BUGGER!

WerTicus · « **on:** 5 Feb 2004, 01:22 pm »

Hey techinal people... I recently did something silly and accidently had the positive shorting directly to the negative on my amp

It was playing for a while before I noticed the clipping light on... but now the clipping light is constantly on - even though it 'sounds' okay... though I have not done extensive listening in an effort to prevent further damange...

This is the schematic for the amp - I was wondering what was likely to be broken given that it was playing for several hours shorted and that the clip indicator lights up with any signal being passed through the amp now.

http://www1.jaycar.com.au/products_uploaded/KA1760%20REV1%20SCREEN.pdf

thanks for any suggestions as i try to trouble shoot this problem! im guessing a mosfet has died?

JohnR · « **Reply #1 on:** 5 Feb 2004, 11:36 pm »

You're saying it still works?

I haven't looked at the circuit, the file is huge. However, I'd suggest measuring the DC voltages around the circuit, starting at the output stage. I'd also suggest disconnecting the speaker until you know what's going on

WerTicus · « **Reply #2 on:** 6 Feb 2004, 04:01 am »

yeah it does still work... and it dosnt sound like it is clipping at all...

TG · « **Reply #3 on:** 6 Feb 2004, 08:50 am »

Quick reply off the top of my head - I have not looked at your documentation, but if the amp incorporates some sort of output short circuit protection and it wasn't shorted for too long a time then you may just have fried the clipping detection rather than the output FET itself - just a guess, as I say.
If you're not comfortable about poking around inside the live chassis and everything sounds OK I'd try just running the amp for a while to see how it goes - however I'd strongly advise buying a couple of $5 drivers and hooking up to them for testing, whatever you do. There is nothing more distressing than hearing a bass or midrange coil cooking in perfect fidelity when a power rail goes open, and if the amp does take it into its mind to oscillate itself to death ... believe me, I've had this happen and it is both expensive and painful. Any DIY amps should always be first tested on something cheap, not your precious main speakers.

Good luck!
TG

WerTicus · « **Reply #4 on:** 6 Feb 2004, 11:43 am »

okay so if i measure the output from the positive and it is 'still' +- 30mv or less. i can probably assume its just the clipping light shat itself?

i would really appreciate it if someone who knows about this sort of thing actually looked at the diagram and told me what was more likely

i will hook it up to $5 drivers before i do any futher testing

JohnR · « **Reply #5 on:** 6 Feb 2004, 12:45 pm »

Did you measure it at less than 30 mV or are you just hypothesizing

WerTicus · « **Reply #6 on:** 6 Feb 2004, 01:45 pm »

i lent out my multimetre so i dont know yet but that would be a good hypothesis right?

TG · « **Reply #7 on:** 7 Feb 2004, 12:27 am »

> lent out my multimetre so i dont know yet but that would
> be a good hypothesis right?

If all the voltages around the circuit check out OK, and the power rails are stable at their nominal voltage, and the DC offset is below 50mV, and you don't pop fuses and all this is maintained after 15 minutes or so then you're pretty safe. There's a transistor near the clipping LED - I bet it's fried.

If everything measures OK (when you get your meter back) do a prolonged signal test - attach a load (a big 8ohm resistor will do in a pinch, but a pair of el cheapo 8ohm drivers from Tandy or Dick's will be cheaper and also present a real world load) and pump something through the amp for 24 hours or so.

Good luck.

WerTicus · « **Reply #8 on:** 7 Feb 2004, 10:49 am »

okay tg ill give this a go when i can... ( the guy just got married and took my multimetre on honey moon i believe)

thanks for you help!

if its someother problem that needs trouble shooting ill post it here!

TG · « **Reply #9 on:** 7 Feb 2004, 02:52 pm »

Quote from: WerTicus

( the guy just got married and took my multimetre on honey moon i believe)

WerT old son, the mind fairly boggles at the implications of that - do you think he wanted to measure his wife's resistance or test the continuity of the relationship?

