Identifying/Replacing Diodes

[Saco]

Hi,

Long time Youtube viewer. First time trying to fix a slightly more than basic device: an audio effects processor. Please be gentle.

At first it seemed like it was "just a blown fuse" but when I replaced it, as soon as I plugged it in, the fuse blew again, so it is some short somewhere.

I started by checking all the diodes with my crappy 2$ multimeter and right off the bat I got different kinds of readings, which got me confused about:
1) which diodes are good and which are not and
2) what type of diodes they are.

I didn't take out the board so I don't know what is going on on the other side of the board, but just to mention the diodes that are easily accessible:

- D1 through D4 show a reading of 710 in both directions.
- D6 through D8 show a reading of 750 in the flow direction and OL on Reverse.
- D9 shows a reading of 70 in both direction.
- ZD1 shows 75 on both directions
- ZD2 and ZD3 show 950 in the flow direction and OL on Reverse.

In terms of their working state, by looking at these readings it looks like almost all diodes are faulty, which is a little hard to believe. Is this true or am I missing something/getting bad results because of in-circuit readings?

When it comes to diode identification the question is more generic: how can I tell if a Diode is a rectifier, zener or just regular? Searching on the internet it seems that this can only be done by looking at their datasheets, but some of these diodes seem to be only marked with two digit numbers (like 48, 27, A2) which I was assuming was their voltage rating until I saw the A2 one and which makes it very difficult for the noob me to even find the datasheet. In this particular case, the Z in ZD1-ZD3 seems to indicate they are Zener diodes but that's pretty much all I've got.

Here are some pictures:

[Audioguru]

When you measure a diode in the circuit then you are measuring everything connected to it. Bad test.
The little orange diodes can probably be replaced with a 1N4148 but they would not cause a fuse to blow.

The larger black rectifiers will be marked 1N400x with the x saying its voltage rating. If one or more is shorted then a fuse will blow. You did not measure a short for D1 to D4 so they are probably fine.

[Twoflower]

Careful! Some of he orange diodes are labelled ZD and also the schematic symbol of a zener diode. This indicate that these are not normal diodes but Zener Diodes. But You have to replace the diodes with the right zener voltage.

But Audioguru is right: They are probably not the case of the blown fuse. Measure diodes in-circuit might cause false readings and depend on the surrounding circuit.

[Saco]

Thank you for the replies.

I also figured the ZD ones would be Zener diodes.

I will have to test them out of circuit, I was just trying to cheat my way in to maybe narrow down on which would be the most likely cause of failure.

As for eventually replacing the remaining orange ones, Audioguru said they could be probably replaced with a 1N4148. Does this mean I could probably replace them all with the same diode regardless of what's written on them?

[rrinker]

 The regular ones, most likely are all the same. But the Zeners, you need to determine the value, you can't just replace those with 1N4148's.

[james_s]

Those small signal diodes are rarely the problem when something is blowing the fuse, at least not the main problem. The very first thing I'd check is the chopper transistor, it will be on a heatsink somewhere in the vicinity of the bulk filter capacitor, the biggest electrolytic cap on the board. If that failed it may take out smaller parts in the vicinity.

[Saco]

Those small signal diodes are rarely the problem when something is blowing the fuse, at least not the main problem. The very first thing I'd check is the chopper transistor, it will be on a heatsink somewhere in the vicinity of the bulk filter capacitor, the biggest electrolytic cap on the board. If that failed it may take out smaller parts in the vicinity.

Wow unless I killed it while de-soldering, it looks like you're spot on. The mosfet is shorted.

Now about replacing it, I actually have the same question as in the diodes: how to go about find a matching replacement.

From my understanding this is a SANYO K2625 which is also sold as 2SK2625.

What I'm not sure is what the other three digits (8B3) mean. Is it just some kind of lot number? I found the same transistor on ebay but with different numbers (eg. 9H2, 0B4)

[james_s]

The 8B3 is probably a date or factory code, it won't matter for sourcing a replacement part. Japanese transistors usually omit the prefix on the printed part number so a 2SKxxx is just going to say Kxxx on it. Careful of parts on ebay, a lot of them are fakes that may not be close to what they say they are. Look up the datasheet and find a mosfet with a similar voltage and current rating if you can't easily get the same part, subs are usually possible.

