# Need help with subwoofer amp blowing fuse

#### Gravydog

Joined Mar 8, 2018
41
Hi, I'm a retired Honda auto tech and I have helped many people in the past, hoping karma comes through for me. I have a pretty good understanding of what different components do, but the interactions are beyond my experience.

This is a KLH ASW10-120 subwoofer that I have had for many years. I hadn't noticed it working and the power led was out so I dove in. I found the 2A fuse was open, not violently blown. I replaced it and it powered up for a few seconds and then failed again. I took the amp apart and gave it a good look.

A long time ago, it had a hum from the speaker and I replaced the 2 large caps in the power supply that were bulged and all seemed fine. The replacements look good still and test very close to their spec (6800uF 50V) when off of the board. Through tireless searching online I found a schematic that looks correct (mostly) but it calls for 4700uF 50V. I think the electronics store didn't have the ones I needed so I substituted. Should the higher uF be OK for this? I am guessing yes.

I say the schematic mostly seems correct because it calls out the exact pcb number and transformer number I have but diodes D5 through D10 don't exist on my board as far as I can tell.

The transformer is always powered. With the leads unsoldered, it reads 72V open circuit and the fuse is not blowing.

Looking at the pair of 2 watt resistors R62 and R63, and I think the color bands have been baked to uselessness. The schematic calls for 680 ohms, one reads 797 and the other is 1.16k. I will replace both. It's odd that the bands don't match each other and they don't look like they could have ever been the colors for 680 (blue, gray, brown?)

The pair of 0.22 ohm 5 watt resistors R50 and R51 each test very close to spec, showing 0.3 ohms on my 40 ohm scale. R51 has the color bands slightly darker from heat.

I removed the 4 power transistors and they test ok using the diode test on my DVM but I know that's not a real-world test. My best clue at this point is that one of the small power transistors Q5 got burning hot in the short time before the fuse failed while its counterpart Q4 did not. Also, you can see in the photo that the little heat sink for Q5 has blackened the glue on the pcb where it contacts it.

This is where my abilities end. I need someone with knowledge to tell me the likely reason(s) for Q5 heating up. I'm guessing just replacing it is not the answer. If anyone feels any of the power transistors should be replaced, the part numbers are kind of outdated and I would welcome suggestions of suitable replacements. They are (Q4) D1563A, (Q5) B1086A, (Q6) A1516, and (Q7) C3907.

Thanks in advance for any help!

Rob

Here is the power supply PCB:

#### Hymie

Joined Mar 30, 2018
976
From the limited information supplied, I would think that possibly transistors Q3 and/or Q8 have failed short circuit (C-E), continually switching on transistors Q5 and Q7 (applying –Vcc to the speaker) causing the fuse to operate.

It is possible to determine the health of the transistors in circuit (especially determining whether they are short), negating the need to remove them from circuit to test.

With regards the two 680 ohm resistors overheating, this would suggest an overload from the circuit connected at the connector on the diagram left hand side.

#### AlbertHall

Joined Jun 4, 2014
11,302
Was the blowing of the fuses with the speaker connected?
I am just thinking that maybe the amplifier is oscillating and dumping lots of power into the speaker at higher than audible frequencies.

#### Gravydog

Joined Mar 8, 2018
41
...possibly transistors Q3 and/or Q8 have failed short circuit (C-E), continually switching on transistors Q5 and Q7 (applying –Vcc to the speaker) causing the fuse to operate.
Thanks for the quick reply. The fuse also blew and Q5 got hot without the speaker connected. I checked Q3 and Q8 in place with ohmmeter and diode tester, nothing obvious to me but I removed them anyway and with the diode tester they seem fine and both the same readings. I unsoldered one leg of D3 (upper left area) and it tests good. I put the ohmmeter on all the resistors on the board (most in place) and none showed low readings like they were shorted, in my opinion.

The power LED is connected to pin 1 of the connector on the left, it's on the second pcb. It stays lit until the fuse blows. I believe I had the off/on/auto switch on the second board in the off position when the fuse blew. Should I post the schematic of the control board as well?

I noticed that the diode at the top of the schematic labeled D11 but on the actual pcb it's labeled ZD2 and is across C20 as shown.

So, any more clues for me, components to try to test? Thanks again.

Rob

#### Hymie

Joined Mar 30, 2018
976
You could considerer the nuclear option of replacing all transistors (Q3 to Q8) – but with careful analysis you should be able to work out what is wrong.

It is still likely that the fuse will operate if transistor Q7 is biased on, should a signal switch transistor Q6 on, then excessive current will flow +Vcc to -Vcc

Firstly, I would attach a voltmeter measuring the voltage (with reference to earth/0V) at the speaker output (join of resistors R50/R51) then power the amp (noting the voltage before the fuse operates).

You could consider reducing the fuse value to 1A or 1.6A while conducting the fault analysis/testing to limit the fault current each time (to limit the possibility of further damage).

#### AlbertHall

Joined Jun 4, 2014
11,302

#### Hymie

Joined Mar 30, 2018
976
From circuit analysis, the only possible high current path (causing the fuse to operate) is via transistors Q6/Q7. Any other transistor failing will not in itself result in the fuse operating – in my experience (if Q6/Q7 are OK) favourite would be the driver transistors Q4/Q5.

#### Gravydog

Joined Mar 8, 2018
41
Thanks for all the help and suggestions. As a mechanic, I always hated "parts replacers" who took the shotgun approach because they couldn't properly diagnose the true problem but I'm afraid that due to the fact that I have so many components removed already and my lack of experience, that is what I am going to do! At this point I would rather spend a little more on parts than reassemble everything for more troubleshooting. My instinct is that Q5 is shorting under load but we will probably never know. I am going to replace Q3 through Q8, R62 and R63, and C28 and C29 (previously replaced) with units of the called-out values.

The power transistors Q4 (D1563A), Q5 (B1086A), Q6 (A1516), and Q7 (C3907) are outdated and not available on Digi-Key. I like DK because they will ship small orders First Class for only $4.99 and I appreciate that. I tried my hand at choosing substitutes transistors, we'll see how I did when I put it together. Less than$23 for all parts including shipping and tax. If it still has a problem then I will have to muddle through active troubleshooting.

Rob