WerTicus · « **Reply #10 on:** 9 Feb 2004, 12:44 pm »

im not sure what he is going to do with my probes and his wife but the guys obviously into this hobby!!

its only a 20$ metre anyways so if it comes back 'damaged' i could always get another one

WerTicus · « **Reply #11 on:** 27 Feb 2004, 06:39 am »

OKay I have gotten my multimetre back and I have tested the output of the 'faulty' channel when its just idleing...

the output DV voltage is 15mv... which is what it is supposed to be and what the other channel is more or less and more or less what the other amp thats the same is....

so this tells me its a faulty clipping detection light correct?

so..... what should i do?

WerTicus · « **Reply #12 on:** 29 Feb 2004, 04:23 pm »

also the voltage going through the circuit is 142.7 v vs 142.3 v on the non stuffed channel

so 'looking good' eh?

WerTicus · « **Reply #13 on:** 29 Feb 2004, 04:31 pm »

so TG would you recommend i do anything other than replace that transistor? (while the case is open?)

TG · « **Reply #14 on:** 2 Mar 2004, 05:02 am »

Hi - sorry, been away for a day or two. Tha amp sounds like it's working fine (you were testing under load, yes?). The problem is almost certainly in the clipping detection. I'd try replacing the LED (it is an LED, isn't it?) first. It may just be shorted "on". If that doesn't work take a look at the transistor. Buy both and save a second trip to Jaycar

neither is expensive.

Good luck.

WerTicus · « **Reply #15 on:** 2 Mar 2004, 10:13 am »

it only clips under a signal... any tiny signal, load or not

i didnt test the volatage across the circuit under a load just without load - i shall do and report back what i find... i guess ill buy a 5 doller woofer to test that

and i suppose test it with a load and a signal too?

WerTicus · « **Reply #16 on:** 6 Mar 2004, 05:50 pm »

Okay!!!

I replaced the clipping led circuit transistors and the clipping problem went away so i put a 5 dollar driver onto the output and played it at the low volumes that i thought suitable for the driver to handle...

she played sweet (for a 5 dollar driver) for about 3 or 4 hours

and then i noticed a very distinct burning smell and also that nothing was playing anymore!

MY FIVE DOLLAR DRIVER WAS COOKED!

(yeah okay im not upset about that it served its purpose)

But i tested the output and now its totally opened up to the 75.5 v rail voltage that is DEFINATLY not ment to be there.... so

... now... i suppose this is where it gets expensive?.....

ummm.....

... HELP!

WerTicus · « **Reply #17 on:** 7 Mar 2004, 04:12 am »

also i just noticed today looking at the amp that its DEFENatly one of the mos fets has pretty much caught fire.

there are 4 of them for each channel so maybe only one needs replacing i am just looking up the schematics to see which...

its q10 the 2sk1058 N- CH mosfet ..... looking at the diagram....

http://www1.jaycar.com.au/products_uploaded/KA1760%20REV1%20SCREEN.pdf

okay that makes sense i suppose....

If q10 dies there is a DIRECT path from + voltage to the output!

so its not longer redirecting power into the curcuit at all... which makes me think its the only one part that needs replacing?

i get the feeling thats 'right' but this is definatly beyond me.

WerTicus · « **Reply #18 on:** 7 Mar 2004, 08:14 am »

also it looks like the diode d4 and R24 and r13 are toasted nicely...

d4 is a 400v diode which may have caught fire killing the resistors or maybe the resistors killed the diode or maybe then all died... im not sure what is incedental damage here...

Actually im glad there was nothing flammable nearby - the appear to have actually caught fire, although i didnt see it happen, there is lots of black residue near the site.

and a signal cable that passed above it it pretty damaged though not through its insulation as far as i can tell so far.

WerTicus · « **Reply #19 on:** 7 Mar 2004, 08:35 am »

upon further inspection its obvious that r15 r 14 r 16 r 17 are all burnt out and unfortunatly r15 took out c17 and c18 with it... though c18 might not be too badly damaged. i can test their capacitance when i remove them...

and it appears that q12 and q9 mostfets are also burnt out in addition to q10

though q11 seems to look okay i have my doubts as to it being okay!

zd1 looks to be cracked and the other zd2 and d2 and d1 may possibly be cracked....

basically ladies and gentlemen this circuit is stuffed

the capacitors that were isolated from any exploding resistors have not being damaged as far as i can tell however so although i am definatly going to have to rebuilt the whole circuit its not going to need total replacement - but i am posting here for advice!

i can certainly give some advice though - do not short your amp from positive to negative with no load!