Finally I can found and Indigo2 teal, but ...

[llopis]

Ok, can you confirm (with the now damaged 400v cap removed from the board) that the holes it went into don’t register as shorted with a multimeter?

Please perform two measuments:

1. on ohms mode, touch the two, now empty, cap holes (vias) and post the result.

2. set the multimeter in diode mode, measure the exact same place (again) and please post the result.

After those, please use diode mode between the legs of the rectifier (google instructions if you need them) and post the resulting measurements.

I checked and there's no short between the places where the 400V cap used to be. I also checked if they were both connected to the diode that was shorted and only one of them is, so the diode didn't short both legs of the cap.
1) 100 Kohms
2) 0 one way 0.555 the other way

Where's the rectifier there? I'm used to a bridge rectifier (either as 4 diodes or packaged together) but I don't see anything quite like that here.

I'll try to reconstruct roughly the connections from the mains to the cap later and report here. Basically, I'd really like to know if the shorted diode caused the explosion was was caused by the explosion. Argh, I wish I had measured it before!

 

[weblacky]

I don’t have great pics of this board myself, can you post more? I thought I saw two diodes plus perhaps a dual-diode body? Perhaps it is the total sum of diodes just sprinkled around? But need pictures to help.

 

[llopis]

I don’t have great pics of this board myself, can you post more? I thought I saw two diodes plus perhaps a dual-diode body? Perhaps it is the total sum of diodes just sprinkled around? But need pictures to help.

There you go.
The HV side:

And the LV side (?) Or at least I think so. The other panel is probably more of a control logic?

 

[weblacky]

Aren’t I seeing some four-pinned ICs under those side-mounted heat sinks?

 

[mamorim01]

There you go.
The HV side:
View attachment 797

Out of curiosity, what's with the two missing fuses on the high voltage side? Were they removed while you were taking readings?

Definitely interested in this thread as well as gathering as much information as possible on this PSU.

 

[llopis]

Aren’t I seeing some four-pinned ICs under those side-mounted heat sinks?

Good eyes. I assumed it was some kind of voltage regulator, but 4 pins is weird. It's an STR81145 which the data sheet says it's a "hybrid auto switch module doubler" "For automatic switch-over from voltage-doubler to bridge rectification and viceversa". What exactly does that do? It sure sounds it should be near the rectifier.

https://datasheetspdf.com/pdf-file/489105/Sanken/STR81145/1

Could the rectifier be that square metal piece with a screw in the middle? You can see it in all the pictures.
I also have no idea what the long yellow component is in this picture. Some kind of filtering cap like a RIFA?



As for mamorim01's question, I removed the fuses because they blew up with the explosion. I have some new ones of the right type on their way already.

 

[mamorim01]

As for mamorim01's question, I removed the fuses because they blew up with the explosion. I have some new ones of the right type on their way already.

The fuses blowing up should have been the end of it, and yet the filter cap blew as well. It is really weird seeing as we have 100 kohm between filter cap terminals, assuming the reading is not affected by the lack of fuses (ie, no shorted components downstream from the fuses, or some other shenanigans). Damn shame we do not have good schematics of this part...

 

[Elf]

This is so basic I hate to mention it, but if it was mentioned above it didn't quite jump out to me: installing the replacement capacitor backwards will also guarantee immediate, violent explosion

Anything else gets a bit more esoteric, of course. A misbehaving power factor correction circuit, as mentioned, could certainly cause excessive voltage across it. Excessive ripple current or draw on the capacitor, though, would be unlikely to pop it immediately. It would take a bit to heat to that point where the electrolyte wanted to boil off.

 

[weblacky]

The picture only shows so much, but the square certainly could be large enough and a single screw is often what’s used. You have nothing to loose, try taking that off and look underneath.

I didn’t know about the blown fuses, that wasn’t stated in an obvious way. If you have blown fuses and high voltage...rectifier failure (or just leaky at certain voltage) is very very likely. As a partially working rectifier will output incorrect voltage + higher?

Yeah, Elf, I really think it’s overly high voltage, what else makes a cap overheat and smoke that fast?

