NEW SEGMENT: EEVsmoke #1

[EEVblog]

Let me know if you like the idea or not

[tautech]

Yeah good Dave, we all learn from someone else's fail.

Translucent cased X rated caps, the cases craze/crack with age and if there's a period of non-powered the humidity gets in and power on....POP. Seen it a few times.

[CaptCrash]

Isn't the missing part from the transistor case laying on the board?
Rather than imploding it looks like it exploded and split the package.

[rollatorwieltje]

Not sure if I still have the high res pics, but this was a APC 3000VA ups that blew up:




The transistors were Fairchild RFP40N10. Note that some of the legs are actually evaporated

[WackyGerman]

I like that .  . You re right . Connecting leds to mains and other stuff would be right in the EEVblowup section 

[tautech]

Isn't the missing part from the transistor case laying on the board?
Rather than imploding it looks like it exploded and split the package.

Quite right.
I was given a 2kW 48-230VAC inverter that came from a marine installation and it had taken a hit of saltwater after a hatch was left open.  At least a dozen of the TO-220 MOSFETS in it looked just like the one Dave shows or even worse.
Must have been a hell of a bang.

Another TO-220 in my bare (not cased) LA battery desulphater suffered some sort of short and at least a 1/3 of the plastic case went somewhere......past my ear. 

[tggzzz]

I skipped through the video until I saw the X2 film capacitor - and thought I recognise that cap, it is in a Tex 24x5 scope. So I rewound a bit and, yes I was right, it is in a 2465.

I replaced exactly that cap in my 2565 last week, but I think mine had rapidly disassembled itself a little more effectively. In my case the identical cap seen in the upper left was even more spectacular - the 1/2W 68ohm series resistor had disassembled itself very effectively, taking the PCB's prepreg with it so the glass fibre is exposed, and depositing a carbon film over a 3"x3" area of the board. The scope continued to work after the magic carbon and smoke escaped, but I didn't tempt fate by leaving it on.

Scope works very nicely since I replaced those caps and the others "hidden" on the other power supply board - which were also crazed.

[bktemp]

Those older X2 caps fail quite often with a bang. It happened to me a couple of times (TV, HP power supply).

Maybe it would be a good idea not only getting the photos but also an explanation why this happened: Was it a design fault, did the user something stupid, or was it due to ageing of some components?

[Arjan Emm]

I  like the new segment. But jeez Dave, it's very obvious  that it exploded, not imploded. The missing bit is right in front of it, with the inprint of the die.

[tautech]

Those older X2 caps fail quite often with a bang. It happened to me a couple of times (TV, HP power supply).

Maybe it would be a good idea not only getting the photos but also an explanation why this happend: Was it a design fault, did the user something stupid, or was it due to ageing of some components?

I've only ever seen those orange/yellow whatever translucent types fail, never the gray or white plastic types so IMO it's the case failing, the material used was not fit for long term useage.

Others experiences ?

[dexters_lab]

Those older X2 caps fail quite often with a bang. It happened to me a couple of times (TV, HP power supply).

Maybe it would be a good idea not only getting the photos but also an explanation why this happened: Was it a design fault, did the user something stupid, or was it due to ageing of some components?

+1 yea i had one go on the power supply for an old (1980s) 8" floppy disk drive

[AmmoJammo]

Here's some smoke for you...



I saw the internals of the amp later, it had melted half the aluminium clamps that held the mosfets to the heatsink....

When you have a battery bank that can supply 1000's of amps, things tend to go wrong

[Ivan7enych]

Maybe it would be a good idea not only getting the photos but also an explanation why this happened: Was it a design fault, did the user something stupid, or was it due to ageing of some components?

Agree.
It's interesting to see why that happens, what should I avoid in my circuit. (I like education style of your videos.)

I think much better would be to reproduce the "magic smoke" on your video. For example make a simplest circuit with a transistor turning on a relay coil (without reverse protection diode), and allow a reverse voltage peak to destroy the transistor...

[Stupid Beard]

Maybe it would be a good idea not only getting the photos but also an explanation why this happened: Was it a design fault, did the user something stupid, or was it due to ageing of some components?

Agree.
It's interesting to see why that happens, what should I avoid in my circuit. (I like education style of your videos.)

