Faulty ICEpower 50ASX2

[almamater]

All 4 of them measure like this? I would say it looks really good. What about output rails? Nothing shorted? Did you measure on the capacitors? What about +5 rail?

Hello.

The values yes, are more or less the same in the 4. 
I took off all the 4 Capacitors and the values are fine with Low ESR.
I dont see nothing shorted, small caps, resistors..

Now, sorry for my question, but where is the +5V Rail?

I dont know if im doing something wrong, im testing this Power Supply module disconnected from the rest of the amplifier unit (signal cables), connected only the 220v input, i assumed that outputs P103 or 104 should have some voltage, or even the Enable pin of P102..

Thanks for all the patience.

[Manul]

I took off all the 4 Capacitors and the values are fine with Low ESR.
I dont see nothing shorted, small caps, resistors..

I was not suggesting to remove capacitors, but thats nothing wrong. Now you really know that they are good. I was mentioning capacitors because they help to find power rails. At least for me, when I try to quickly troubleshoot something. If I see 5 electrolytic capacitors, I just quickly measure resistance on them. Some power rails may lack electrolytics and have just ceramics, but those often can also be recognized just by looking. For me it is quickest way, because I can test that no voltages on the board is shorted. If someone gives me failed unknown board I just spend frst 30 seconds doing that.

Now, sorry for my question, but where is the +5V Rail?

I don't know, but it has 5V logic level on enable, right? So I assume that some logic is working from local 5V supply. Maybe also 3.3V. Poke around capacitors near IC's. Just to see if nothing is shorted.

I dont know if im doing something wrong, im testing this Power Supply module disconnected from the rest of the amplifier unit (signal cables), connected only the 220v input, i assumed that outputs P103 or 104 should have some voltage, or even the Enable pin of P102..

I do believe, that this board should work on its own.

Anyway, somehow I have suspicion from begining, that the problem is something on the primary side. Do you have scope?

[almamater]

Yes i have one scope (bought very recently but still learning  from UNIT-T), i didn't test anything from this board with him because i dont have one Isolated transformer..

[Manul]

Yes i have one scope (bought very recently but still learning  from UNIT-T), i didn't test anything from this board with him because i dont have one Isolated transformer..

Yeah, but there are always some creative things to do. Connect ground clip to probe tip (short the probe), and move that wire loop very near transformer in a bit different directions, without connecting electrically. Be sure that 20Mhz BW limit is not set, have probe at 10X and have scope on low millivolts. Do you see any bursts of noise from transformer? You may try this with some working SMPS to see how it works.

Idea is: to see if your boards primary side does something, or nothing at all.

[almamater]

Hi.

I made the suggested test with the oscilloscope   

First i test with a good SMPS (salvaged from one printer i guess) and then i test with this faulty one:


The result with the good SMPS:


But with the Fender Module SMPS i don't get anything, only when i turn ON the Power Button, but nothing more..:
 

[Manul]

You probed both transformers, also that one with silver sticker? I saw you had 500mV vertical, did you try lower like 10mV vertical (trigger on auto)? No pulses, no noise? This is what I was thinking, it does not even try to start I guess. There could be something with mains voltage detection and undervoltage lockout or something which prevents starting...

By the way, there were a few cases when there was no tools available, I used AM radio to listen for circuit "activity". Noise which is heard (or not heard) can sometimes be a useful insight. AM radio picks up well not only transformer noise, but also some digital signal activity.

[KevinA]

B&O do not release any service data these days, all B&O dealers repair products by replacing complete modules (not that any of them would have a technician who could repair something).

Even companies who use B&O modules for their sound systems are not given service data - everything is to module level replacement.

You can (I believe) still download Karsten Ballhorn's dissertation on class D amplification, which all their modules are based upon, but there are no schematics there.

[almamater]

You probed both transformers, also that one with silver sticker? I saw you had 500mV vertical, did you try lower like 10mV vertical (trigger on auto)? No pulses, no noise? This is what I was thinking, it does not even try to start I guess. There could be something with mains voltage detection and undervoltage lockout or something which prevents starting...

By the way, there were a few cases when there was no tools available, I used AM radio to listen for circuit "activity". Noise which is heard (or not heard) can sometimes be a useful insight. AM radio picks up well not only transformer noise, but also some digital signal activity.

  yes i try again with 50mV and nothing.. and no sound\noise from transformer..

It seems very complicated to detect the problem of this module...

[Manul]

It seems very complicated to detect the problem of this module...

