Yes you are right. Distortion of 0.03 percent is due to my output stage. The output stage in the original circuit provides 0.12% distortion and a fairly impressive range of high harmonics. So far, I'm in no hurry to share the output stage circuit in the public domain. But I can say that it is parallel on two pairs of output transitors, with modified current sources and positive power feedback to correct the minimum distortion at the first watt.
great, it will be interesting to know your opinion!
a 24 kΩ resistor must be of high quality and with a power of at least 3 W
I completely agree with Hattori, agree to raise the quiescent current of the output transistors. Domenico do your boards have 2 grounds? I do not think these will be better.
my PCBs are purchased on Alipress or Aliexpress, in short, we understand each other.
Then, in a frenzy, I developed it with 4 output pairs, changed the polarizations, of course, resistors on the emitters, VCC stabilization for the input BJTs with LM 317 instead of the zener, stabilized power supply for the driver and final stage, diodes fast recovery on lines and exits, another final than 108, but always a mess to check the offset without the supercap , this scheme
Attachments
< I will study on the "passive servo" whose starting point I posted in #1431. >
...........................
A passive servo, not a guilty one! No power supply from 78/79xx most hated by most of the audiophile. Not any op-amp. No... nothing bad! a serie of two very long time constant with the right cut and Q, with the best quality components, as good to let pass only very subsounds an DC. To be studied on. I have not the solution yet. But the idea should be a good one.
Why
?
Last edited:
< A passive servo, not a guilty one! >
Okay
we make it faster:
who has tried to compare servo / cap and, if possible without cap and without servo?
If you need something to check the offset because the scheme just does not allow you to control it except to resort to caps and servos, you can always try, TO UNDERSTAND, the damage that the servo and the cap causes, checking the output offset with a tester . Of course, these are only experiments to be done carefully, but they serve to understand.
Once this is done, one wonders if it is worth going crazy with some schemes rather than looking beyond and looking for a scheme that does not have compensations, limitations and the need for various caps and servants.
Okay
we make it faster:
who has tried to compare servo / cap and, if possible without cap and without servo?
If you need something to check the offset because the scheme just does not allow you to control it except to resort to caps and servos, you can always try, TO UNDERSTAND, the damage that the servo and the cap causes, checking the output offset with a tester . Of course, these are only experiments to be done carefully, but they serve to understand.
Once this is done, one wonders if it is worth going crazy with some schemes rather than looking beyond and looking for a scheme that does not have compensations, limitations and the need for various caps and servants.
I'm just watching from a distance, but isn't the DC offset a thermal issue as the junction temperature fluctuates so does the offset, right? I'm think that part of the solution is to thermally couple the semiconductors with a smaller slab of copper and keep them warm/hot deliberately so that the junction temperature does not fluctuate with load.
A huge slab of heat-sink just makes the thermal difference bigger between the semiconductors. It's need so the temperature doesn't run away, but nothing is said about running hot at 85C (185F). Any thoughts?
A huge slab of heat-sink just makes the thermal difference bigger between the semiconductors. It's need so the temperature doesn't run away, but nothing is said about running hot at 85C (185F). Any thoughts?
all the problems related to the BJTs of the final stage and various differences that in a scheme like the 108 with a single pair are very important can be solved with the automatic closed-circuit bias that obviates the problems of the BJTs of a closed open loop, bjt PNP and NPN , chips like 1166
I found this russian build thread ????????? Dartzeel NHB-108 (????????? ?????, ?????? ????????? ? ????) [2] - ??????????? iXBT.com
Looks like they have been looking at thermally coupling things as well.
Looks like they have been looking at thermally coupling things as well.
Dear piotrkundu, you did not read this forum well, I have long shared this photo and others, how to reduce the constant drift at the output without a super-duper capacitor. My amplifier is built just on such transistors, I will not say anything bad about them, the letter G means lead-free technology manufacturing.
Attachments
