Small Push Pull tube amp build, advice sought.

deltafred · Post by **deltafred** » 11 Dec 2020, 18:36

I used EL84s. As they are out of my spares box and untested I tried 4 different ones with no change in output.

Thanks for the correction on parallel/series, it was bothering me that parallel didn't seem right.

deltafred · Post by **deltafred** » 11 Dec 2020, 23:44

Manfred, I got a bit of time on it with the scope.

It's the PI that is clipping not the output stage, it is biased too high so is clipping to the HT. I'm using the one from the 18Watt Lite IIb. When I get some more free time I need to get that sorted out.

Manfred · Post by **Manfred** » 12 Dec 2020, 16:31

Here is the corrected dimensioning.
To reach the Ra of 4.8 kOhm you now have to connect an 8 Ohm speaker to the 15 Ohm tap or a 4 Ohm speaker to the 7.5 Ohm tap.
The maximum output power that can be reached is about 3.4 Watts.

: El84_PP_TriodeMod_PP_ClassA.jpg (103.23 KiB) Viewed 1786 times

Manfred · Post by **Manfred** » 13 Dec 2020, 04:21

Hi Deltafred
you had written:

The MT was just warm to the touch after about 30 minutes at full power (5W), the OT still felt cold.

With the new circuit, the current consumption is higher with the new quiescent current setting.
The transformer must supply about 40 watts for the power output stage including heater voltage supply and should be
should have a power value of 50VA for this alone. In addition, there is the power consumption of the preamplifier and the phase inverter.
What is the power rating of your transformer?
If your transformer has a smaller power value,
then you could make the power of the output stage less.
It is a disadvantage of the class A power amplifier that the current consumption in idle mode is high.

Manfred · Post by **Manfred** » 13 Dec 2020, 16:59

I made a mistake in the output power calculation.
I took (Ua~_peak * Ia~_peak) / sqrt(2) this is wrong, corredt is Ua~peak/sqrt(2) * Ia~_peak) / sqrt(2) = (Ua~peak* Ia~_peak)/2.
So the output power is about 1.7 watts per tube so 3.4 watts total.
I have corrected the diagram above accordingly.

Manfred · Post by **Manfred** » 14 Dec 2020, 00:54

Here the Class B version provides more output power without quiescent current.
If the transformer has a 12.6 VAC winding for the heating filament, you can generate the bias voltage by means of a double-wave rectification.

: El84_PP_TriodeMod_PP_ClassB.jpg (99.55 KiB) Viewed 1759 times

deltafred · Post by **deltafred** » 14 Dec 2020, 16:11

Thanks for all this design work Manfred, it is really appreciated.

I can generate a -22v bias voltage easily with a separate small transformer, I have lots to choose from and am sure I can find one suitable.

I haven't got back to it yet as other jobs keep getting in the way.

Manfred · Post by **Manfred** » 14 Dec 2020, 22:44

In the circuit diagram I had entered the bias voltage value with -22VDC instead of -20VDC I have corrected the circuit diagram above.
You can also generate the bias voltage from the high voltage winding.

deltafred · Post by **deltafred** » 16 Feb 2024, 18:05

No progress has been made on the amp, I see it every time I go in my store cupboard and it reminds me that one day I will have to get back to it.

I've just had a PM for info on Partridge transformers so am posting what I have in case anyone else needs it.

Partridge info.pdf: (3.27 MiB) Downloaded 27 times