The +5.5V supply is for the CCS. And yes, it is PLUS 5.5V.
Primaries correctly connected. Same dcr on both phases.
I swapped all tubes in all possible ways. Replaced with new ones etc etc. No significant difference.
I measured the output tube grids, they both are around 0.001V.
I haven't taken out the coupling caps for further measurement, but they seem to do their job.
I had some moments of red plating earlier, so I was thinking the OPT might need demagnetizing. But like I said I put in another OPT and the issue is still there.
Perhaps you might extrapolate. I quiver with antici-pation. (Birthday greets to Tim Curry).
All good fortune,
Chris
Yes. I had a little bit of oscillation with a different driver operating point / anode load. (30K and 300K fb R) at the bottoms of the sine. Gain was very high. So I went back to 10K / 100K.
Output cathodes are both at 21,5V.
If everything is fine with both amps and the signal is equal on the primary side of the transformers then the secondaries are not.
What kind of current source are you using for the input tubes?
Schematic of the CCS circuit, please.
You have checked everything else; try checking the CCS.
Why does the current source need +5V.
Most current sources start with a Negative voltage at the bottom, then the top of the current source goes to the cathodes.
That way, with the cathodes at perhaps +3V for example, the current source has -5V to +3V across it (8V), plenty of voltage to meet the burden voltage of most current sources.
Check that both supplies are equal, either +5V and +5V; or -5V and -5V.
Then swap the current sources.
Schematic of the CCS circuit, please.
You have checked everything else; try checking the CCS.
Why does the current source need +5V.
Most current sources start with a Negative voltage at the bottom, then the top of the current source goes to the cathodes.
That way, with the cathodes at perhaps +3V for example, the current source has -5V to +3V across it (8V), plenty of voltage to meet the burden voltage of most current sources.
Check that both supplies are equal, either +5V and +5V; or -5V and -5V.
Then swap the current sources.
You have labelled the voltages on the rectified supply for the CCS as +5.5V and 0V, but it seems like that is not what they'd actually be, because of those 100ohm resistors that are tying the midpoint of the 6.3Vac heater supply to ground. Wouldn't the two recified supply outputs be at something like +2.25V and -2.25V? Maybe this is OK, but without seeing the schematic for the CCS it is difficult to tell.
I have used this circuit many times myself and it works just fine.
https://www.diyaudio.com/community/threads/zotl-and-ferrites.73135/post-5665425
https://www.diyaudio.com/community/threads/zotl-and-ferrites.73135/post-5665425
it needs a negative supply, cause transistor may be underbiased, with some tubes it may work but in most cases the tube will only droping a couple of volts, then if the tubes are unmatched the available voltage for the ccs will be less, the red led bias at the transistor is 1.8V so the cathode available voltage will be less cause it would need to be higher than the base to conduct, when driving signal, it will become negative and abrupt cut or plain oscilate
It's a NPN and the bias is fine. In case of overload it will be bad but so is without.
An issue with CCSs is called voltage overhead (among other names), meaning how little voltage can it have across itself and still be a CCS. The voltage at the "top" of your CCS is the DC common cathode voltage, at idle (no signal) plus and minus 1/2 of the input signal voltage to the amp, with signal. First thing is to measure that DCV. Then figure the signal voltage added to that. Let's pretend signal voltage at the input jack is 1.0 VRMS, so signal voltage at common cathodes is half of that: 0.707 VPeak positive and negative.
The CCS needs to operate properly down to DC common cathode voltage minus 0.707 Volts. Does yours do that? Let's assume that yours works well enough down to a voltage (at it's CCS output) of the voltage across a red LED plus a few tens of a Volt, with zero margin for error.
If one were to measure the DC common cathode voltage and the DCV across the red LED, one would resolve this issue. Or converting to a negative DCV supply, which would give plenty of headroom.
Don't know if this will solve your issue, but needs to be explored. And painless enough to do.
All good fortune,
Chris
The CCS needs to operate properly down to DC common cathode voltage minus 0.707 Volts. Does yours do that? Let's assume that yours works well enough down to a voltage (at it's CCS output) of the voltage across a red LED plus a few tens of a Volt, with zero margin for error.
If one were to measure the DC common cathode voltage and the DCV across the red LED, one would resolve this issue. Or converting to a negative DCV supply, which would give plenty of headroom.
Don't know if this will solve your issue, but needs to be explored. And painless enough to do.
All good fortune,
Chris
So the idea works fine just the implementation of the second amp differs. Now it can be still the CCS on that one so try another BC547.