Hi Alex,
Thank you for stepping in. No smoke, no fuse blowing (but I am not switching the amp On for too long periods). As far as I can tell, R123 is fine. Except for the random crackling, the sound is fine.
This fault is driving me nuts as I have replaced so many components but it is still there!
Thank you for stepping in. No smoke, no fuse blowing (but I am not switching the amp On for too long periods). As far as I can tell, R123 is fine. Except for the random crackling, the sound is fine.
This fault is driving me nuts as I have replaced so many components but it is still there!
Q109 replaced as well as Q103 & Q104 (BC546) as the crackling was linked with one of these. Thank you, Alex!
All working now and crackle free but I can’t adjust P101 to read 12mV as the lowest I can get is 28mV across R123 or R122 ( I can adjust to 12mV on the left channel). Is this going to be a problem long term? I have already replaced many resistors on the right channel…
Dominique
All working now and crackle free but I can’t adjust P101 to read 12mV as the lowest I can get is 28mV across R123 or R122 ( I can adjust to 12mV on the left channel). Is this going to be a problem long term? I have already replaced many resistors on the right channel…
Dominique
If you can't adjust the emitter current in R123 down to the correct value, that suggests either a fault in the resistors/pot around q9, or the current souce q10/LED 1 etc is pulling too big a current through them.
That might mean Q7 and Q8 are not sharing the current set by LED 2 properly?
What are the LED voltages?
Q8 and Q10 may start running hot again.
I am thinking of either reducing the Q8-Q10 current a bit or reducing the rails. 40W would be enough for me.
I have all the parts to fix mine now (I hope!), just need to clear the bench of 3D printer parts and get my car working.
That might mean Q7 and Q8 are not sharing the current set by LED 2 properly?
What are the LED voltages?
Q8 and Q10 may start running hot again.
I am thinking of either reducing the Q8-Q10 current a bit or reducing the rails. 40W would be enough for me.
I have all the parts to fix mine now (I hope!), just need to clear the bench of 3D printer parts and get my car working.
The LED voltages seem all fine (all showing 2V). The voltage rails for the right channel are higher (39V for the left side and 38V for the right side), all voltages measured at different points right up to the Varipot are higher on the right side, but all resistor values are correct. Any clues why this is and also how to remedy it? Thanks in advance!
I think the rails are common across the L/R channels, the output stages are direct to the smoothing caps, D211 and D210 supply the low current sections?
If your meter is measuring different rails volts, maybe something is oscillating?
It sounds like it's not far from working properly.
You have two major nodes in the circuit, the bases of the Bd139 and BD140 transistors.
The voltages across B-E of the four transdistors in the darlingtons, plus the voltages across the emitter resistors R122, R123 have to add up to the volts across q9 C to E. In turn that must equal the volts across the path P101, R119 R120, 3.2 x
If you're set for minimum idle current, then the trim pot should be max resistance.
So R119+ POT has the Vbe of Q109 across it, and R120, having much the same current should have 1k5/(470+470) times this voltage =1.6 x Vbe
Sounds like something in the darlingtons has a much lower Vbe than q109?
I'd measure all the Vbe's.
If your meter is measuring different rails volts, maybe something is oscillating?
It sounds like it's not far from working properly.
You have two major nodes in the circuit, the bases of the Bd139 and BD140 transistors.
The voltages across B-E of the four transdistors in the darlingtons, plus the voltages across the emitter resistors R122, R123 have to add up to the volts across q9 C to E. In turn that must equal the volts across the path P101, R119 R120, 3.2 x
If you're set for minimum idle current, then the trim pot should be max resistance.
So R119+ POT has the Vbe of Q109 across it, and R120, having much the same current should have 1k5/(470+470) times this voltage =1.6 x Vbe
Sounds like something in the darlingtons has a much lower Vbe than q109?
I'd measure all the Vbe's.
I've made a little progress with mine, but I'm finding that the PCB is heat damaged in several places, some copper tracks are cracked, they seem to be made of very thin foil or the etching process wasn't great. It's not just change a few components, it's check every connection and many of the components. As the input stage emitter resistors are burned out, what else has had a hard time and changed value? I'm actually starting to consider binning the PCB and building two amps from scratch, using the power transistors.
If we were going to make a new PCB, how much of the original design should be kept?
How good was this amp when it was working properly?
This particular one is never going to be a collector's item, I would use it in a 'second room' system, or maybe if our other amp needs a service.
I can't claim to fully understand the merits of the FET VAS implementation. Is it good?
It's been a long time since I did my own PCB layout, so I'm having a play with KiCad.
Considering putting the existing schematic and layout into KiCad as an exercise, might make debugging easier.
But a lot of parts need to come out and be checked or replaced, and the PCB is going to be a mess after that.
I'm not sure where to go with it. What looked like a simple few burned out components is turning into a big job of rebuilding something which may be intrinsically rubbish.
Maybe I should build a pair of completely different amps in the box?
