Soundcraft mixer - sparks n smoke!

[Phil Mcintyre]

A few weeks ago my Soundcraft M12 mixer went bang. It had been left turned on over night, and stopped working the next day - no power. I checked the fuse which had blown, and so tried replacing it - that's when the fun started. A blue spark, and a puff of smoke came out of the back after turning it on with the new fuse, but the fuse was still intact.

I don't really need the mixer so I had just put it away, planning on selling it on Ebay as faulty. But, I decided today to have a look inside it. I've removed the PSU board and can see that a few resistors have been blown, and also one of the smallish caps has some solid, clear gunky looking stuff underneath it (is that what's meant when I hear people saying a 'leaky cap'?) There's also a small brown mark on the bottom of that large cap.

Anyway, after looking at it, I've decided I'm not going to attempt repairing it 'cos I wouldn't know where to start trying to figure out what's caused it. I'm going to contact Soundcraft tommorow to see if they can supply me with a replacement for this board, and wondered if it is likely that more than just this board is damaged?

 

[PRR]

It's a switcher power supply, which means bone-headed logic won't be enough to fix it. The damage suggests a power semiconductor has gone short.

See if the company will sell you a new board.

Odds are the rest of the mixer is fine, but I can't think of a simple way to be sure.

 

[mcs]

A repair kit for a switch mode supply often includes all the electrolytics, some resistors and most of the semiconductors. So in other words: to repair it replace all parts :wink:

Best regards,

Mikkel C. Simonsen

 

[Phil Mcintyre]

Thanks for the replys! I'll phone them later for a new board - hopefully it wont be too expensive. There's a few surface mount caps and resistors on there anyway, so probably impossible for me to fix :grin:

Thanks
Phil

 

[Phil Mcintyre]

£81.94 it cost me! It better bloody work :shock: It does also have a A/D converter built on it aswell, so maybe thats why its so pricey.

How can you tell its a switch mode PSU, and what does that even mean? I got to pretend I was clever on the phone when he asked me if it was a linear or switch mode one I had :wink:

 

[mcs]

If the AC power goes directly into a big transformer, you have a linear supply. If the AC is rectified and goes into a small transformer, then you have a switcher.

Best regards,

Mikkel C. Simonsen

 

[Phil Mcintyre]

I see.. thanks again :razz:

 

[PRR]

> How can you tell its a switch mode PSU

In a "linear" power supply, making low-voltage DC, the AC wall-power input goes through switch and fuse to a big power transformer. The other side of that is low-volt AC. This goes to a rectifier, then to a big capacitor, and now you have low-volt DC. This may need to be further smoothed or regulated. But you can go through step-by-step and see what is happening, what isn't happening, and most of it is on the "safer" side of the transformer.

In a switcher, the AC wall-power input (almost always) goes to a filter-choke, because we don't want what is about to happen to leak back to the wall wires. You have a power switch here (many switchers don't). Then your 120VAC or 230VAC is directly rectified to 320VDC on the main capacitor. You could check this, but it is dangerous like sticking your finger in an outlet (even worse), and this part rarely fails. This 320VDC is "buzzed" at ~20KHz by a chip and a transistor or two, to excite a relatively small (compared to "linear" of similar power) transformer. The transformer isolates the wall-power from the chassis (this board has a White Line showing the division between un-safe wall-power parts and safer circuit-side parts, running under the transformer). The other side of the transformer has low-volt 20KHz AC. The bits on the left are rectifiers and caps to turn this into smooth low-volt DC. As an added frill, there is probably a feedback part from the low-volt side to the "buzz" transistor, to control its "ON" time and regulate the voltage at the low-volt side despite variations of load or line-voltage.

It is barely possible the actual path zig-zags from wall-power through switch to the transformer, and back to the rectifier and cap, then back to the left again. This seems unlikely, also the transformer looks too small (tranny size is related to frequency, and 20KHz AC takes a lot less core than 50Hz AC to carry the same power), also that choke right AT the power inlet is characteristic of nasty switchers that would wipe-out every AM radio in the house if their buzz got out to the power cord. (So are the box-caps below it, which seem to have been omitted here; perhaps their function is included in the large power-inlet connector.)

There are people here who actually design switching power controllers. I'm not one of them. I know the basic idea, but I don't understand the details nearly enough to diagnose one. And what seems to happen is a cascade of failures, whereas "linear" supplies usually fail at one point. If you have AC, but no DC, replace the rectifier. If it fails, you have a short; start cutting parts out until you get DC again, and then replace whatever was shorted. If you have DC but it is not smooth, replace the cap. But switcher theory is all dI/dV and hysteresis curves and special-spec transistors working at 101% of their ability..... as MCS says, a repair kit (if there is one) is usually all the power parts. Probably most of them failed, and it is easier to replace them all than to figure out which ones are truly bad.

 

[henk]

The best thing to do is buying or building a other linear powersupply.
Switch mode powersupply is not good for audio , the frequentie from the trafo 20Khz to 80Khz is audio trough the table , because the filtering at the end is not good enough.
These powersupply's are designed for computers not for audio!!!!!

 

[Phil Mcintyre]

Thanks for the explanation PRR, that certainly satisfied my curiosity :grin:

henk: They actually gave me the option when I spoke to them to buy a linear one, but I thought it'd be best to get the same as I had :sad: I did think about looking into building a PSU, but on the same board there is also an A/D converter which I may want to use.

 

[Svart]

Your board is definately a switcher.

I'm sorry to have to disagree with PRR on part of this but the main cap after the rectifier tends to go bad a lot. with heat and usage they tend to shift ESR greatly causing noise and eventually a misbehaving PSU if they just don't fail outright. Quality caps here matter greatly but the size/value of the cap is so large that manufacturers tend to cut cost here by getting something cheap. big mistake.


However in this case it is likely that the transformer driving transistor( more than likely a MOSFET) has gone bad, probably punching through on the gate and thus burning the resistors up. It's not likely that such small resistors are used in the actual power path, more likely in the gate drive path. this is another common failure mode and is also linked to cutting corners in design.

The one thing I don't see is a feedback loop. I might be wrong but I don't see ay type of optocoupler/optoisolator from the output side feeding back to the PWM IC(the 8 pin DIP part near the "buzz") it is also common in switching PSUs to leave this out and just let the PSU push electrons without care. 75% of the PSUs will probably never see a problem, the others might just die a quick death from part failure and be replaced while still others might die from being shorted somewhere else in the board.

Hmm that brings up another point.. I don't see any isolation for the gate of the MOSFET from the drive IC.. usually you will see some kind of small pulse trafo or other assorted goodies between them.. I still say this was a quick and dirty cut n' paste from the app notes design.

I say check your board for possible problems elsewhere first but where the parts actually burned tells me that it is likely just part failure in the PSU.

:guinness: