there are 192ma bias on amplifier, I think more bias can make more distortion on amplifier.
My simulation shows that more bias gives less distortion.
I have used 200mA. THD 0.00060%
At 250mA the distortion is only half from that. THD 0.00027%
This is tested with 1 Watt into 8 Ohm.
I guess with Class A bias this amplifier has really low distortion.
But Class A with +/-40 Volt is not to recommend.
My testing is with +/-20 Volt. For this Class A is possible without to much heating.
Should not be very hot. Make sure there is no oscillation.
It happened to me with the PeeCeeBee which has a similar topology.
Did you equip the output with a Zobel network?
it looks like a production\fabrication problem. also; the pads have very little soldering area.
It is drawed on eagle cadsoft and it works good. I dont have any solder problem. Signal and low current size could be little but high current and onboard caps have a big solder pad… if you use good flux you can solder top and bottom. But i will care and revise second time printed pcb thanks
LSK189, KSC3503, KSA1381, TLVH431 can be bought from https:www.mouser.com
This is a SPICE test with VSSA as Class A amp.
The voltage supply is set to +/-20 Volt.
The bias is 1.25 Ampere.
The gain is reduced to 13.2. Set by R6/R2 and R7/R3.
Result at 1 Watt level using same circuit for both cases.
Bias .... THD
0.200A 0.000591% Class AB
1.250A 0.000085% Class A
This is the circuit I used:
Correction for the best: R8 pot should be 2 kOhm.
This is a SPICE test with VSSA as Class A amp.
The voltage supply is set to +/-20 Volt.
The bias is 1.25 Ampere.
The gain is reduced to 13.2. Set by R6/R2 and R7/R3.
Result at 1 Watt level using same circuit for both cases.
Bias .... THD
0.200A 0.000591% Class AB
1.250A 0.000085% Class A
This is the circuit I used:
Correction for the best: R8 pot should be 2 kOhm.
Last edited: