Nice! Digging all the output options of late. MOS, SIT, JFET, BJT, GER… what’s next??
GAN, we need GAN outputs in amp that plays up to 10kHz
ZM, does the P103 adjusts Iq?yup, as I said , bad influence of Cyclotron emanation , all the way from Bubba Bob's Davis ...... 'Talian names popping as mushrooms after the rain
anyway, what's better to put on pizza than mushrooms .......
So, last one in series of autoformer gain amps (you can call them all - M2 on Mushrooms) , this one differing on outputs being classic BJT complementary pair, so sorta blasphemy in Papaland ......... he sinned heavily eons ago ....... while my Sin is recent........
Anyway, here's funny diagram, showing nothing else than bunch of multilegged fuses and some nice iron
hm, yes, my prototype equipped with first genuine ones I snatched from drawer - 2SA1491/2SC3855 ....... but pretty much any "for Audio amp" complementary pair should work well ........ 2SA1943/2SC5200 , TTA/TTC (newest Toshiba incarnation of 2SA/2SC)
rails +/-23V, Iq 1A5
2nd is clearly neg phase, not seen from my gigglywiggly, but I stumbled on nice routine in LTSpice showing just that behavior and didn't yet found reason to doubt it
edit on 20.01.2022. - C109 and C110 are not fitted, neither here nor in schm in post #6, so ignore them; these were "precautionary parts"...... which I initially forgot they're needing investigation (with or without); measurements shown were made with them in place, but after trying without, I can happily toss them out; as result, Freq. linearity (logically) being better without them ..... for instance , 8R load - everything stays in 1db envelope
What about three pairs of outputs? I got plenty C5200/A1943...
P103 offset, P104 Iq
name of the game is Square Law; in this power scale format - can't see how much you can make it more Berserk than it is already with "just" one pair of outputs, so reasoning that way - really no possible gains in multiplying number of outputs , just increasing complexity...... which inevitably goes to getting less detailed sound
generally - doable for multiple pairs, but that is feasible only if going for higher power amp - different front end, higher rails, and only then is possible to think how best to arrange biasing - as one circuit for all of them or separate for each pair; most likely latter approach, as pretty much only way of avoiding use of emiter resistors
so, see, even my own Greediness is having boundaries, sometimes
in any case, to quote one clever guy and his thinking :
http://edsaudiopages.blogspot.com/2021/07/how-much-amplifier-power-do-you-need.html
name of the game is Square Law; in this power scale format - can't see how much you can make it more Berserk than it is already with "just" one pair of outputs, so reasoning that way - really no possible gains in multiplying number of outputs , just increasing complexity...... which inevitably goes to getting less detailed sound
generally - doable for multiple pairs, but that is feasible only if going for higher power amp - different front end, higher rails, and only then is possible to think how best to arrange biasing - as one circuit for all of them or separate for each pair; most likely latter approach, as pretty much only way of avoiding use of emiter resistors
so, see, even my own Greediness is having boundaries, sometimes
in any case, to quote one clever guy and his thinking :
http://edsaudiopages.blogspot.com/2021/07/how-much-amplifier-power-do-you-need.html
I have tried and tried and tried, but no matter how much I turn these pots my IQ test score remains the same, with a huuuuge offset…
Not about power, i just wanted to use more partsP103 offset, P104 Iq
name of the game is Square Law; in this power scale format - can't see how much you can make it more Berserk than it is already with "just" one pair of outputs, so reasoning that way - really no possible gains in multiplying number of outputs , just increasing complexity...... which inevitably goes to getting less detailed sound
generally - doable for multiple pairs, but that is feasible only if going for higher power amp - different front end, higher rails, and only then is possible to think how best to arrange biasing - as one circuit for all of them or separate for each pair; most likely latter approach, as pretty much only way of avoiding use of emiter resistors
so, see, even my own Greediness is having boundaries, sometimes
in any case, to quote one clever guy and his thinking :
http://edsaudiopages.blogspot.com/2021/07/how-much-amplifier-power-do-you-need.html
This amp needs more parts.
No ballast=pure square law=more linear. And more power/current and huge class A pocket. Mighty will tell me if I turned the Iq pot the right or wrong way
damnvikinggreenhorn replied almost right - source resistors are sorta ballast but ...... calling them that way would be misleading
yup, Square Law
Papa started it with PL XA25, though basic arrangement is well known from before
example - Hitachi lateral audio mosfets - smaller amps from Yore (first ones, published in application sheets) made with them - didn't had source resistors at all, easily solvable due to friendly TempCo of laterals - 100mA is sweet spot and you can bias them with piece of string and two safety pins
verticals - no way
pucks - nononoway
visited today my OSSoE (Official Second Set of Ears)
he said - after smaaaaaaalll Rakija and biiiiiiiiiiig amount of Tea, and some listening while chatting, and 2 tunes carefully listened - "Well, you didn't embarrassed your self, even with this one"
he said - after smaaaaaaalll Rakija and biiiiiiiiiiig amount of Tea, and some listening while chatting, and 2 tunes carefully listened - "Well, you didn't embarrassed your self, even with this one"
Attachments
Yeah, but where can i get a Rakija in US?visited today my OSSoE (Official Second Set of Ears)
he said - after smaaaaaaalll Rakija and biiiiiiiiiiig amount of Tea, and some listening while chatting, and 2 tunes carefully listened - "Well, you didn't embarrassed your self, even with this one"