[Paul Moir]

A shorted main drive transistor in a switch mode power supply is more often a symptom of another problem than not.  For example, if the drive for it does not switch it off and on like it's supposed to but rather leaves it on, then it'll burn out again in a tenth of a second.  To protect against that when powering it up again, google a "dim bulb tester". 

I would check out whatever drives this transistor and the driver's power supply as best you can.  Is it a chip? 

[james_s]

I've seen bad solder joints kill chopper transistors on many occasions. Power glitches or surges, overheating and other faults can also do it. Not always another bad part.

[gibbled]

R4 appears fractured, could just be superficial though.

[Saco]

Looking at the back of the PCB there don't seem to be any bad joints, but then again, I'm a rather beginner.

My EE experience only goes as far as changing bad resistors/caps/switches and fixing loose joints. Actually analyzing circuits is not my forte at all, so I'm not really sure what to look for to find the transistor's driver.

Looking upstream from the transistor's Source all I see are different kinds of caps, resistors, diodes and the power supply. Here is a picture of the full board (The two big holes and the 3 smaller ones in the middle is where the mosfet and heatsink were):

[james_s]

Sometimes stuff just fails for no obvious reason, I would suggest checking carefully for any other bad parts and replace anything you find, then put it together with a new mosfet and try it out. If it blows up again then head back to the drawing board.

[SeanB]

Before putting in a new transistor check the big power resistors, especially R7, R6 and the one next to D5, as they often go open circuit and blow up the pass transistor. the ones like R7 are probably a low value, typically under 10R, R6 likely is 100k or so and the one next next to D5 will probably be somewhere from 22k to 47k.

R7 does the switch off as the current ramps up per cycle, R6 is a start up supply and the other is part of the snubber on the transformer primary.

[Saco]

Re-hello everyone.

So I finally received the replacement mosfet and proceeded to replace it.

The good news is the unit turns on and the fuse no longer blows.
The bad news is that the power supply (I think it's the power supply) is buzzing a lot (it didn't use to) and after around 1 minute I can start to "smell some heat" (no visible smoke, though).

So back to the drawing board:

The LEDs and LCD connected to the DC board are all dead so I would assume the Power Supply is not... well... supplying power.
Would I be correct to assume that it is probably just the new mosfet that is not up to the task (switching fast enough) thus making the power supply switch at a lower, audible frequency?

[james_s]

I think it's more likely that another fault remains. Rather than making assumptions, I would suggest measuring the output(s) of the power supply and see what's happening there. If there are any signs of life then that will offer clues.

[Jwillis]

Basically a switching PSU works by taking the AC at 60hz ,rectifies it to DC ,puts that into  Pulse Width Modulator (PWM) or transistors to produce a square wave at say around 30Khz.This allows for the use of a smaller transformer.The higher the frequency the smaller the transformer can be .The transformer reduces the AC square wave voltage to what ever is required  then its rectified again to DC. So on either side of the transformer you have a high side and a low side.On the high side their will be voltage regulators for the PWM and or other discretes .Generally they can be zener diodes but not always.There will also be Mosfets to pass current from  the high side to the low side.These are also pulsed with a square wave . The 2 optocouplers PC1 and PC2 are used to keep the high side and low side balanced at the PWM or transistors and mosfets by switching them on or off depending on the load on the low side .If those fail one of 2 things can happen.Voltage run away or complete shut down.

[mjwurtz]

[ Specified attachment is not available ] TC_electronic_SMPS JKS020Q1.pdf (15.53 kB - downloaded 403 times.)Hi there,
I'm probably a bit late on this thread, but had the same problem.  Replacing the MOSFET didn't change anything.  I suspect that the transformer was hurted too and tried to replace the whole power supply, only to discover that it's a taiwanaise manufacturer (and so no luck with the usual chinese sellers ;-)) which sells them by quantity of 1000.  At least, I managed to get a schematic (see attachment) for the Q1 model.  I have the Q2, but they seems very similar.
Since I don't know what cause the problem (diode, transistor, optocoupler, etc) and wanted to avoid tampering with high voltage, I just replaced the whole thing by a 2x15V transformer, using the 7815 and 7915 voltage regulators from the old board, and 2 chinese buck converters for the +5V and +3.3V...

[james_s]

Speaking from experience, the transformer is almost never the problem. I never saw the OP come back to this thread but I suspect the power supply had faulty electrolytic capacitors, which is probably what caused the original mosfet to blow and probably why the PSU was buzzing and overheating after replacing the obviously blown part.