I think much better would be to reproduce the "magic smoke" on your video. For example make a simplest circuit with a transistor turning on a relay coil (without reverse protection diode), and allow a reverse voltage peak to destroy the transistor...

I agree with that too. I think that it's an interesting idea for a segment, but if it's just looking at photos then I think it's more suited to a forum thread.

[SNGLinks]

My HP8656B sig gen IEC mains filter. You better have the correct fuse in your mains feed when this burns up as it's before the fuse.

[jonovid]

Quote from: Ivan7enych on Today at 08:43:18 PM

Quote from: bktemp on Today at 07:01:45 PM
Maybe it would be a good idea not only getting the photos but also an explanation why this happened: Was it a design fault, did the user something stupid, or was it due to ageing of some components?

Agree.
It's interesting to see why that happens, what should I avoid in my circuit. (I like education style of your videos.)

I think much better would be to reproduce the "magic smoke" on your video. For example make a simplest circuit with a transistor turning on a relay coil (without reverse protection diode), and allow a reverse voltage peak to destroy the transistor...

I agree with that too. I think that it's an interesting idea for a segment, but if it's just looking at photos then I think it's more suited to a forum thread.

maybe some on screen video of Stress Testing of Dodgy Components.  as  like I got a batch of TO92's from China. As the pins unknown to Me was BASE, EMITTER , COLLECTOR   not COLLECTOR ,BASE in the middle  then EMITTER .

[setq]

Surprised they weren't base, base, smoke emitter.

Good section. I like it!

X2 film caps I don't like. My old Telequipment scope blew one a few years ago and filled my bench corner with smoke. They go out in style and cause much worry. Fortunately easy to fix!

[bktemp]

maybe some on screen video of Stress Testing of Dodgy Components.

Most semiconductors fail short circuit without any sparks, so it will be quite boring.
I don't think we need another Youtube channel blowing up circuits or components. Photonicinduction already did a great job in applying to much voltage to almost everything until it pops. Mehdi Sadaghdar (ElectroBOOM) also has sparks on most of his videos. They are always fun to watch.

Explaining why it did happen and how to avoid it is probably more adequate for Dave's typical audience, because they are interested in learning things.

The most spectacular failures are probably in mains powered circuits, because there you have large caps after the rectifier charged to 160V or 320V. Those caps can deliver many 10 or 100kW for a short period of time and vaporize almost anything.

[_Andrew_]

When it comes to smoke it has to be:   Lucas the Prince Of Darkens

This sums up the theory of smoke quite well http://hdabob.com/the-vehicle/electrical-system/smoke-theory/

[Don Hills]

Fortunately, you can get Lucas smoke replacement kits:

http://www3.telus.net/bc_triumph_registry/smoke.htm

[setq]

More Lucas thoughts: http://www4.ncsu.edu/~mtmorris/index3.html



I actually incinerated the entire ignition system on my Land Rover Series III back in the 1990s thanks to Lucas. Now that was smoke - clouds of it as the battery dumped through the entire shorted loom!

[G7PSK]

I have had many vehicles with Lucas electrics and mechanical parts, never had a problem with them now wico that is another story.

[Len]

You don't see this very often

I've seen it several times on this very forum. It's a Rifa brand capacitor, isn't it?

[maelli]

Nice segment, we want more of this!
Regarding the Rifa PME271, these are a classic for failures. The potting material appers to be too brittle, cracks form, moisture gets in.
The metalised paper winding starts increasing capacitance (good?) and increasing loss factor (sure not good).
When connected to the mains, things get ugly.
The cap on the picture has not seen voltage for years, it will go up in smoke within seconds or minutes if connected to mains ;-) should I try?

[madires]

You don't see this very often

I've seen it several times on this very forum. It's a Rifa brand capacitor, isn't it?

Not just on this forum. I've seen several cracked X2 Rifas myself, luckily none completely vaporized like in some older IEC line filters. If I see any Rifa cap, I feel the sudden urge to replace it with a Wima.

[f4eru]

Here's a classic one :

[Towger]

You don't see this very often

I've seen it several times on this very forum. It's a Rifa brand capacitor, isn't it?

Not just on this forum. I've seen several cracked X2 Rifas myself, luckily none completely vaporized like in some older IEC line filters.

