Ett tidigare svar från konstruktören ang matchning:
Fick svar från Federico (konstruktören) direkt. Nån som tror det spelar någon roll?
T: These are however matched as I can see and I wonder how important that is?
F: Well, I admit that I've never known how important is; I've always matched them, and many literature report that it is important to have power transistors matched BY A POINT OF VIEW OF STRESS AND THERMAL BEHAVIOUR.
But nothing is reported about the sound.
Matching them it's a difficult task; you should do it using a "typical" current (say: 1A), and I made a circuit that did this in impulsive mode (10mS ON, 990mS OFF), in order not to be masked by the thermal drift.
More simply (you have to match only few transistors, not thousands like me) you can make a DC circuit that measure the beta with the power transistors on a heathsink.
Or .... you can simply put new transistors and live happy
Ciao
Federico
p.s. I remember that N and N were similar, so well P and P; the problem was to match N with P
Ang. offset är den mycket låg, multimeter visar 0.00 på lägsta skalan (mV). Förutom den kanal som nu är trasig som uppvisar 25 mV - men det kanske nu är löst (hittat trasig diod i speaker protection).