• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Output Impedance Driving Headphones

Does somebody have a better suggestion for the OPT - they are difficult to come by. Something like 40H primary (15K CT), 50 ohm secondary. Use used a mains transformer 1.6VA toroid which works OK but started to roll off (without NFB) at 8KHz. The current one is 24H which is a bit low.
 
Added 100R in series with each headphone output and what a transformation - the metallic sound has gone. So thanks. So half decent quality headphones should be driven from 120R or so not 0R. I guess this is what in line with what modern amps do. I did not know this I assumed it was like speakers.
Output impedance may be a red herring; the blame may lie with the amp and not the headphones, since driving a reactive load is always unpleasant for tubes. The metallic sound may simply have been distortion. But putting 100R in series, the tube sees a higher and more consistent load impedance, meaning lower and more consistent distortion.
 
baudouin0,

Your circuit in Post # 19 has a self inverting output stage.
I have done a couple of self inverting output stages.

Perhaps the operating plate resistance, rp, of the ECC88 is high with the cathodes tied together to the current sink (current sink, which makes them a self inverting output stage).

And, the negative feedback is affected by the fact that the headphone impedance is 'in-series' with the feedback circuit, instead of being across the secondary.

The plate resistance, rp, would be lower if the cathodes were connected to AC ground.
You could use individual self bias resistors with individual bypass caps.
Or tie the cathodes directly to individual current sense resistors to ground, and use individual fixed adjustable bias on the grids to make the plate currents to the output transformer equal.
But either of these 2 solutions would require driving the output stage with a phase inverter.

The higher plate resistance, rp, of the self inverting output stage has to drive the 470pF from plate to plate. That may be part of the high frequency roll off.
Or, he 150k driver plate load and driver plate impedance, rp, and the output stages Miller Effect Capacitance at high frequencies is causing the high frequency rolloff.
Or, perhaps, the toroid output transformer leakage reactance from primary to secondary is causing the high frequency rolloff.
(if each of these 3 factors causes -1dB high frequency rolloff, then the 3 factors together causes a -3dB rolloff).

And, the plate impedance, rp of the output tubes has to drive the 24H primary (6H per 1/2 of the primary, but effectively 12H before either tube cuts off). That could cause low frequency roll off.
 
Last edited:
Certainly the source impedance is changing the sound. I know this because I did the same experiment on the low impedance output of a solid state op-amp headphone output and had the same results. The fact seems to be that my headphones are better driven from a current rather than voltage source. The fact they sound different must mean they are not purely resistive. So this is something that can be measured which is what I will do. The 470pF is intended to model the effective primary capacitance of the transformer. You are correct the HF roll-off is a bit of both. The NFB help with this a great deal. The 24H is just not quite enough at the bass end I think. I will try a mains toroid 1VA6 230V CT 0-12 0-12 on secondary to experiment. I think the primary effective cap is nearer 2nF which makes the HF end more difficult to drive.

And I don't need the hazard thing at the top my HT is only 100V, which can give you a nip on a good day.