They crack over time, I have a few unused ones I got about ~30 years ago and they have all split.

[bktemp]

The cap on the picture has not seen voltage for years, it will go up in smoke within seconds or minutes if connected to mains ;-) should I try?

You could make a video, or send it to Dave, so he can make a video using his Sony RX100 IV camera to record the cap blowing up at 1000fps.

[gardner]

record the cap blowing up at 1000fps.

Maybe ask Mike to make a guest appearance on the Slow Mo Guys with his Destruct-O-Tron (tm).

[jaromir]

Let me know if you like the idea or not

Though I love most of your other videos, I'm not sure about this segment.
I agree with some other folks suggesting to simply keep the photos of failed components in forum instead of doing commented slideshow. On the other hand, discussing failure modes in real life applications (that's what you did) has more value than deliberately destructing components just for video, as some other commenters suggested.

[TheWelly888]

A 30 year old Philips analogue scope (PM something) at work released the magic smoke whilst being used on my colleague's bench. The cloud hazed the workshop nicely and I was worried about the smoke detectors not setting off the fire alarm until I realised they were heat detectors!

The magic smoke was released from one of those X2 caps across the mains input.

[richnormand]

@ Dave about the video:

How about expanding to look deeper into root cause, failure analysis, x-ray perhaps, some measurements of adjacent circuits, etc...
The idea would be to learn and figure out what happened.

That is the point that left me meh... about it.
Sorry, you did ask for reactions....

[EEVblog]

How about expanding to look deeper into root cause, failure analysis, x-ray perhaps, some measurements of adjacent circuits, etc...
The idea would be to learn and figure out what happened.

Kinda impossible to do with just a photo, and not easy even if you have the board.

[rx8pilot]

I have smoked many PCB's, MOSFETS, DIODES, and other parts in the process of learning power electronics. Some of those events were a total mystery because the even unfolds so fast. It can take considerable amounts of time to figure out what went wrong or at least how to measure the system in s useful way just before the smoke escapes.

I would like to see the effect followed by cause if at all possible. Burnouts are an excellent learning experience and great discussion topic. For photos that I could send - I could also include the analysis of what happened or at least the best guess.

[Tinkerer]

Recently we had a failure under testing.

A 70V 8A rated diode went poof; it was sitting between VDC In and return. What caused this? Well, it was sitting at 35V for 30 min and then the voltage was dropped to 20V; note this was not with a constant current input. After about 15 seconds, it went poof.
Everyone is rather baffled by this because there isnt a reason this part should have died. The current thoughts are wrong part, defective part, or I think human error may have been ruled out at this point.
Although, it was hit by a 600volt transient in testing before this time. However one of the senior guys had already done thorough testing to prove the part could survive that so we are still baffled.

[meeko]

No pictures, unfortunately, but back when I used to work at a computer store, in the service department, we had a PSU go bang.  The timing could hardly have been worse, either:  right as I turned the PC on to show the customer it was fixed!  Needless to say, the customer wasn't too impressed.  Of course, if it hadn't blown up then, it probably would have done soon, so it was probably for the best, really.

I swapped the PSU on the spot for him, and noticed that it rattled when I took it out.  After he left, I took it apart, and found that the seal on one of the output caps had failed catastrophically, and it was the perfectly intact can that was rattling around the PSU casing.  Even the foil was still intact!

[jonovid]

Quote from: richnormand on Today at 10:26:41 AM
How about expanding to look deeper into root cause, failure analysis, x-ray perhaps, some measurements of adjacent circuits, etc...
The idea would be to learn and figure out what happened.

Kinda impossible to do with just a photo, and not easy even if you have the board.

Your most helpful failure analysis video .                      EEVblog #735 - Mailbag  stop the video at 40:00 then play it to 50:50   the total video time = 54:22
most helpful LED lamp teardown.  I did use this EEVblog video to help avoid some unseen but common pitfalls in my LED lighting designs.
why you need a good thermal image camera in the lab,  if the budget will allow this. 

[Brumby]

Certainly, it is interesting, but looking at the first EEVsmoke video, I can't say it stands out as being viable for it's own channel - at least not with the level of detail offered in this example.  However, stepping it up is no simple exercise.