The pre-amp is OK and is conveniently a separate PCB.
The power transformer is OK or I have a toroid I could put in but that is higher voltage still.
The power transistors are probably OK, I could perhaps strip it down to the output stage and just test that?
Would these transistors be a good upgrade for a pair of Chinese kit amps?
How good was this amp when it was working properly?
This particular one is never going to be a collector's item, I would use it in a 'second room' system, or maybe if our other amp needs a service.
I can't claim to fully understand the merits of the FET VAS implementation. Is it good?
It's been a long time since I did my own PCB layout, so I'm having a play with KiCad.
Considering putting the existing schematic and layout into KiCad as an exercise, might make debugging easier.
But a lot of parts need to come out and be checked or replaced, and the PCB is going to be a mess after that.
I'm not sure where to go with it. What looked like a simple few burned out components is turning into a big job of rebuilding something which may be intrinsically rubbish.
Maybe I should build a pair of completely different amps in the box?
The pre-amp is OK and is conveniently a separate PCB.
The power transformer is OK or I have a toroid I could put in but that is higher voltage still.
The power transistors are probably OK, I could perhaps strip it down to the output stage and just test that?
Would these transistors be a good upgrade for a pair of Chinese kit amps?
It may be more effort to 'just fix it' back to its original state, than to build something more interesting into the box.
Its original build state seems to be fundamentally flawed, with several components running at high dissipations for their nominal ratings.
I started out thinking it was an easy fix of one channel, but the other channel is also in need of looking at, it oscillates, but only sometimes!
For me it's an interesting thing to do, get to understand an amp and learn from it, more than actually needing another amp.
I guess I should properly test the output stages, because buying replacement power transistors might not be worthwhile.
I might set it aside for a while and put some time into my test equipment first.
Its original build state seems to be fundamentally flawed, with several components running at high dissipations for their nominal ratings.
I started out thinking it was an easy fix of one channel, but the other channel is also in need of looking at, it oscillates, but only sometimes!
For me it's an interesting thing to do, get to understand an amp and learn from it, more than actually needing another amp.
I guess I should properly test the output stages, because buying replacement power transistors might not be worthwhile.
I might set it aside for a while and put some time into my test equipment first.
I think what I was asking a few posts back is:
Is this a gem of an amp which I should want in a system, or is it just A. N. Other 40-60W 'value' amp?
Would you cherish the sound of one of these over a Pioneer or Arcam?
Will it be miles better than the Sherwood amp which 'does the job' in my workshop?
Is it special in any way?
Is this a gem of an amp which I should want in a system, or is it just A. N. Other 40-60W 'value' amp?
Would you cherish the sound of one of these over a Pioneer or Arcam?
Will it be miles better than the Sherwood amp which 'does the job' in my workshop?
Is it special in any way?
As I've designed this amp over a quarter of a century ago, I might try to answer these questions
It would be nice to make a replacement PA board for the A3i, from FR4 material, implementing all the mods, providing for a good thermal performance on the PCB.
Cheers
Alex
So happens that A.N. are my initials (true). And A3i is the design I am simultaneously proud and somewhat ashamed of, due to a horrible build quality in production and some "management decisions" leading to local overheating.
Yes, absolutely, after the mods. It was and still is a good amp with excellent sound quality. I still own a modified A3i and for about fifteen years it was on 24x7 as the main amp in my home system, driving a pair of Epos M12. Right now it needs a new output relay and I have to use my second amp - Creek Audio 4330R. Sound quality - wise these two are comparable.
Yes.
Yes, in several ways. The VAS is pretty unique (I've designed it specially for A3i and later a modified version of this VAS topology also was used in Creek Audio 5250 and A52 amplifiers). The main aim was to provide a good linearity with a symmetrical output current limit. The circuit is simple but even at a reasonably high gain gives an excellent distortion performance. Output devices are real Sankens and worth probably more that the rest of the amp. The preamp section can be fully by-passed leaving only the switches and the volume control in the signal path before the power amplifier (that is how I run it) .
It would be nice to make a replacement PA board for the A3i, from FR4 material, implementing all the mods, providing for a good thermal performance on the PCB.
Cheers
Alex
Thanks for your reply Alex.
I think what I intend to do is test the Sankens, then work towards a new PCB.
Around the time you were designing this, I was designing/developing quite different electronic hardware. In some ways it's good to know that what I worked on became obsolete, got scrapped and isn't going to haunt me into my retirement. So I understand your feeling about this product.
I'd like to understand more about the FET VAS, I might have to read a textbook or two!
Thanks again,
B
I think what I intend to do is test the Sankens, then work towards a new PCB.
Around the time you were designing this, I was designing/developing quite different electronic hardware. In some ways it's good to know that what I worked on became obsolete, got scrapped and isn't going to haunt me into my retirement. So I understand your feeling about this product.
I'd like to understand more about the FET VAS, I might have to read a textbook or two!
Thanks again,
B