 

[llopis]

This is so basic I hate to mention it, but if it was mentioned above it didn't quite jump out to me: installing the replacement capacitor backwards will also guarantee immediate, violent explosion

That would be a good explanation, but unfortunately that's not it. I took a lot of pains to make sure I didn't screw that up. Here's the photographic evidence.
Before:


After:


Yeah, I forgot to explicitly mention then blown fuses earlier. That only happened with the actual explosion. The first smoking didn't blow them up.

I'll check underneath that square metal and see what I find.

Thanks everybody for your help and suggestions. Super helpful!

 

[llopis]

The picture only shows so much, but the square certainly could be large enough and a single screw is often what’s used. You have nothing to loose, try taking that off and look underneath.

I didn’t know about the blown fuses, that wasn’t stated in an obvious way. If you have blown fuses and high voltage...rectifier failure (or just leaky at certain voltage) is very very likely. As a partially working rectifier will output incorrect voltage + higher?

The square component under the heat sink seems to be indeed a rectifier! No markings, but it seems pretty clear from the 4 terminals and the readings.

I checked in diode mode and... exactly what you'd expect. No shorts anywhere, and either no connectivity, or showing 0.45 depending on the direction. So I think the rectifier seems fine!

 

[mamorim01]

Something obvious from my end just as well, but just in case... Keep in mind that since you have no fuses on, when you checked for continuity between terminals in the blown filter cap (ie, to check for shorted components) you are not getting a good picture of what is going on beyond the fuse holders. In other words, there could be a shorted component downstream from the fuses which will not show up while checking from the filter cap because the circuit is open. There has to be something wrong beyond the fuses or they would not have blown, but you may not be able to find it from the filter cap's position until the circuit is closed again. Hope that makes sense...

 

[weblacky]

That's good advice actually, you'd normally want to test paths with completed connections, if you're in investigation-mode.

OK, I'd still check - the NTC, X-Caps, and coils right near the fuse holders. If those short, the X-Caps could bridge hot and neutral (although heir job is not to do that). If your fuse blew, you need to check the transformer windings, and the MOSFET connected to the transformer windings. Basically all the smarts in the world detect things at the low-voltage side and perform shutdowns of the (working) high voltage side, if your fuse blew (not due to rust or water to bridge odd stuff) then you have a problem on the high-voltage side. No question. Though if you fix it, you still need to replace filter caps on LV side anyway.

I'd just shot-gun stuff at this point. Test everything for short (both in Diode mode and Ohm). You could just have like a shorted resistor for all we know. But there aren't that many components on that board.


While I'm trying not to "make-up" equipment you should be using, but don't own....

For something like this (even though this PSU is TERRIBLE to try to work on live due to its construction) where you have problems on the high voltage side and it's not super obvious, I'd have broken out my current-limiting incandescent light bulb assembly to basically attach an AC current limited source to the PSU. Then by using like a 60w or 100w bulb (to start), I'd monitor for heat. I have a couple thermal imagers so I'd both do heat mapping but also some basic voltage checking.

Since you had a brand new cap just pop and smoke, I still think you have VERY high voltage on your voltage doubler. But it might just stupidly be something like a resistor or diode that configures that circuit being shorted or something to provide the wrong input.

Yeah, I'd start shotgunning all the basic parts you can, and eliminating them as suspects. In the end we know damage was done outside of aging (due to smoking and explosions), so you're looking for both again issues and semiconductor or passive faults.

 

[llopis]

All right, I got new fuses today. Popped them in and measured across the cap, and everything was as before. No changes there.

But... remember how I had a shorted transient voltage protection diode that I had removed? Just for testing, I soldered a wire between the terminals where the diode was, tested for continuity across the capacitor vias and... shorted!! So it seems that the shorted diode would explain the capacitor blowing up. But of course, the root cause could be something else...

I will look around at the rest of the components later to see if there's anything else shorted.

In any case, I'm going to be freaking out when it's time to power it up again

 

[mamorim01]

I am not convinced that diode is the only faulty part. When you soldered the wire you were closing a circuit, the fact that you now get continuity at the filter cap may also be a consequence of something else being shorted that is now being detected because the circuit is now complete, if that makes sense. It is important to restore the circuit to the original condition (completing connections as weblacky mentioned above) prior to starting troubleshooting. If you have removed parts or have parts of the circuit unknowingly disabled, it is so very easy to go off track.