As you have indicated, Dave, having it being driven by photographic contributions does limit what you can do and this is going to hamper the interest and usefulness to a fairly low level, IMO.  Fault analysis and root cause identification would make it better - but that is really impractical if not impossible.  Even researching other examples of the failure of the component under examination is going to demand a lot of time and effort - and then you would have to present that in the video.

If anything, I might suggest you save up a number of EEVsmoke candidates and then, when you have enough for a production, do a standard blog video with 'EEVsmoke' in the title.  You can then gauge reaction and interest as to whether there is value in taking it further - or dropping it altogether.

I think it's a bit early to make the call after just one video.

[Brumby]

Having said that - I've just had a look at one of the 1500VA Line Interactive UPS's I picked up from the Jaycar warehouse clearance.

I checked out the SLA batteries in the remote hope they might have some life in them - but less than half a volt open circuit isn't too encouraging.  Connected a power supply - 12V limited to 100mA and the terminal voltage jumped instantly to 12V, then dropped back to under a volt in under 2 seconds when disconnected.

However, the first thing that greeted me when I opened the case was this:
(I've brightened up the image so you can see a bit more detail)


Don't think I'll be trying to resurrect the UPS ... but the transformer at the bottom of the case has me thinking....

It's big and very heavy.

[JoeN]

I liked the vid.  I don't see why it needs to be on a different channel.  It could easily be thrown in with the other vids on the main channel.  That's just one man's opinion.

[SeanB]

How about this, on a 2kVA UPS in scrapped last week. It still worked with the slightly charred board and the diode on each half of the output stage sitting on the base of the cabinet.

[bitwelder]

This new EEVsmoke section as it is on its first episode is very much 'meh' quality.
I don't find it particularly entertaining to look at pictures of failed components, especially if there is no analysis or explanation of why that it happened.
As somebody commented already, it could perhaps be working better e.g. on the forum, where other people can comment and post similar failures as what Dave showed, but as video... Meh!

[Fungus]

What's the fun in pictures of dead components? Where's the video.

[Towger]

What's the fun in pictures of dead components? Where's the video.

None.  To put bread on the table it needs to show exploding components and have catchy key words in the title. EEVSmoke means nothing to the great unwashed.   Along the line of Phontonic's or the Press channel (complete with cackling wife in the background).  Dave, I can send you some NOS of those X2 caps, complete with cracked cases. Just apply power, if 240v does not release the magic, I am sure you can rig up something with more umph.

[Cyberdragon]

I don't think he's into blowing things up, he said he only wanted "naturally released smoke". Even if he was it might set off the fire alarm in the building.

[tautech]

Dave, I can send you some NOS of those X2 caps, complete with cracked cases. Just apply power, if 240v does not release the magic, I am sure you can rig up something with more umph.

I don't think he's into blowing things up, he said he only wanted "naturally released smoke". Even if he was it might set off the fire alarm in the building.

Maybe, but it'd be a good followup to EEVsmoke #1 demonstrating an actual failure of this type of X2 cap. Dissect a cracked one looking at the breached case and apply power to another showing the end result.
If there's not a vid of this already on the net, then it might be the first. 

[Bud]

Go for it, i will be watching this segment.

[XynxNet]

Not my kind of segment without failure analysis.
But I don't have to watch it. So go for it. 45k views indicate a lot of people like it.

[gardner]

My feeling is also "meh".  I think the main problem is working from photos.  Without being able to measure voltages and currents and so forth and do real analysis of a failed circuit, it just amounts to 15 minutes of "gosh, look at that!" which admittedly, Dave can pull off reasonably entertainingly.  But what you can get from a photo to really expound on usefully is just too limited.  I was probably about the 200th person to shout at Dave "that TO220 is not "emploded" -- you can see the front of the package laying on the board right in front of it!"  Working strictly from photos is just going to be more of the same.

[Cerebus]

I, for one, definitely want to see more.

Two motivations:

1) Seeing and understanding failure modes.

2) What bloke doesn't like seeing things go BOOM!

[Cerebus]

My HP8656B sig gen IEC mains filter. You better have the correct fuse in your mains feed when this burns up as it's before the fuse.