Depends with what you compare. Of course it is not the simplest thing in the world. But also it does not look to me, that you spent a lot of time on this problem. You measured that there is nothing at the output side, capacitors are good, rails not shorted, output mosfets good. You also measured that main PWM controller is not powering. And that is pretty much all as far as I understand. Have you traced how the PNP transistor is controlled? Mains undervoltage circuit?

A lot of people in this forum would likely troubleshoot such board in one hour or less, but they will not be able to tell exactly how to do it, because there is no exact algorithm, it is just probing, thinking, reverse engineering sections of PCB, finding some datasheets, drawing schematic sketches on a piece of paper and probing again until something is found and the problem understood.

Also I will note, that people generally hesitate to offer ideas which might be risky. For example in this case you might consider to force start the SG3525 by bypassing the circuit which controlls 15V. But it is risky, bad things might happen, more damage to the board. So if I suggest this idea and you respond that your SMPS went on fire and now it is totally destroyed, I would feel really bad about it.

You might consider such repair as opportunity to learn. It really depends what is your goal. If you want just to save some money, then maybe it is not worth.

[almamater]

Thanks for the message.

Yes I've already tried to understand how the PNP is being controlled but I haven't drawn any schematics... anyway, the idea of forcing the SG3525  to start up doesn't seem bad idea to me, what better way to supply the 15V to him, removing the PNP from the circuit with a shunt wire?

[Manul]

Well, yes, you could short emitter and collector to bypass. If you want, do such experiment, but on your own risk. The logic is that it is not powering the SG3525 probably because it thinks that mains voltage is too low or temperature too high. If it will not explode and start running and you will get output voltages, turn it off and try to analyze what exactly is preventing it to run.

[Vicus]

In my opinion something load too much the output of the pnp. It can be the 3525 or something else. I suggest you to power the 3525 from a external and current limited psu WITHOUT apply main voltage.

[Vicus]

Someone know what's the purpose of the smaller transformer?

[almamater]

I tested connecting ~10v directly to the IC 3525  and it seems to work, at least there was no excessive consumption, from the ~300mA that I provided him  it only consumed 50mA.

And the IC output stays active with a voltage of ~3.8v



Possibly I will now make the Manul suggestion to bypass the Transistor.. 

[almamater]

So.. i bypassed the transistor (shorting between emitter-collector) nothing exploded but still no voltage at Enable Pin 

[Manul]

And how about SG3525 supply voltage? Is it normal? How about Vref? Should be 5.1V.

[almamater]

Hello.

The SGS3525 has:

Vref= 1.13v
+Vi= 2.52v
OutputA= 1.2v

  Looks like it's not very well...

[Manul]

Hello.

The SGS3525 has:

Vref= 1.13v
+Vi= 2.52v
OutputA= 1.2v

  Looks like it's not very well...

You wrote, that applying 10V externally it consumed 50mA. This is quite a lot. I would expect way less. This could mean that SG3525 is bad. Something else could be draining the current, but I thing there are not a lot of devices powered from same supply. You could desolder the SG3525 and try again to apply 10V. What is current consumtion? Also you can trace and measure things interfacing to that IC. Or you could just buy a new SG3525 and try. Not very expensive part.

[almamater]

Before buying a new IC, i removed the SG3525 from the circuit and tested again with 10v and seems now Ok, a consumption of 10mA and 5v at Vref.

[Manul]

Before buying a new IC, i removed the SG3525 from the circuit and tested again with 10v and seems now Ok, a consumption of 10mA and 5v at Vref.

I would say it is not a conclusive test, you would need to build a full circuit. For example the output stage of SG3525 could be damaged. You should also check what else on the board is powered from same source.

[almamater]

I  only now managed to test this with a new SGS3525, (I had to buy it online because there wasn't one like the original here).

But same result..  i assume its to difficult (for me) to find the problem of this module.

I appreciate the time you took for this.

[Dorian]

I know the topic is old, but maybe someone will need it, there is a general solution and it is very simple. Coil L401 should be temporarily removed and under the white glue are 4 smd resistors, 2x62k and 2x5k. The 5K resistors are always intermittent and should be simply changed, they will work with 5K1 as well. after changing those two resistances, put the coil L401 back as it was and the module is correct and works. That's all and it's checked.

[almamater]

Ok, Thank you!

I'm going to buy these resistors and try it out..

[kevinj72]

Hey, were you able to try it and did it work?

[almamater]

Yes i tried it but no luck..