I also think the capacitor blew up due to some significant increase in voltage (produced by a faulty part, as others have suggested) and not due to a current surge produced by a short, it kind of looks like two different problems which are probably related. I wish there was some sort of schematic at hand, it would be so much easier to start narrowing down the problem.

Once you have replaced that diode, assume the worst and keep checking before powering up. I would also use a current limiting device similar to what weblacky suggests, you will be working in much safer conditions.

 

[weblacky]

Also since we don’t really know the function of that diode, I’ll point out that usually the use of a TVS diode is often as a protection device where it conducts above a certain voltage. I’ve seen them installed backwards for this reason. So it’s possible that diode suppresses the transformers main winding back emf. Again, I have no proof, but if this diode is the lone fault you find on the high voltage side. It’s possible it’s function was to block/direct to hot ground the reverse emf being released after field collapse from the main transformer. When this happens not only is polarity reversed but it voltage is higher.

So this could be the source of the higher voltage and the loss of the TVS diode (not failing open) exposes the filter caps to very high reverse voltage every time the MOSFET fired.

I mean, it’s possible. I’m not a PSU designer and I’d have believe this is not supposed to happen, so I’m making guesses, but I can tell you that the negative leg of the filter cap has the same potential as ground on the HV side.

This could still be it.

I’d be still recommend you solder a light bulb in place if you aren’t going to build a current limiting setup for first time power up. Also doide and many semiconductor failure is caused by bad caps. So while I can understand the want to startup everything to verify functionality before you do more work. Please don’t try to run the Indigo2 without also replacing all electrolytics on the low voltage side if the PSU as well.

 

[llopis]

As I'm looking around the power supply, one thing that I thought was a voltage regulator was really a thyristor. Which I have to admit I wasn't familiar with them before. I looked them up and I thought this was interesting:

Thyristor - Wikipedia

en.wikipedia.org

They can also be found in power supplies for digital circuits, where they are used as a sort of "enhanced circuit breaker" to prevent a failure in the power supply from damaging downstream components. A thyristor is used in conjunction with a Zener diode attached to its gate, and if the output voltage of the supply rises above the Zener voltage, the thyristor will conduct and short-circuit the power supply output to ground (in general also tripping an upstream breaker or fuse). This kind of protection circuit is known as a crowbar, and has the advantage over a standard circuit breaker or fuse in that it creates a high-conductance path to ground for the damaging supply voltage and potentially for stored energy in the system being powered.

I think that's exactly the configuration we have here. The shorted diode (fancy Zenner) was connected to the gate of this thyristor, which shorted things and caused the fuses to blow.

So things are making sense with the results we got, but I still need to find the root cause of all of this

 

[mamorim01]

As I'm looking around the power supply, one thing that I thought was a voltage regulator was really a thyristor. Which I have to admit I wasn't familiar with them before. I looked them up and I thought this was interesting:

Thyristor - Wikipedia

en.wikipedia.org

I think that's exactly the configuration we have here. The shorted diode (fancy Zenner) was connected to the gate of this thyristor, which shorted things and caused the fuses to blow.

So things are making sense with the results we got, but I still need to find the root cause of all of this

I think this is basically what weblacky was describing, only with a zener instead of a TVS diode. This could have easily blown the fuses but we would still have the underlying problem, tracking down what originated the voltage surge that triggered that fault and also blew the filter cap. Do not disregard weblacky's advice, even if you think you have it nailed down, wiring a bulb on the live wire or something similar will keep things in check.

 

[llopis]

I think this is basically what weblacky was describing, only with a zener instead of a TVS diode. This could have easily blown the fuses but we would still have the underlying problem, tracking down what originated the voltage surge that triggered that fault and also blew the filter cap. Do not disregard weblacky's advice, even if you think you have it nailed down, wiring a bulb on the live wire or something similar will keep things in check.

I'm not disregarding anything. I'm reading everything and taking notes. Just that for now I'm in the phase of examining the board and reporting as I go.

 

[mamorim01]

I'm not disregarding anything. I'm reading everything and taking notes. Just that for now I'm in the phase of examining the board and reporting as I go.

For what it is worth, I think you are on the right track. But you are facing something that blew a brand new 400V rated capacitor in a few seconds, on a largely undocumented board. It is a delicate job to say the least