Not the first Schaffner filter I've seen that happen to. There's a period - I forget exactly when - where the caps they fitted to those aged badly, even on the shelf. Recently, I had one that had been on the shelf for years go up about 15 minutes after power was applied. As I was sitting next to it at the time I got a face full of capacitor smoke, the shit scared out of me and clothes that still smelt evil after a couple of washes. It's still on the back step waiting to be autopsied once the smell has faded a bit.

[EEVblog]

I got 50 emails with new submissions!, please no more for now.

[Cerebus]

I got 50 emails with new submissions!, please no more for now.

Aww, I was going to conduct an 'autopsy' that I'd been putting off. Just for you.

If you've got a big pile of them, collecting similar failures together could be quite instructive. Either similar components with a range of failure modes, or alternatively similar failure modes across dissimilar components - scratch that, the latter sounds a bit dumb.

It's already been quite instructive that Rifa X class caps have had a lot of discussion in this thread.

[madires]

Not the first Schaffner filter I've seen that happen to. There's a period - I forget exactly when - where the caps they fitted to those aged badly, even on the shelf. Recently, I had one that had been on the shelf for years go up about 15 minutes after power was applied. As I was sitting next to it at the time I got a face full of capacitor smoke, the shit scared out of me and clothes that still smelt evil after a couple of washes. It's still on the back step waiting to be autopsied once the smell has faded a bit.

Congratulations! You've found a Rifa cap

[SeanB]

Rifa, Real Intense Fire Action.

[max_torque]

"exploded" IGBT, after the soot has been washed off:




Lots of melty carnage on the Gate/Source boundary! 

[ion]

However, the first thing that greeted me when I opened the case was this:
(I've brightened up the image so you can see a bit more detail)


Don't think I'll be trying to resurrect the UPS ...

That looks like my UPS did a few months ago (something went wrong while probing an intermittent fault).  After the flames died down I cleaned it up, replaced the MOSFETs and a shorted diode and it came back to life - so even looking like that, it might not be that hard to repair.  Pity I didn't take a picture of my unit at the time.

[tautech]

However, the first thing that greeted me when I opened the case was this:
(I've brightened up the image so you can see a bit more detail)


Don't think I'll be trying to resurrect the UPS ...

That looks like my UPS did a few months ago (something went wrong while probing an intermittent fault).  After the flames died down I cleaned it up, replaced the MOSFETs and a shorted diode and it came back to life - so even looking like that, it might not be that hard to repair.  Pity I didn't take a picture of my unit at the time.

The UPS or your underpants ? 

[larrybl]

Speaking of UPS's, Batteries having a bad day. 

[Cerebus]

Speaking of UPS's, Batteries having a bad day. 

Ooo, that is nasty. It's bad enough having to clean up any mess, but when that mess contains sulphuric acid and dissolved lead salts it really is a bad day.

[jonovid]

Quote from: larrybl on Today at 01:02:45 AM

Speaking of UPS's, Batteries having a bad day. 

Ooo, that is nasty. It's bad enough having to clean up any mess, but when that mess contains sulphuric acid and dissolved lead salts it really is a bad day.

Hydrogen .......... Boom!

[Cerebus]

Quote from: larrybl on Today at 01:02:45 AM

Speaking of UPS's, Batteries having a bad day. 

Ooo, that is nasty. It's bad enough having to clean up any mess, but when that mess contains sulphuric acid and dissolved lead salts it really is a bad day.

Hydrogen .......... Boom!

If they're in a submarine, Chlorine. Even worse...

[fable]

HV transformer from crt tv

[Kfj]

What happens when you don't pay attention to the voltage rating on an electrolytic capacitor

[mcs_5]

Regarding the Rifa PME271, these are a classic for failures. The potting material appers to be too brittle, cracks form, moisture gets in.
The metalised paper winding starts increasing capacitance (good?) and increasing loss factor (sure not good).
When connected to the mains, things get ugly.

Not only Rifa/Kemet have this problem - Wima MP3 caps do the same thing. They release a lot of white smoke, a terrible smell, and spew brown stuff on the board area next to them.

[EPTech]

"exploded" IGBT, after the soot has been washed off:




Lots of melty carnage on the Gate/Source boundary!

