Repairing Alesis M1 Active mk2 Monitor Speakers

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Thanks for the idea Momaka. Its a clever one but I am not sure what additional information it will provide. The different readings I am getting already tell me something is seriously wrong, that I will have as much current as the power input will deliver when there is supposed to be resistance and there is none. What I need is a quick method to track down exactly where this lack of resistance is so I can correct it. Or even a slowish method, other than removing just about everything from the board and going from there.

[Khron]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Ok, so you're measuring lower-than-normal resistances between... what and what?

Some sort of list might be the clearest option.

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Okay Mr Khron66 et al.

I have measured the resistance of almost all of the components anywhere near the two explosion zones. What I did was check the ohms between the legs of each resistor, cap, diode and a couple of other bits. "Normal" means they had a degree of resistance indicated. "Infinite" means the resistance was total as far as I can tell, as close as dammit to an open circuit, or at least higher than my measly meter can detect. "Short" means I measured no resistance whatsoever between the legs of the component, suggesting the part itself was offering no resistance, or that the legs are connected to traces that are shorted somewhere or to a shorted component. Here we go:

C33 - infinite
C34 - short
C39 - short
R28 - short
D13 - normal
C37 - infinite
C36 - infinite
C5 - infinite
C3 - infinite
C4 - infinite
C2 - infinite
R15 - short
R5 - short
D5 - short
C8 - short
R4 - normal
R3 - normal
R11 - normal
R18 - normal
R16 - normal
D6 - short
D7 - normal
C16 - normal
RT1 - short
D1 - short
D2 - short
D3 - normal
D4 - normal
D16 - normal
L4 - normal
D14 - short
C40 - short
Q3 - normal
Q4 - normal
C15 - normal
D9 - normal
R13 - normal
C10 - normal
R10 - normal
C12 - normal

All the components above were tested in place. Diodes and caps were tested both ways. I was expecting a partial resistance one way and total resistance the other way. I would assume that a diode conducting 100% in any direction must be faulty or shorted.

So now, how would this tell me exactly where the problem/s are?

Thanks for any suggestions. I will give it a few days to see if I can, with your kind assistance, find a solution. Otherwise I will just have to order a new power supply board.

The other monitor is now back together and working perfectly with the two large 2W resistors now sticking out the back of the board through a hole I cut in the steel backing sheet.

So at least I am half happy!

[Khron]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

C39 - shouldn't be shorted, but that also shouldn't really impede the functioning of the PSU
R28 - should measure 100 ohms, as per the schematic

Speaking of schematic, you DID download the service manual that i've attached, on page 1 or 2 of this thread, didn't you?

D5 - most definitely shouldn't be shorted, it's the rectifier for the auxiliary winding on the main transformer, which supplies the Vcc to the UC3824, after the startup phase
C8 - if this shows no resistance as well... I'm starting to suspect that maybe the 3824 is toast (again). Or, less likely, the totem-pole on the output (Q3/Q4)

D1/D2 - shorted = BAD, consider replacing them, as well as D3 and D4 - they all form the input rectifier bridge (230VAC -> 315VDC)

D14 / C40 - shouldn't be shorted, leads me to suspect that actually Q2 might be shorted, perhaps?

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Hi Khron666. Thanks for taking the time to go through that. I did indeed download the service manual thank you. Incredibly helpful.

My thinking is as follows: If there is only one short its going to be in a part of the circuit that is common to all the components marked "shorted" above. Secondly the short will either be in one of the components or in the board itself. Thirdly, if I had to remove all the "shorted" components from the board I would be able locate the short on the board easily if it is in fact on the board. If its not on the board I should then be able to identify which components have gone short. Once I know exactly where the faults are I should easily be able to fix it.

Does this sound like a reasonable approach to this? Before I lay into my board with my soldering iron, sucker and wick...

[Khron]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Yes, that sounds like the most sane, and most certain approach.

But it might be wisest to start with the components i've mentioned, that shouldn't be shorted; the others should indeed measure as quite low resistance (either because that's the correct value, or because they're paralleled with a transformer winding).

[momaka]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Thanks for the idea Momaka. Its a clever one but I am not sure what additional information it will provide.

It's simple:

If the PSU is fine (or at least nothing is critically shorted to cause a BANG), the light bulb with flash quickly and then turn OFF or become very dim.

If the PSU is not fine and there is something very wrong with it, the light bulb will stay lit at full or nearly-full brightness.

With this trick, you can safely test the PSU and be reassured that it won't blow your new components.

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Thank you Momaka, that's what I figured. With all the apparent shorts on the board there doesn't seem to be a point in trying the light bulb trick until I have solved them first. When I think I have it fixed I shall give it a try.

