Simple amplifiers have always been a challenge. Here is the one, single supply operated (24V), 8W amplifier with only 4 transistors, singleton input and bootstrap, with coupling capacitors. The schematics follows:
The amp uses Vbe multiplier (Q10) to stabilize idle current with temperature. Idle current is set to about 60mA.
Frequency response is flat
Stability is very good and feedback factor is low and flat
Spectrum has mostly low harmonic order of distortion components
and THD+N vs. output power
If you asked me if I built it - I did, however lower power 2W version supplied from 15V PSU.
The amp uses Vbe multiplier (Q10) to stabilize idle current with temperature. Idle current is set to about 60mA.
Frequency response is flat
Stability is very good and feedback factor is low and flat
Spectrum has mostly low harmonic order of distortion components
and THD+N vs. output power
If you asked me if I built it - I did, however lower power 2W version supplied from 15V PSU.
Nice, @PMA
It is a bit like NPN Only old school.
See my thread: https://www.diyaudio.com/community/threads/npn-only-amplifiers-with-only-npn-transistors.405544/
It is a bit like NPN Only old school.
See my thread: https://www.diyaudio.com/community/threads/npn-only-amplifiers-with-only-npn-transistors.405544/
You should use a higher value for R10, 22k to 47k, and the bootstrap cap should be connected to Q3 emitter
rather than to the output, also this kind of topology is inferior to the usual complementary input IPS + VAS
arrangement.
I would add that given the 24V supply it would be better ro reduce the VAS current to 4mA and use a pair
of BDW93/94 darlingtons, these are cheaper than the MJE and will provide much lower distorsion., here
your VAS drain 23mA and dissipate 0.26W, the BC547B will be 52°C over ambiant, usualy a BC140/141
was used for such high current VAS + EF1 OS.
rather than to the output, also this kind of topology is inferior to the usual complementary input IPS + VAS
arrangement.
I would add that given the 24V supply it would be better ro reduce the VAS current to 4mA and use a pair
of BDW93/94 darlingtons, these are cheaper than the MJE and will provide much lower distorsion., here
your VAS drain 23mA and dissipate 0.26W, the BC547B will be 52°C over ambiant, usualy a BC140/141
was used for such high current VAS + EF1 OS.
One might be surprised that it really works, right? I found the old sample and below is a measurement at 8.35W/4.7ohm. THD 0.21%. Heatsink is too small for infinite operation at this power level 😀. But it is really idiotensicher.
And quite good 60+70kHz IMD 😉. Try 60+70kHz with high NFB amplifier 😉. And spectrum is clean up to 10MHz (meas. limit).
And quite good 60+70kHz IMD 😉. Try 60+70kHz with high NFB amplifier 😉. And spectrum is clean up to 10MHz (meas. limit).
Last edited:
Black-Devil
I can't stop thinking about the so-called basic circuits of applied electronics. And I also assume that the quasi-complementary output stage with only one line type, npn or pnp, is not as bad as one would like to assume. Why not even without a complementary output pair?
Anyway,
I'm pinning a German circuit from the 1980s.
HBt.
Yeah, we can make it a fun thread, no problem. In fact the tested circuit in post #5 is a bit different from post #1, it uses another trick with similar effect as the bootstrap, which probably leads to lower distortion. Note that the speaker is within the FB loop. (The test PSU voltage was 24V, not 18V as in the image)
But, I do not like quasi-complementary topology! 😀😀😀
I built this when I was 14, in 1969, with GC521K/GC511K (Tesla CZ) germanium complementary pair. But it was not idiotensicher as now with MJL15030/031, those GC germanium parts burned as a paper, even at 12V.
But, I do not like quasi-complementary topology! 😀😀😀
I built this when I was 14, in 1969, with GC521K/GC511K (Tesla CZ) germanium complementary pair. But it was not idiotensicher as now with MJL15030/031, those GC germanium parts burned as a paper, even at 12V.
Last edited: