with the error on the PCB, my amp had no compensation and was NOT unstable 🙂 (59° phase margin without cap c3 c4)
even with those 2SA in my opinion you can go down to 33p
these 33p is really to make it super stable on difficult charge (> 80 ° phase marging)
even with those 2SA in my opinion you can go down to 33p
these 33p is really to make it super stable on difficult charge (> 80 ° phase marging)
Last edited:
I connected the differential output from the AK4399. (Chinese solid-state player DV-10).
Now I listen to music without a transformer.
Now I listen to music without a transformer.
I immediately put 33 pF. Stable current and deviation at the output of the amplifier + - 20 mV.
even with 100p the stability remains good
if there is no problem in measure, stay like this
i want to cry when i see people trying to run goldmund clones that oscillates without charge
they waste time and money
they waste time and money
Last edited:
2 kHz (test disk).But I will later measure using a generator through a transformer.
I had to add a passive filter at the output of the DAC. The amplifier sounds great.
When turning the amplifier on and off, there is no shock in the speakers. Everything ingenious is simple.
Seb, what is the optimal output current of the stage?
At a frequency of 50 kHz, I removed the “step” and the current turned out to be equal to 100 mA.
When turning the amplifier on and off, there is no shock in the speakers. Everything ingenious is simple.
Seb, what is the optimal output current of the stage?
At a frequency of 50 kHz, I removed the “step” and the current turned out to be equal to 100 mA.
Last edited:
Mine is 300ma, but Michele use 500 or more
but this chosen according to our heatsink
Sorry don't understand "removed the “step” "
but this chosen according to our heatsink
Sorry don't understand "removed the “step” "
Last edited:
i think in a circlotron topology, that step must not be present.
my two cents is that derive from a not ideal match between the two mosfets.
increasing current i think you can fix that. but best do a close match on the device first...
my two cents is that derive from a not ideal match between the two mosfets.
increasing current i think you can fix that. but best do a close match on the device first...