sparks then smoke - help

atenolol50 · « **on:** 18 May 2004, 02:04 am »

i completed 1 board of the aksa55 and decided to test it.
all hookup was done with crocodile clips.
in step 3 of amplifier testing was where the problems occurred.

upon powering up, i got sparks on the copper side of the board, then smoke and the dreaded burnt smell.
the smoke seems to come from the non-polarised C13 and C9.

the voltage across R21, is about 3.4V.
the power supply output voltage is approx 38V.

whats next now?

EchiDna · « **Reply #1 on:** 18 May 2004, 05:49 am »

can't directly help you as I don't have the 55, but I do have the 100...
sounds pretty serious if you are seeing sparks!

small piece of advice, make a coffee, sit back and relax, what is a day spent waiting afterall?
the grand master Hugh Dean will get you going soon enough

GL!

welcome to the world of Singaporean AKSAphiles

-saw your post on echoloft!

AKSA · « **Reply #2 on:** 18 May 2004, 07:52 am »

Try this:

Remove the rail capacitors - the electros right near the spade connectors -and hook up again and switch on, with fuses out, of course.

You very likely connected the rail caps wrong way around, and there will be no other damage. Try good quality 100uF 63VW electros in this role, see what happens.

If all is well, buy new caps and put 'em in (brand not too critical).

Only 3.4V across the fuse resistors is not in any way dangerous; the electronics sounds fine, only the rail caps are destroyed.

Cheers,

Hugh

EchiDna · « **Reply #3 on:** 18 May 2004, 08:25 am »

atenolol50,
head down to KOBA at people's park centre 4th floor. They have this value in 5+ brands (around S$1 each), even muse and gold tunes if you hunt about (up to S$2.5 each)

mb · « **Reply #4 on:** 18 May 2004, 09:15 am »

I've had ok experience with Koba, but have also seen many items that I would never put into an amp. Dodgy looking caps that are dented, look very old, are obsolete, and some that look like outright fakes. Check carefully before purchasing and installing into your AKSA.

atenolol50 · « **Reply #5 on:** 18 May 2004, 10:09 am »

thanks.

i'll relook my rail caps connections.

how about the C13? thought i saw some smoke from it.
do i need to have it replaced?

kyrill · « **Reply #6 on:** 18 May 2004, 12:50 pm »

Hi atenolol50

You gave 3.4 v over R21, I suppose you read the same value for R22.

C13 and C9 are physically close to the smoke of C12 but C9 not electrically. SO it is hard to tell in the few seconds ( I hope : ) you allowed the voltage on the board, if there was smoke from C9. C13 is a non polarised capacitor and should take the voltages with no harm, as it was designed for Follow up Hugh's advice first, before going to Koba. As I understand it (as an amateur) you have to desolder C12 and C10, remove them and do yr testing until setting the bias, with fuses removed. IF no fire and smoke happens, replace C12 and check the polarity of C10. Probably C10 is configured right on the board. If not, I would replace it too, to be sure.

atenolol50 · « **Reply #7 on:** 18 May 2004, 01:22 pm »

rechecked all the rail caps, the polarity are correct.

i noticed that my R21 and R22 are burnt.
they no longer give a resistance of 100R now. more like 91R.
is it a different problem?

i would need replacements for these too.
they are 1% tolerance, how many watts?

i've soldered new R21 and R22 in place now.

should i proceed with the tests that hugh suggested?

kyrill · « **Reply #8 on:** 18 May 2004, 03:00 pm »

The parts and the pcb are proven problem free.
99.999% of the faults is in soldering (something does not made a connection) or parts on the wrong place or right place but not with the good polarity.
Check now the power supply wires to the main amp pcb. is the + and _ connected the right way?

The question seems trivial but with extreme logical things like machines or amps you check from bottom up.

If r21 and r22 are burned than you must have higher voltages than 3.44 or so.

Also just as a shot in the dark, as it happened to me. All the transistors T6-T8 can only be at one place. T7 and T8 and T5 and T6 look identical, but they are not. You cannot interchange them on the pcb, which physically is possible .

I also had to check carefully every solder joint . Especially the wires with the shrink tube over them from T4 (in my case) one was so short, I had a good looking solder joint on the copper side of the pcb, but the wire was within the thickness of the pcb and did not made contact. Argghh.. Anothter reason for a burned R21 or r22.

You have to wait for Hugh. It is better if you email him and share your findings. It is quicker that way.and having a private email with him is part of the service.

AKSA · « **Reply #9 on:** 18 May 2004, 11:10 pm »

Hi Atenolol,

Send me a private email; we will get to the bottom of this. It's possible you might need to send it to me for 'surgery'. Explaining complicated tests, particularly if it 'half operates', is tricky stuff and having it on my workbench I can fix things much, much quicker.

Cheers,

Hugh

atenolol50 · « **Reply #10 on:** 19 May 2004, 01:36 am »

my T5 and T6 are swapped!
i hope this is the problem.

any damage to the transistors?

the last time i tried to power up, voltage i saw ... its more like 35V, must have been blinded by the smoke earlier. kyrill is right 3.4V couldn't have fried the resistors.

PJ · « **Reply #11 on:** 19 May 2004, 02:39 am »

Just as an aside....to measure those resistors accurately they must be taken out of the circuit. I dont have the schematic with me (I think I may have lost it anyway), but usually (often?) measuring resistor values in circuit will not reveal their true values as you measure other parallel resistors, transistors, etc.

AKSA · « **Reply #12 on:** 19 May 2004, 03:24 am »

Ate,

You are right; the transistors should be fine, as the 100R resistors limits current to no more than about 320mA. However, you might have fried R15, so check it UNDER the pcb and replace if 'browned out'.

You may also have damaged an output, though the chances are slim.

If R21 and R22 have been browned somewhat, and resistance is down, don't be too concerned about replacing them as once the fuse is in place, they are bypassed and out of circuit anyway. Their sonic impact is negligible.

cheers,

Hugh

atenolol50 · « **Reply #13 on:** 19 May 2004, 10:17 am »

i put T5 and T6 to where they belong.
and with the rail caps in place i now get a voltage across R21 and R22 of 2.7V. no sparks no smoke

phew ....
can i leave it as it is?

in step 4.
the output offset voltage didn't come up on my DMM - scale 20V.
set to a 2000mV scale, it measures approx 10mV.

step 5.
after setting the bias pot such that my voltage drop betw TP1 and TP2 to 55mV, the voltage betw output and earth reads -11mV.
do i need to make any more adjustments?

i must thank everyone here for the valuable advise and concern.
i'm more confident working on the 2nd board now.

kyrill · « **Reply #14 on:** 19 May 2004, 11:35 am »

+/- 20 mv between output and ground is very fine. Less is on the borderline of perfect. Anything under 30 mv is good. And it varies depending on outlet voltages from the power company. You have less than i have.

The second bord will be easier, now you have experience.

EchiDna · « **Reply #15 on:** 19 May 2004, 12:10 pm »

yep Kyrill is completely right here, less than +/- 20mv for offset is considered very good in any amp. In my AKSA 100, I managed under 10 for both channels, but I think that is rare!

I bet you breathed a sigh of relief when you powered it up

AKSA · « **Reply #16 on:** 19 May 2004, 01:35 pm »

Ate,

Your job is done; the amps are perfectly fine and operational, and after putting into a suitable box, you are ready to rock 'n roll..........

Congratulations!! Another AKSA is born........

Cheers,

Hugh

stvnharr · « **Reply #17 on:** 20 May 2004, 01:53 am »

Another AKSA gets born after some smoke and squawk.
Just like my first!!!!