Author Topic: EEVblog #1036 - PSU Fire PCB Repair  (Read 20088 times)

0 Members and 1 Guest are viewing this topic.

Offline ChrisLX200

  • Supporter
  • ****
  • Posts: 458
  • Country: gb
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #25 on: November 07, 2017, 09:28:52 am »
Quote from: Brumby on Today at 10:59:31 AM>Quote from: ChrisLX200 on Today at 05:05:39 AM
Given the modest cost of these things I would not have attempted a repair anyway and just binned it.

A lot of people looking at a low cost solution like these modules is not going to be so cavalier.  They will see a problem like this and want to repair the module, if at all reasonably possible.

By having one of these units "blow up", Dave has done the manufacturer a favour.  He has caught the attention of a lot of people who view such cheap modules with suspicion and has activated their "AHA!!" button.

By then successfully pursuing the repair, he has demonstrated the basic robustness of the design - other than the poor placement of a single MLCC cap (which only occurs on the 50V 20A version).  The response from the designer/manufacturer has only added to the positive vibe about them.

The repair itself is another plus.  It was very simple (for this fault) and could be done by anyone with a modicum of skill.


All these things add up to a more positive view of these modules.  Sure, something went wrong - but by having something go wrong, we have been able to see the Achilles' heel we so often suspect is lurking in low cost gear - and we can fix it.  Better still, we can prevent it.

Tell me you don't have a more positive attitude about these modules because of all this.....




You missed my point from your quote - the person interested in a low cost solution is also entitled to a free replacement so does not need to repair anything. There's nothing cavalier about that, you don't really know what other components have been stressed by this event. But sure, with a replacement in hand, then they are free to do whatever they like with the burned unit. Would you really have a positive attitude about walking away from one of these units and leaving it powered up?

I actually own one of the cheaper units in this series but seeing this one self-destruct doesn't make me feel warm and fuzzy inside..
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12806
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #26 on: November 07, 2017, 09:52:47 am »
The damage was localised to the area around the failed cap.  The fire did not spread to the rest of the PCB.  I have seen similar damaged PCBs from reputable manufacturers of consumer electronics sold in the EU, with similar localoisation of the damage, so its is extremely unlikely that the PCB laminate was made with a non-self-extinguishing resin.  Whether or not it meets EU and US standards for PCB flammability would require testing, but I think we can assume that if it does not, it isn't far short of meeting them.

Would I be happy to have one of these specific board version PSUs (assuming I had a suitable raw DC supply to run it from)?  Yes, but the first thing I'd do would be to remove the offending cap and fit a suitable low ESR electrolytic across the output terminals.  Its obviously not got the build quality of a comparable lab grade PSU from a leading test equipment manufacturer, but as its an order of magnitude cheaper, seems reasonably robust apart from an easily rectified design flaw, and has the 'grunt' you need if you are doing stuff like robotics on a budget, IMHO its still a good deal.  If the manufacturer did a recall offering exchange or partial credit if you remove the cap yourself, and also reworked existing stock to remove the ceramic cap and supplied them with a suitable leaded cap + fitting instructions, it would be an entirely satisfactory resolution to the issue.

Would I repair it?  Yes, if the manufacturer didn't want the failed board back or if it was out of warrenty.   Depending on the extent of the damage, once I'd confirmed it was savable,  I might choose to move the fusing off-board - depopulate and link out the SMD fuses then fit an inline holder and a 15A fuse.
« Last Edit: November 07, 2017, 10:05:15 am by Ian.M »
 
The following users thanked this post: thm_w, Freelander

Offline Freelander

  • Regular Contributor
  • *
  • Posts: 146
  • Country: 00
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #27 on: November 07, 2017, 12:58:30 pm »
The damage was localised to the area around the failed cap.  The fire did not spread to the rest of the PCB.  I have seen similar damaged PCBs from reputable manufacturers of consumer electronics sold in the EU, with similar localoisation of the damage, so its is extremely unlikely that the PCB laminate was made with a non-self-extinguishing resin.  Whether or not it meets EU and US standards for PCB flammability would require testing, but I think we can assume that if it does not, it isn't far short of meeting them.

Would I be happy to have one of these specific board version PSUs (assuming I had a suitable raw DC supply to run it from)?  Yes, but the first thing I'd do would be to remove the offending cap and fit a suitable low ESR electrolytic across the output terminals.  Its obviously not got the build quality of a comparable lab grade PSU from a leading test equipment manufacturer, but as its an order of magnitude cheaper, seems reasonably robust apart from an easily rectified design flaw, and has the 'grunt' you need if you are doing stuff like robotics on a budget, IMHO its still a good deal.  If the manufacturer did a recall offering exchange or partial credit if you remove the cap yourself, and also reworked existing stock to remove the ceramic cap and supplied them with a suitable leaded cap + fitting instructions, it would be an entirely satisfactory resolution to the issue.