Khron666 - Here is what I have now established. I have removed all the possibly offending components as per your/my shortlist, plus the extra couple indicated as "short". I tested them simply for resistance (both ways for the caps and diodes). Four diodes are OK and three are offering no resistance in either direction (D1, D2 and D14). RT1 is also shorted inside. Unless 10 ohm is so small my meter can't see it. The rest of the components are clocking fine.

The board is a different matter! I have a broken trace next to the hole where the short burnt through the board (where it passes through alongside C6). That I can bypass easily. Its one line from Q2 to C40. Measuring across the solder points for C40 and D14 I get an open circuit which I expected.

Next, beginning from T1B I find a short across where R5 was, across where D5 was, through where C8 was to the heatshield connection back to T1A (the square terminal). So despite all components removed in that sequence there is a short between T1A and T1B somewhere. I imagine close to the "explosion" site.

Picking up again at the fork at D5, I find the proper resistance across R4 and R3, and of course that goes to L4 then to T1D, or through D16 back to T1A. D16 is giving me no current one way and resistance the other as I would expect (it was not on the short list). The trace paths from L4 to C40 and D14 also seem to behave like they are meant to. So that pathway seems fine.

Also down that path from D5 is the branch from R3 that goes to C6 (one of the big caps) which is not in the board. There is no short along this route as far as the board is concerned.

The third route from D5 is to U1. This route is intact and there doesn't seem to be any obvious misbehaving from U1.

So it seems to me the problem is somewhere along that first path. To recap, when I test the continuity between any pair of component connectors such as D5's, or even between connections that are further apart along this path, I am registering a short.

I am not sure from here how to narrow it down further. If I am to repair this thing I will have to work out how to break the shorted path within the board somehow. Then I will have to see how to bypass the "cancerous" area and connect the components correctly. Obviously having first replaced the dead ones with good ones.

I sense I am making progress but only in the area of diagnostics. I am imagining that I might be able to cut away the evil part of the board and then use a small modified piece of prototyping board to mount and connect the components for that area. Like an electronic prosthesis.

Is that a sensible idea? Or am I daft in the extreme? Will H&S come down on me like a ton of bricks?

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

I just realised I didn't say anything about C39 which I also removed and tested, as well as R28 next to it. These are in the vicinity of C33 which exploded sideways. Testing connections around this area I am finding continuous current in all sorts of places that I am not expecting it, and that I didn't test before. Judging by the thickness of the traces I am guessing this is the high voltage area. I am also beginning to realise that, as ambitious as I might be about this project, it is way beyond my level of expertise and that I should rather just replace the board. It might be that given enough time I could puzzle it out and make it work, but I have a lot of other things I need to do, so I think its best I call it quits on this.

Am I being realistic or a defeatist?

[Khron]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

D1, D2 shorted - it's good that you confirmed that, out-of-circuit. As i mentioned, replace those, as well as D3 and D4, just to be on the safe side. Plain old 1N4007's are (nearly) a dime a dozen anyway.

D14 = 1 amp 600V fast-recovery diode - that definitely needs replacing.

RT1 = NTC thermistor, used for "soft start", as it were. They usually fail open, and/but if your meter sees 10 ohms as "short", i'd be a bit worried But these should be cheap as chips too - hell, i've got a whole baggie of them, "harvested" from old/busted computer supplies

T1B-area shorts - is that with R5 and D5 removed?

Actually, C39 & R28 form the "snubber" circuit on the SECONDARY of the T1 transformer, so the low-voltage side (the high-voltage side doesn't have as much thick traces, as it has large gaps between traces). That particular pair of windings supplies the tweeter amplifier and the line-level section (opamps in the crossover).

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Hi Khron666. You are very kind to give me so much of your time and expertise, and its been very helpful so far. Yes, I can replace the shorted components relatively easily as you say. Probably at quite low cost too. In fact up to ten people who need the same parts can have them from me for nothing. You have to buy them in quantities as you know.

The T1B area shorts are all with R5 and D5 removed. My (possibly erroneous) deduction is that the board has some serious shorts around the carbonised area and perhaps elsewhere too. Only the new R3 and R4 resistors are in the board now in that area.

Would I be incorrect to guess that, where the current surged through the board for a second or three during the "explosion", the PCB has been turned to carbon which, soaked with melted solder, is now conductive and shorting the two layers of the board at perhaps a number of points?

If my hunch is correct, the only solution I can think of is to cut away the area where its shorted, or certainly disconnect any traces from it, and find another way to connect the components to each other.

If that makes sense, I could perhaps use a piece of prototype board, or make a small PCB to mount just the homeless components on, or make a new PCB completely.

