High voltages in final test on CS4 board

[dehringer]

With the 6922 tube out,

Between C1 and C3: 73.6k
Between C6 and C8: 73.7k

[Paul Birkeland]

How about C1 to C2, C2 and C3, C6 to C7, C7 to C8?

[dehringer]

C1-C2: 121.8k
C2-C3: 102k
C6-C7: 121.8k
C7-C8: 102k

[Paul Birkeland]

Can you run the amp with just the 12AU7 (no 6922, and no EF86s), then let me know what the OA and OB voltages are on the front C4S board? OKA and OKB voltages would also be of interest.

[dehringer]

Sure thing.

With only the 12AU7:
OA: 214.8 V
OB: 213.6 V
OKA: 0.2 mV
OKB 0.1 mV

[Paul Birkeland]

I would attempt to procure a different 6922 or 6DJ8 and try that in the C socket.  When your amp is running, it's acting like there's a short between plate and cathode on that side of the 6922, which could be from debris in the socket, but that would appear as a DC resistance difference between sides on the socket and would be persistent with the tube out of the socket.

[dehringer]

Ok, sounds like a plan.  I’ll do that and when it arrives report back.  Thanks Paul!

[dehringer]

I tried a different 6922 in the C socket and unfortunately did not resolve the high voltages.  Updated voltages:

IA: 217
IB: 230
OA: 145
OB: 160
OC: 140
OD: 98
OkA: 139
OkB: 100
OkC: 4.7
OkD: 1.6

Are there any other things I should try?

[Paul Birkeland]

Perhaps PJ has some ideas about why this might be happening. 

[Paul Joppa]

I'll take a closer look, but it may take a few days - PB revised the layout and the PC board to make the Eros II much easier to assemble. It's a great improvement, but I'm not yet intimately familiar with the new board.

[Paul Birkeland]

IA: 217
6DJ8 plate A: 145
EF86 plate C: 140
EF86 screen/DJ8 cathode A: 139
EF86 cathode C: 4.7

This is the voltage set on the offending side.

[Paul Joppa]

4.7v on OkC tells us that the EF86 is cut off - the 2N2222 bias servo is not conducting.

102k ohms between C2 and C3 tells us that the 174-ohm resistor is connected at least.

But the 174-ohm resistor may be shorted, or the 2N2222 base may not be connected to that resistor. Check the voltage across that resistor - it should be around 0.6v. If it's closer to zero, the resistor is shorted. If it's closer to 0.8, the transistor may not be connected at the base.

Or the 2N2222 collector may not be connected to the EF86 cathode (A3). The collector should also be connected to one end of the 0.1uF Z5U on the PC board, and to the + end of the cathode capacitor, which is (I think) the 10000uF/6v cap (shown on the circuit diagram as 2200uF/4v as was originally used (I'm not certain of that change). Check those capacitor voltages to identify where the break may be.

I notice that the original measurement of OkC was 11.8v, which could have damaged the 10000uF cap. There's probably not enough current available to damage it severely, but you might want to replace it after the real problem is corrected, just to be sure.

That's my first shot at identifying the problem, anyhow!