Would I repair it?  Yes, if the manufacturer didn't want the failed board back or if it was out of warrenty.   Depending on the extent of the damage, once I'd confirmed it was savable,  I might choose to move the fusing off-board - depopulate and link out the SMD fuses then fit an inline holder and a 15A fuse.

Agreed ! - the other point that people have happily and conveniently failed to mention is that the 'supply' - for the DPS5020 was NOT fused!!!. it was constant current limited. There is a major difference. Anyone building a unit such as this should indeed have a fuse in the supply that feeds to the DPS.  To not have such protection mechanism is simply poor practice and asking for trouble.
Had such an absolutely normal, safe and standard practice been in place during the tests then there would have been absolutely no drama at all. Often it is not the gun it is the gunner!
In normal use the fuse that SHOULD have been fitted in the supply to the unit would have done its job.
In any other 'experiment; with a similar result I would expect people to jump on the fact that the setup was totally wrong in the first place. The drama ONLY 'happened'  due to poor practice -it was not 'CAUSED' by the DPS unit. the fault was of the DPS, yes, but -NOT the 'DRAMA'. Much ado about nothing and blown out of proportion by lack of standard safeguards and practice. (all IMHO of course) ;)

edit - cannot splelllll ...
« Last Edit: November 07, 2017, 01:05:13 pm by Freelander »
 

Online Ian.M

  • Super Contributor
  • ***
  • Posts: 12806
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #28 on: November 07, 2017, 02:15:25 pm »
Although the input *should* be fused, it wouldn't have stopped any of the drama if the DPS5020 was suitably fused to permit its full specification to be utilised.  Once the cap failed and it went into current limiting, I would expect that the output voltage collapsed considerably, limiting the power.  This would have happened quickly enough that a 20A fuse on the input would have been unlikely to blow. Any lesser fuse woudn't permit full output current at full voltage.

Edit: thanks Brumby for reminding me the module is capable of 20A output, so needs a 20A input fuse when running at max output voltage with minimum headroom.
« Last Edit: November 07, 2017, 02:57:14 pm by Ian.M »
 
The following users thanked this post: Freelander

Offline Brumby

  • Supporter
  • ****
  • Posts: 12288
  • Country: au
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #29 on: November 07, 2017, 02:45:38 pm »
Although the input *should* be fused, it wouldn't have stopped any of the drama if the DPS5020 was suitably fused to permit its full specification to be utilised.
This, I thought, was somewhat obvious and critiquing the lack of a fuse is, essentially, irrelevant.

The module was rated to deliver 50V @ 20A.  Put any fusing to allow operation in that range and there would still have been a magic smoke show.  1000W is several times the power necessary to cook that PCB.
 
The following users thanked this post: Freelander

Offline thm_w

  • Super Contributor
  • ***
  • Posts: 6272
  • Country: ca
  • Non-expert
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #30 on: November 07, 2017, 08:11:57 pm »
Thanks for the follow-up!~ Was originally going to get one of these; but now I'm not so sure.

Mistakes happen but what made me :palm: was when RD Tech admitted that this isn't the first time they've encountered precisely this method of failure but yet never bothered to correct the issue until now.

If you have hundreds or thousands of sales, and a few fail, its generally not worth looking into. Of course if you have free time go for it. I can tell you for my job unless I see 5+ failures of the same type, its not worth the time to investigate the fault.

They may have also not got the PSU's back, as shipping to china is expensive.
Profile -> Modify profile -> Look and Layout ->  Don't show users' signatures
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14072
  • Country: de
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #31 on: November 07, 2017, 09:06:38 pm »
If they don't get the board back, they probably get at least a picture back.  Looking at the burnt board and the new one should make it obvious what happened.

With a product sold to hobby customers, there is a chance that some of the reported failures are not due to a problem in the product. With just a few failures it may not be worth the investigation.
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12288
  • Country: au
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #32 on: November 07, 2017, 11:26:33 pm »
With a product sold to hobby customers, there is a chance that some of the reported failures are not due to a problem in the product. With just a few failures it may not be worth the investigation.
This was the thought I often have with "reported failures".

In this case, Dave has some credentials - plus video - to back up his experience.  The manufacturer would certainly respond to this - which is exactly what happened.
 

Offline rymannphilippe

  • Newbie
  • Posts: 4
  • Country: ch
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #33 on: November 08, 2017, 06:38:38 am »
Hello all

I will order this DPS5020 and found it on Aliexpress.
Dave's Board Version is V2.3 ( 2:10Minutes Position).

The one i found on Aliexpress is V2.2 on the backside without the CAP.
Is this a good or a bad idea? ;-)

Anybody know some more news, maybe about V2.4 (new design)?

