this resolved before considering any mods.
The inability to get them balanced may be caused by a bad resistor, eg the 470 screen grids, 1k5 grid stoppers, or 47k / 68k bias feed resistors.
In any case, a bias balance control is often useful, it may be better to fit an overall bias control, in addition. That could be done by making the 33k resistor to ground into a 22k resistor in series with a 22k pot, wired as a variable resistor.

[pontiacpete]

Thanks for the response. I checked/cleaned the bias pot, changed the screen resistors and the grid resistors are good.
This amp came to me with a hole through the glass of one of the 6l6s. I put in a new set and did a few things and tried starting it up on a limiter bulb, but there was a short. Then I found that the terminal strip with the 30k/20w resistor and the 2 100K resistors on it just below the pilot light next to the PT had broke apart where the 100ks meet. Well to make a very long story short, someone had wired something wrong before me. So that has to get squared away first, because it now runs without those 3 resistors connected and in there probably lies the problem with the bias of the tubes being out of balance.
I'll be back when I figure this out.

[Amp Kat]

Are you sure the tubes are perfectly matched ? Just because they tell you they are doesn't mean they are. You may have one set matched to a certain ma # and the other 2 matched at a higher #. Did you check the bias voltage on pin 5 of each tube. If the bias balance is at O and each tube has the same bias voltage the only difference can be the tube itself.

[pontiacpete]

thanks, again, for the help and reminding me that matched quads might not necessarily matched. I shouldn't trust that they all come balanced because they say so.

I sorted out the wiring to the of the 30k/20w resistor according to the schematic. Then I turned the bias pot and found that as I turned it up, the grids on both sides did in fact balance out at -56v, but the current measured at 19ma. At one end of the pot the voltages were at -52 and -57 and in the middle they equal out at -56v. As I turned the up more the balance went out again. So I, then, returned it to where it was balanced at -56v on both sides and add a 68k resistor in parallel with the 33k resistor on top of the bias pot. This allowed me to bring the bias up to 29ma, where I like to set it, with the plates at 470v, while maintaining the balance of the quad. pete