please help--serious troubles with my 55

zachslc · « **on:** 2 Jul 2003, 05:11 am »

My aksa has been running happily ever since I completed it (with a fair share of problems) last summer. Today I was listening for a bit and suddenly one channel got very, very distorted. I tried switching everything in the chain, including the power supply. I resoldered all connections, used different speaker cables, etc. I know it is something on the board itself. sometimes I can switch it off then back on and it will play fine for a few seconds, then develops the distortion again. Also, it usually takes a very long time to discharge the caps after switching it off.

Please tell me this is probably not a transistor problem. Where should I start for trouble-shooting? What do you think it might be?

Thanks a lot,
Zach.

AKSA · « **Reply #1 on:** 2 Jul 2003, 07:02 am »

Zach,

This does sound like a transistor problem, but it could be almost anything.

You really need to get technical about it, and measure voltages all over the amp - with input disconnected, but power on.

You will need:

1. Voltage across R1.
2. Voltage at top of R9 wrt earth.
3. Voltage across R3.
4. Voltage across R4.
5. Voltage across R5.
6. Voltage across R6.
7. Voltage across R15.
8. Voltage from TP1 to TP2.
9. Voltage at output wrt earth.

This prescribes the entire DC operating conditions of the amplifier, and once known gives me a complete X-ray of the amplifier operation.

In the meantime, check the driver resistors, T5 and T6. Often the connection to the pcb is faulty because the leads are short and people fail to push them right through the pcb for soldering. The symptoms you describe could well be this; I had one a week ago in the Philippines behaved exactly like this.

Please make the measurements, then I can diagnose it.

Cheers,

Hugh

zachslc · « **Reply #2 on:** 13 Jul 2003, 03:54 am »

Hugh, thanks for the time. I finally got some emasurements, and have taken them a few times with similar results each time. They are not good.

across r1--.02v
r9 to earth--.01v
across r3--3.7v
across r4--.45v
between r5, r6, r19, and the trim pot leads there was no voltage my dmm would measure (eek!)
the output starts at about 1.5 then steadily declines until I repower the amp. This makes sense (sort of) since the spl slowly declines to the point where there is only a low humming distortion when connected.

Any ideas?

Thanks.

AKSA · « **Reply #3 on:** 13 Jul 2003, 04:55 am »

Hi Zach,

OK, thanks for the measurements, we're homing in....

Now, can you measure the voltage at the top of R3 wrt earth, then the bottom of R3 wrt earth?

This will confirm things for me.

Cheers,

Hugh

zachslc · « **Reply #4 on:** 13 Jul 2003, 05:59 am »

.6v from the top to ground, and 36v from the underside to ground. Can that be right? Seems screwy to me, but I am pretty sure I was in the right place. I'll revisit it in the morning with a bit more sleep.

Thanks again.

AKSA · « **Reply #5 on:** 13 Jul 2003, 08:46 am »

Zach,

I suspect that one of the driver leads is not properly soldered to the pcb.

Please redo all six joints on T5 and T6.

This should improve things. Indications are that the first three transistors are OK, although a really accurate reading across R4 might be a good idea.

Cheers,

Hugh

zachslc · « **Reply #6 on:** 13 Jul 2003, 07:03 pm »

Hugh, I removed and re-installed both T5 and T6. No audible change. The voltage across r4 now reads .37, but FWIW my mains voltage varies quite a bit and the DC in varied between 35 and almost 38 volts. I probably should be more aware of this in making future measurements.

So, Hugh, how do you suggest I proceed? Would it be foolish to take t5 and t6 from the other amp and install them on the problematic board, just to see what happens?

Thanks again for the time and help.

AKSA · « **Reply #7 on:** 13 Jul 2003, 11:07 pm »

Hi Zach,

Next step is to measure the following voltages with the amp on, no speaker connected, and both fuses in and good:

1. Across R15 (should be 1.15 to 1.25 volts)
2. Voltage across R5 (should be 13.5 to 15 volts)
3. Voltage across R6 (should be 20.5 to 22.5 volts)
4. Output offset with respect to earth (should be less than 30mV, positive or negative, let me know which and ensure black probe is on earth).
5. Voltage between top of R13 and top of R14 (should be 2.1 to 2.3 volts).
6. Voltage from TP1 to TP2 (should be 55mV as per bias instruction).

This prescribes the entire DC operation of the amplifier and will give me an X-ray of the problem. At least that's the hope!!

Cheers,

Hugh

zachslc · « **Reply #8 on:** 14 Jul 2003, 03:45 am »

hugh
across r15, r5, 56, and the trim pot I can not measure any voltage at all.

Now it seems that the output voltage is relatively stable at -.12v

there is nothing my dmm will measure between the tops of r13 and r14.

Unless you have better ideas, I am contemplating disassembling the whole amp, cleaning the PCB well, and starting from the ground up again. However, I find it odd that it ran fine for a year then suddenly started acting up. It's a mystery to me.

THanks.

Ciao.

AKSA · « **Reply #9 on:** 14 Jul 2003, 03:55 am »

Hi Zach,

From what you report, it appears to me that the output stage is not passing any bias current at all.

Problems with the AKSA are always confined to active devices. Nothing is overstressed, but misadventure will occasionally break things in this game, particularly when the output stage is not protected, as it is not on the AKSA for sonic reasons.

I believe that the feedback transistor, T2, is faulty. It's most unlikely to be outputs or drivers, which nearly always fail short circuit.

This is a matched pair with T1, and you need the two as a pair. I can send them on to you, but first you should check something.

Try resoldering all the connections around these two transistors, and also around T3, the voltage amplifier, and R4, the base/emitter resistor. It is quite possible you have a faulty joint there, and I would not recommend tearing it all down and starting afresh. To avoid the prospect of ruining the pcb with immoderate disassembly it's always better to try to trace the single component causing problems.

Cheers,

Hugh

fred · « **Reply #10 on:** 14 Jul 2003, 03:43 pm »

Hugh - this reminds me of our e-mails when I was assembling my AKSA. Therefore, it occurs to me that you must go through essentially the same troubleshooting questions over and over. Have you considered saving yourself some time (in the long run) by documenting a troubleshooting procedure? It would be nice to be able to determine which cap or transistor is bad on our own. It's bound to save you a host of work, after the intial effort of documentation.

AKSA · « **Reply #11 on:** 14 Jul 2003, 09:52 pm »

Fred,

Thanks for your post. Your GK-1R ships today at long last!

Yes, I've thought of this, documented a more thorough analysis in the instructions, but unless I give specific direction via email I find it doesn't seem to work. People almost demand personal communication. In the roughly 30% of cases where there is a problem this happens every time, and I'm at a loss to know what to do about it. Nothing, I suspect.

Good thing I type fast!!

Cheers,

Hugh

fred · « **Reply #12 on:** 15 Jul 2003, 02:18 am »

Quote

Your GK-1R ships today at long last!