Greetings from Swiss
Philippe
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #34 on: November 08, 2017, 08:17:27 am »
I will order this DPS5020

Dave's Board Version is V2.3.

The one i found on Aliexpress is V2.2

Anybody know some more news, maybe about V2.4 (new design)?

This video is only 2 days old and you want the next revision today? While the sellers are still on the previous revision?  :o
 

Online T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #35 on: November 08, 2017, 08:47:26 am »
Why put fuses at the end, at all?  It's current limited, isn't it?  I hope it's current mode, anyway... that TL598 isn't a good sign though...

If it's for short protection (pass device fails), the fuse should be at the input, where a bigass capacitor or battery can supply clearing fault current.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline rymannphilippe

  • Newbie
  • Posts: 4
  • Country: ch
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #36 on: November 08, 2017, 09:19:46 am »
Dear Fungus

You are right  :-DD
Sorry I'm stuck in my business and we are only do some emergency manufacuring and 24h is nothing.... but you are right.

By the way:
I have some contact with RD STORE and the told me, THE CAP IS CHANGED.
 

Online tszaboo

  • Super Contributor
  • ***
  • Posts: 7304
  • Country: nl
  • Current job: ATEX product design
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #37 on: November 08, 2017, 10:29:24 am »
Why put fuses at the end, at all?  It's current limited, isn't it?  I hope it's current mode, anyway... that TL598 isn't a good sign though...

If it's for short protection (pass device fails), the fuse should be at the input, where a bigass capacitor or battery can supply clearing fault current.

Tim
The pass transistor can fail short. Also, people can use this to charge batteries so in case a failure, power can be pumped in from the output terminal. Imagine a 24V li-ion battery pack, with 10mOhm resistance charged.
When I was designing battery testing system, we had a fuse for every single pass transistor.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #38 on: November 08, 2017, 02:02:09 pm »
You are right  :-DD
Sorry I'm stuck in my business and we are only do some emergency manufacuring and 24h is nothing.... but you are right.

Less than 24h to manufacture a product revision and destroy all the old ones? That must be expensive.

I have some contact with RD STORE and the told me, THE CAP IS CHANGED.

Maybe you can work something out to get one of the new ones.

Or... get one of the old ones and substitute an old-fashioned cap. One with long, flexible metal legs.  :popcorn:
 

Offline Freelander

  • Regular Contributor
  • *
  • Posts: 146
  • Country: 00
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #39 on: November 08, 2017, 05:08:34 pm »
You are right  :-DD
Sorry I'm stuck in my business and we are only do some emergency manufacuring and 24h is nothing.... but you are right.

Less than 24h to manufacture a product revision and destroy all the old ones? That must be expensive.

I have some contact with RD STORE and the told me, THE CAP IS CHANGED.

Maybe you can work something out to get one of the new ones.

Or... get one of the old ones and substitute an old-fashioned cap. One with long, flexible metal legs.  :popcorn:

The version for sale now HAS the cap with old fashioned 'bendy' legs ;). Simple replacement. All sorted. - Also,the cap is only (and was only) fitted on the DPS5020/C I believe.
Sorted. :)
 

Online paulca

  • Super Contributor
  • ***
  • Posts: 3994
  • Country: gb
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #40 on: November 25, 2017, 09:50:04 am »
I just bought one of these.  For some reason I assumed the video was older and a new revision would be out.

However RDTech did say they had stopped selling the one Dave tested, but yet AliExpress's "official" RDTech store is selling them.

I am hoping to run it with a 48V 10A DC power supply (https://www.ebay.co.uk/itm/282556976292).  I don't mind that I won't get the full 50V 20A.

So, when it arrives how would I check the board is safe?

For protection should I fuse the DC supply input to the DPS5020, or would I be better with a reset-able breaker to prevent overloading the 10A supply?  So if I do try and draw more then the 48V/10A the breaker will pop.

"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Fungus

  • Super Contributor
  • ***
  • Posts: 16560
  • Country: 00
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #41 on: November 25, 2017, 10:14:50 am »
I just bought one of these.  For some reason I assumed the video was older and a new revision would be out.

However RDTech did say they had stopped selling the one Dave tested, but yet AliExpress's "official" RDTech store is selling them.

AFAIK they're not doing a product recall. It will take a while for the supply chains to clear and the new version to trickle down.

I am hoping to run it with a 48V 10A DC power supply.

So, when it arrives how would I check the board is safe?

Dave's was due to a cracked capacitor that shorted out two pins. The only way to be sure yours isn't similarly cracked is to power it up. Be sure to record the whole thing for posterity.  :popcorn:

For protection should I fuse the DC supply input to the DPS5020

Where's the fun in that?

It might only be half-shorted. Even if there's no instant flameage you could still look at the unloaded current draw with a multimeter.
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12288
  • Country: au
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #42 on: November 25, 2017, 10:22:27 am »
So, when it arrives how would I check the board is safe?