Hi there,

I never cleaned of the soot of one of those, not to mention the sticky gel that comes with it . very interesting. What type of IGBT was it? How many amps?
Are you sure it is the gate source that failed? Looks more like the source drain to me. The gate usaully has the smallest surface on an IGBT. It could be the structure to the left side of your picture though I usually see the gate in the middle of an IGBT. Also, gate source failures are usually more subtle. Sometimes they can be seen as a very faint darker line or spot escaping from the gate structure.

I have a couple of failed power modules lying around. I tried taking some pictures but I am sorry to say I do not have the equipment to take that kind of a macro shot. Your picture is amazing. Thanks for sharing.

[EEVblog]

Not my kind of segment without failure analysis.
But I don't have to watch it. So go for it. 45k views indicate a lot of people like it.

75k views now and only 10% thumbs down, so more than enough to do another one.
Another one should show the true interest. Lots are going to click on the first one just to see what's going on.

[SeanB]

"exploded" IGBT, after the soot has been washed off:




Lots of melty carnage on the Gate/Source boundary!

Hi there,

I never cleaned of the soot of one of those, not to mention the sticky gel that comes with it . very interesting. What type of IGBT was it? How many amps?
Are you sure it is the gate source that failed? Looks more like the source drain to me. The gate usaully has the smallest surface on an IGBT. It could be the structure to the left side of your picture though I usually see the gate in the middle of an IGBT. Also, gate source failures are usually more subtle. Sometimes they can be seen as a very faint darker line or spot escaping from the gate structure.

I have a couple of failed power modules lying around. I tried taking some pictures but I am sorry to say I do not have the equipment to take that kind of a macro shot. Your picture is amazing. Thanks for sharing.

Yes, more like the integrated PNP power transistor itself had a SOA failure. IGBT devices are funny, the only place since they stopped making commercial Ge devices where you find a power PNP device.

http://www.ixys.com/Documents/AppNotes/IXYS_IGBT_Basic_I.pdf

[Cerebus]

Yes, more like the integrated PNP power transistor itself had a SOA failure. IGBT devices are funny, the only place since they stopped making commercial Ge devices where you find a power PNP device.

I'm sure that can't be what you meant to say. There are an awful lot of PNP power transistors in the world.

[SeanB]

Yes, more like the integrated PNP power transistor itself had a SOA failure. IGBT devices are funny, the only place since they stopped making commercial Ge devices where you find a power PNP device.

I'm sure that can't be what you meant to say. There are an awful lot of PNP power transistors in the world.

True there are power PNP devices, but almost all SMPS applications use 2 NPN power devices, and there are no PNP transistors designed for line output switch application as far as I know. My bet was if there ever was one it would have been in a Sony TV set, as a complement to those lovely GCS thyristors that they were enamoured of for a decade or three.

[Stupid Beard]

Not my kind of segment without failure analysis.
But I don't have to watch it. So go for it. 45k views indicate a lot of people like it.

75k views now and only 10% thumbs down, so more than enough to do another one.
Another one should show the true interest. Lots are going to click on the first one just to see what's going on.

For what it's worth, I watched the second one to see if my opinion would be different. I still think that it's a subject that's better suited to a forum thread than a video. I did watch the whole of it, but for the last couple of minutes I found myself switching to another tab and reading other stuff. I don't think I'll continue watching eevsmoke unless I'm bored and there's nothing else to watch.

I also did not thumbs down either video, because giving a thumbs down just due to my personal preference seems like a shit reason to thumbs down a video.

One way I think that this could work in a video format would be as part of a small segment in a regular larger segmented format show type thing. But that would be more work and is not really the way you like to film things so maybe that's a crap suggestion.

[WackyGerman]

Here we see the control board of a led bulb after a delayed new years eve bang

[Godzil]

Is this the bigdavedotcom channel???

[tpowell1830]

I think this would be boring unless there is video of the actual event.
i have seen plenty of burned components in my career.

PEACE===>T

[Brumby]

Burned is boring.

What works here is the exceptional - detonations, capturing magic smoke and other intriguing failures.