I have a new board on the way, so from here on its an experiment to fix this one. I feel better about that. It means I can do it as and when its convenient.

I still need to work out how to go about seeing if there are issues around your "snubber" circuit, involving C39 and R28. As its a low voltage side would it be unlikely to have burnt the board? Or blown any components? Or is it still reasonable that Chernobyl took out components in this area too?

[Khron]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

I have quite serious doubts that the board itself turned conductive. I mean, it's fiberglass+epoxy, and as far as i know, that's not the best solder-wick material.

If anything, i'm starting to suspect that the UC384x is shorted internally. Extract that, and see if that short persists, on the board - my bet is that it won't

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Hi Khron666. Unfortunately your doubts are not well founded. A phenomenon called Conductive Anodic Filaments has been shown to occur in PCB's under certain conditions, which include high voltage and heat, wherein metallic copper or conductive copper salts are deposited on the glass filaments within the PCB leading to catastrophic failure.



http://www.ami.ac.uk/courses/topics/0152_caf/

This of course does not necessarily mean that your theory of UC384x is incorrect. I shall extract it and retest the pathway. Let's hope you are correct as it would simplify the repair.

Of course at this point it may be of value to recommend that your bet should not have any monetary consideration attached.

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Well isn't that just typical! Just received an email from the people who sell the boards. "Sorry, we are out of stock and so are our suppliers and we expect to have stock in August." August???? No indication on their website they have no stock. They let you order and pay first, with the promise of delivery within 3-5 business days. Then they tell you they don't actually have any.

Sigh.

I'll have to tell my customers, "Sorry I can't do your mixes until August." Wonder if they will wear that? After they have paid me...

[Khron]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

I don't suppose you've bothered removing the UC384x from the board, have you?

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Dear Mr Khron666. It is alas not a matter of whether I bothered. I have had a busy few days with band rehearsals and the like which have kept me from the drama of the board. I shall indeed bother in due course when I can set aside some time. And when I do you shall be the first to know the result. Or the second actually. I'll be the first.

I was just having a little rant about certain supplier attitudes. I'm sure you've endured it at some point.

[Khron]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

More or less, i suppose. I was just a bit surprised you had had time to research that CAF phenomenon, instead of popping that chip out and be done with it But i wish you nothing but good fortune in that

And yes, i know all too well how "lovely" it is when things just pile up and pile up in one's schedule >_<

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Well, I managed (bothered) to make the time to pop that U1 out as you suggested. It may be early days for you to gloat or point fingers at me and jeer, but I tested the resistance across the various terminals of the chip and I get zero resistance across at least two of the pins.

Furthermore, I no longer get zero resistance on the board between the T1B terminal and the square terminal on C8. I am measuring a resistance that I might expect to see due to some components being in the board still. I have no idea of what it is supposed to be though. I do notice however that this line continues on to U1 and C14.

Apart from all the seemingly low resistances I mentioned earlier across the other end of the board (did you say I shouldn't worry too much about them at this stage?) I am feeling much more upbeat about how things are looking.

Would it make sense now to renew all the components I removed and then test the board with that magic light bulb test (if I can find one)? Or a light bulb and suitable resistor in series?

[Khron]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Nah, no gloating "necessary" Just a minor case of "i hate to say i told you so, but... i told you so"

But i'm glad you finally got to the bottom of it And to be totally frank, a blown chip was WAY more likely than some semi-obscure PCB-substrate-going-conductive phenomenon.

But yes, with a new chip, and the other shorted components replaced, and especially with a series lightbulb, you should be quite safe for a testing session

[Ozyris]

Re: Repairing Alesis M1 Active mk2 Monitor Speakers

Well, the weeks have flown. I had a whole lot of stuff I needed to attend to so was not able to proceed with this project. But now the time has come to conclude my story.

I have finally completed the damage-bypassing and component-replacement exercise on my blown speaker, guided by the wise men who advised me on my journey with particular gratitude to Khron666.

Here is what the back of the board looks like now.


The cleanness of the result is deceptive. I have replaced at least 15 components. You can see the hole through the board where the explosion burnt it. There are three places where this happened, this one being the worst.

If you want to know what I had to go through to reach this point, please read back. I have in fact documented the entire exercise photographically. If anyone thinks it would be helpful to have this story visually online let me know and I will put it all on a webpage and link to it from here.

So the outcome? Today I finally assembled the board back into the chassis, cutting a hole in the metal sheet for the 2W resistors to poke through, and fitted it all back into the cabinet.

Tentatively I connected it up and turned it on...

It works perfectly. Who would have thought?

Thank you again to everyone who helped me with this. Specially Khron666.

Great site!