Remove the cap.

RD tech has said this will not affect the operation - and the capacitor only reduced the output ripple a bit.

You can always replace the MLCC with another type.
 

Online paulca

  • Super Contributor
  • ***
  • Posts: 3994
  • Country: gb
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #43 on: November 25, 2017, 02:52:27 pm »
So, when it arrives how would I check the board is safe?

Remove the cap.

RD tech has said this will not affect the operation - and the capacitor only reduced the output ripple a bit.

You can always replace the MLCC with another type.

Could I not measure the cap isn't shorted and reads something like it should?  I mean if it's not cracked it should be fine right?  If it is cracked then it should read funny???
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Online paulca

  • Super Contributor
  • ***
  • Posts: 3994
  • Country: gb
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #44 on: November 25, 2017, 02:54:47 pm »
Dave's was due to a cracked capacitor that shorted out two pins. The only way to be sure yours isn't similarly cracked is to power it up. Be sure to record the whole thing for posterity.  :popcorn:

For protection should I fuse the DC supply input to the DPS5020

Where's the fun in that?

It might only be half-shorted. Even if there's no instant flameage you could still look at the unloaded current draw with a multimeter.

I think I might need to put a latex glove over the smoke alarm for this.  Embarressingly it's a building alarm so both my neighbours know full well when someone burns the toast (or toasts a PSU).

Anyway, I think I'll power it up on the bench supply first and limit the current so see how she goes before give it the full 10A supply.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline Brumby

  • Supporter
  • ****
  • Posts: 12288
  • Country: au
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #45 on: November 26, 2017, 12:06:18 am »
So, when it arrives how would I check the board is safe?

Remove the cap.

RD tech has said this will not affect the operation - and the capacitor only reduced the output ripple a bit.

You can always replace the MLCC with another type.

Could I not measure the cap isn't shorted and reads something like it should?  I mean if it's not cracked it should be fine right?  If it is cracked then it should read funny???

The capacitor in question is at risk because it can easily be subjected to mechanical stress that will cause it to fail.  This stress is applied when the screw terminals for the output are tightened.

You may receive a board that is perfectly fine - until you tighten the terminals ... for the first? ... third? ...ninth? time ... ... ... or it might never fail.

Removing or replacing the cap seems cheap insurance.
 

Online paulca

  • Super Contributor
  • ***
  • Posts: 3994
  • Country: gb
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #46 on: December 05, 2017, 06:16:39 pm »
Don't want to celebrate too soon, but it looks like they have hand reworked the board.

Pretty sure the offending cap is gone and poly of some kind has been soldered to the bottom of the board.

If you want more photos, just ask.  Although I am about to assemble it and find a rubber glove for the smoke alarm! :D


"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 

Offline glarsson

  • Frequent Contributor
  • **
  • Posts: 814
  • Country: se
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #47 on: December 05, 2017, 07:03:09 pm »
Pretty sure the offending cap is gone and poly of some kind has been soldered to the bottom of the board.
The fire happy capacitor was C34. The new bodge capacitor is not replacing C34.
 

Offline Nusa

  • Super Contributor
  • ***
  • Posts: 2416
  • Country: us
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #48 on: December 05, 2017, 08:49:11 pm »
Pretty sure the offending cap is gone and poly of some kind has been soldered to the bottom of the board.
The fire happy capacitor was C34. The new bodge capacitor is not replacing C34.
You're right, it's technically C33, which was previously unpopulated. Now it's C34 that's unpopulated. C33 functionally replaces it, just on the other side of the 20A fuses.
 

Online paulca

  • Super Contributor
  • ***
  • Posts: 3994
  • Country: gb
Re: EEVblog #1036 - PSU Fire PCB Repair
« Reply #49 on: December 05, 2017, 09:03:11 pm »
It works.  The only downside so far is they are so tight fisted they give you EXACTLY enough wire and no more.  So things are pretty tight to solder and connect.

Other than that, it powers on and seems to function.  It's a bit short on max volts.  My bench supply is providing it with 30V, it allows me to set 29.0V but registers 27.15V.  Although it has no load.  The DMM confers to within a few tens of millivolts with what it reads.

Also, a lesson for myself this evening.  If you want to test a power supplies response to a short circuit by stuffing you pliers in the binding posts.

DO NOT DO IT ON 30V!    :-[

Scared the living piss out of myself as sparks shot across the room.  I thought I was safe setting the current limit to 100mA.  Forgot about the 30V (or 27.15V whatever).  Lovely smell of ozone though :)

Dialled back to 3V it responds fine to the pliers, volts drop to some near zero amount, CC lights up.
"What could possibly go wrong?"
Current Open Projects:  STM32F411RE+ESP32+TFT for home IoT (NoT) projects.  Child's advent xmas countdown toy.  Digital audio routing board.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf