DSL bias problem

[mhuss]

I had to replace three main boards on a set of JCM2000s due to the 'creeping copper' problem. Two went just fine (didn't even need my soldering iron) but the third one is having an odd problem. The one sides biases up fine (using the little trim pots and 'three pins') but the other keeps reading 0v. Changed tubes, no difference. I pulled it apart again and re-flowed all the solder joints on the little bias bias board (with real solder, not that ROHSs stuff ). Measurements at the jumper pins seemed to make sense (~10k across the pins to the two 22k pots in parallel, turning the pots changes the resistance at pins 2 and 6, zero ohms from pins 3,4,5 to their respective test point pins). I put it back together, still the same. Recall that this is a brand new main board fresh from Milton Keynes.

Probing further, I noticed something odd -- the bias board jumper wires are mirror-imaged (pin 1 goes to pin 7, 2, goes to 6, etc.) Since the connector wiring is symmetrical, it still "works," but the pots need to turn the opposite direction and the left pot controls the right tubes and vice-versa. This caused a bit of confusion when I was testing at pin 5 on the tube sockets.

As many of you know, powered testing is an adventure because of that %$#%$ output jack board. Insulated foam board to the rescue.

More testing:
- The 1 ohm cathode resistors measure 1 ohm, there is continuity between the high side of these resistors and the external test pins and the cold side and ground.
- Turning the trimmers changes the voltage at the 220k grid resistors.
- The voltage at the (L-R swapped) pin 5 changes on one side, but stays fixed at -45v or so on the "0 volt cathode resistor" side.
- Continuity from the 220k Rs to individual 5k1 grid stoppers, and from the stoppers to the socket pin 5s.
- With caps discharged, all pin 5s measure about 300k to ground, about what you would expect.

Any suggestions?

[mhuss]

More testing:
Voltage looks good on plates and screens
OPT measures 15 ohms DCR from either side to CT.
Getting signal on all grids, but output (as you'd expect) looks like class B.
I'm getting an inverted version of the output on the 'bad side' plates, so the output seems to be getting reflected from the 'good side.'
I rechecked that pins 1&8 both measure ~1 ohm to ground.

I'm thinking there's a tear in the space-time continuum.

I'm testing using a single pair of tubes, the bad side acts the same with the single tube plugged into either socket.

To recap - The OT seems OK, I have HV on plate & screen, reasonable bias and signal on the grid, and the cathode and suppressor grids are grounded -- but the tubes I've tried refuse to conduct. (Recall I did swap L&R tubes).

Any and all suggestions welcome...

[The Dude]

............The voltage at the (L-R swapped) pin 5 changes on one side, but stays fixed at -45v or so on the "0 volt cathode resistor" side............

That part makes me wonder if you don't have bad connections or a bad bias pot. So, if I have this right, one bias pot causes a change in tube current, but the other does not? Check the pot itself. Check connections from the bias board all the way on to the amp board.
Also curious: One side changes with the bias pot, the other stays fixed at -45. On the side that works, what voltage gives you reasonable bias current? In other words, is it in the range of the -45 on the bad side or something vastly different?

[mhuss]

Thanks for the reply. Actually, the bias voltage seems to be varying correctly now. The information above (about the bias being 'stuck') is incorrect, and was my mistake due to the 'backwards' jumper problem. (I would fix it, but apparently you only get one edit? I cannot edit the original post now. )

So, the bias seems to be working OK, the voltages and input signals all seem OK, but the tubes I've tried stubbornly refuse to conduct any current whatsoever.

[The Dude]

My only other thought is: OT wired correctly? When you checked DCR to center tap, did you check it from the tube sockets or at the OT wires? I would verify at the sockets to be sure.

[mhuss]

Verified at the sockets. I have two working ones here to compare, so I'm positive the wires are connected correctly, and it still has the original OT that had been working fine before the runaway bias problem.

[g1]

In case the cathode circuit is only allowing tiny current to measure ohms, maybe try jumpering 1&8 directly to ground.

[mhuss]

Solved!

I connected a 25k 25w resistor between the plate and the cathode...and it warmed right up. Tried a third tube in the same position, and now it's working!

Three morals of the story:
- Always use known good tubes to test! (duh)
- So much for simple emission testing, this one tested OK but obviously does not work.
- "When you have eliminated the impossible, whatever remains, however improbable, must be the truth."