[Godzil]

By the way, the X2 capacitor is a paper based capacitor, and as it dry with time they all fail in a similar manner. All of the failing one nowadays are form device from the 80s or really early 90s. It's quite common in vintage Thomson TO7/MO5 monitor to have the X2 filter capacitor to fail like this, my Apple //c monitor had his X2 filter failing in such a way too, it's not critical as it's just a filter cap, but I used that as an occasion to replace all the electrolytic capacitors with newer one and replace the faulty X2 with a non paper X2 one, and another paper caps with a ceramic one.

It works way better now

I still have the failed X2, I should take a picture of it, it did some really nice disgusting magic black things©

(it is that sort of caps: http://uk.farnell.com/kemet/pme271m547mr04/cap-film-paper-47nf-275vac-rad/dp/2332428 )


Dave: also does huge battery failure would fill that segment? (I have a board where a NiCd battery failed in a really uncontrolled manner, it's like if a nuke happened in the device, that's absolutely impressive)

[akos_nemeth]

Washing machine drama

The ST ACS102-6T "Transient protected AC switch"(http://www.st.com/content/ccc/resource/technical/document/datasheet/64/e8/a1/ac/ad/58/41/e2/CD00083453.pdf/files/CD00083453.pdf/jcr:content/translations/en.CD00083453.pdf) on the control panel PCB of an Electrolux front loader washing machine (EWM 2500 platform http://www.4mcculloch.co.uk/images/mediator/559/599527452.pdf) was blown up probably due to a short  around the door interlock (PTC bimetal switch?).

The trace on the PCB between pin 2 of the ACS102-6T (OUT) and R3 is misteriously evaporated (why only this section?). Another interesting feature of this PCB is the rectangular fiducial-like marking (what could be a purpose of it?). Capacitor C3 is not populated on the PCB but there is soldering on the pads, maybe the designer decided not to use it, but the stencil was already made?

Regards,
Ákos

[VinzC]

The story of a mishap...

Here's one of my projects at work, of which I blew up the MCU quite stupidly. The board is a motor controlling unit that uses stall currents to stop the controlled motors. It drives up to 2 motors and has 4 SPST relay outputs. Motors are powered by an IC that I re-used from my predecessors' project. The difference is my board uses miniature OMRON H-bridge relays instead. The absolute maximum current is 23 amps, as per MEGA-FIT connector specs.

The microcontroller is an Atmel ATmega64M1 as shown on the picture. The firmware has several configuration options such as inrush delay, power on delay and threshold (stall) current. There's also a stall delay that is zero by default. The latter delay allows a motor to run on even though the stall current was reached. The configuration can be changed by plugging a USB cable and connecting it to any computer with a VT100 terminal emulator (that's the purpose of the MCP2221).

When motor currents fall within the specs even when stalled, all is well. All motors my unit drives stall at no more than 20 amps... with one exception: our sunroof motor. The latter is a rather old unit that stalls at more than 30 amps! As I didn't have time to upgrade the board for stronger currents, I decided to give it a go.

Unfortunately the rooftop didn't close even with a current limit of 20 amps. I don't know what I had in mind that day but I enabled the stall delay and set it to 1 second... I lost all communications with the unit all of a sudden. I took the board off the car and plugged it to my test bench and then... smoke!

I quickly realized the VND600, which is rated for 25 amps maximum was blown without a hint, which incidentally caused its inputs to be shorted to ground. Had I driven the VND600 inputs with the MCU internal pull-ups instead of setting the outputs directly, I would have spared the microcontroller.

That experience also stresses on the fact you need to be extra careful also designing the firmware as well. I didn't use intermediate resistors to drive the VND600 here, which would not have been a problem if I had enabled the MCU port pull-up circuitry. Also the now obsolete VND600 has no current limitation internals, which modern drivers, like the VND5xx series have.

Call it a lesson learnt ;-) .

[bktemp]

Had I driven the VND600 inputs by with the MCU internal pull-ups instead of setting the outputs directly, I would have spared the microcontroller.

I doubt it was the outputs driving into a short. The output transistors of typical microcontrollers are too weak to let any significant current flow.
More likely when the VND600 failed it put the motor supply voltage onto its input pins and also into the AVR.
Only adding series resistors could have saved the microcontroller.

[VinzC]

More likely when the VND600 failed it put the motor supply voltage onto its input pins and also into the AVR.

That, I can check, thanks for the hint.