Hot capacitor with 18m2 low power board (CHI030)

#1
I provided a lighting kit for a customer a month ago and it was working happy as larry during extensive programming and testing.

The 18m2 low power board (CHI030).

However, they ran it for 10-15 mins and the capacitor got very hot and then nothing worked. I've got it back for testing / fixing and can confirm the capacitor is getting red hot but everything else looks fine. No obvious shorts, things are wired as I sent it so it shouldn't be having any issues really.

It's being run off 12v mains transformer with the standard 5v voltage regulator as the outputs are LED strips, but within capacity for the darlington driver (300mA spread over 5 outputs ie 60mA each output)

Why would a capacitor suddenly get hot? Is it just a faulty capacitor?

Note: The board come pre-assembled so the only items I've added are the voltage regulator and the connector blocks for the power and outputs. The capacitor therefore should be the right way round.

The customer is in France but I supplied the transformer and so it's a 12v regulated power supply, only difference is it's EU plug rather than UK plug. I had wondered if it could possibly be suppling more than the 16v the capacitor is rated to but it would seem unlikely.

Everything is as I tested it apart from the transformer.

Thanks.
 

hippy

Technical Support
Staff member
#2
Why would a capacitor suddenly get hot? Is it just a faulty capacitor?
You have not detailed what or where this capacitor is, how it fits in with the circuit so it is difficult to say. Doing so might help to identify the issue.

Things tend to only get hot when current flows through them and, hotter than expected, usually means more current than expected. For a decoupling capacitor or something across a power rail and 0V that would seem to suggest a capacitor fault, or it's not actually wired to what it should be wired to.
 

sghioto

Senior Member
#3
From your description I think you mean C2 on the CH1030 board. Very possible that the 5 volt regulator circuit you supplied may be at fault. Is the mains supply in France the same as in the UK?
SG
 
#4
From your description I think you mean C2 on the CH1030 board. Very possible that the 5 volt regulator circuit you supplied may be at fault. Is the mains supply in France the same as in the UK?
SG
Yes, it's the standard 100nf 16v capacitor supplied on the board. The 5v regulator is the one recommended by Picaxe (5v 100mA - 78L05) in the space on the board for the said regulator, just above the capacitor, and I've been using that set up for years so unless there's something else causing that to be insufficient it should be okay?

According to Wikipedia it's 230v 50 Hz so same as UK.
 
#5
You have not detailed what or where this capacitor is, how it fits in with the circuit so it is difficult to say. Doing so might help to identify the issue.

Things tend to only get hot when current flows through them and, hotter than expected, usually means more current than expected. For a decoupling capacitor or something across a power rail and 0V that would seem to suggest a capacitor fault, or it's not actually wired to what it should be wired to.
I assumed, since it was the standard capacitor that you would know, but it's C2 as per the schematic, standard capacitor for Picaxe boards 100uf 16v http://www.picaxe.com/docs/CHI030A.pdf
 

The bear

Senior Member
#6
@ sghioto, Electricity in France is 230 Volts, alternating at 50 cycles per second.
So its not the wrong sort of mains power.

@Jen, I guess this was a one-off, I know you soak tested the original. Just wondering if its worth duplicating.
Has the unit been returned to you?
Have faith in hippy.

Good luck, bear..
 
#7
Bear - yes, I've got the unit and confirmed that the capacitor is getting red hot within seconds of power being applied. This wasn't happening when I did the original testing so it's why I thought it may be a capacitor fault but don't want it to happen again when it goes back to France.

I can either try and replace the capacitor or replace the whole board.
 

premelec

Senior Member
#8
In my experience electrolytic capacitors that get hot usually pop their case [they are made that way]. They can fail by some defect in the component OR the power supply allowing AC voltage OR overvoltage - I suggest you simply replace the capacitor [100uF] with a known good one. For interest connect voltage to the removed one and see if it draws current with voltage on it - current should be very low in a good capacitor.

I also would check that the voltage regulator is working withing rated specifications - what is the DC input voltage and power dissipated in the regulator... There should be capacitors _near_ the regulator input and output - regulators can oscillate without proper capacitors [see spec sheet for your particular unit].
 
Last edited:
#9
In my experience electrolytic capacitors that get hot usually pop their case [they are made that way]. They can fail by some defect in the component OR the power supply allowing AC voltage OR overvoltage - I suggest you simply replace the capacitor [100uF] with a known good one. For interest connect voltage to the removed one and see if it draws current with voltage on it - current should be very low in a good capacitor.

I also would check that the voltage regulator is working withing rated specifications - what is the DC input voltage and power dissipated in the regulator... There should be capacitors _near_ the regulator input and output - regulators can oscillate without proper capacitors [see spec sheet for your particular unit].
The regulator doesn't have the capacitors, other than those already on the PCB. It's the Picaxe circuit and therefore there's nowhere to add anything extra. It's worked fine in the past on all the other boards I've used.

I'll try replacing the capacitor.
 

hippy

Technical Support
Staff member
#10
It could have been a faulty capacitor but it seems a little too coincidental that it was fine until the customer got it. I don't recall anyone ever having reporting a faulty electrolytic on any of our boards.

I think it's going to be hard to figure out what's gone wrong. The 100uF 16V cap is on the 5V regulated side of the circuit so should never have encountered anything more than 5V, no matter what supply the user was using.

Possibilities are the regulator has been fitted the wrong way round, the R9 link has been left in place, the 12V voltage has been applied to V1 rather than V2. Depending on how it was used in development and tested it is possible there could have been an issue lurking which only manifested itself once in the customers hands.

It's also potentially possible the customer did something which they are not admitting to.
 

hippy

Technical Support
Staff member
#11
It's being run off 12v mains transformer with the standard 5v voltage regulator

The customer is in France but I supplied the transformer and so it's a 12v regulated power supply

Everything is as I tested it apart from the transformer.
Lots of mentions of "transformer" there. Do you have details of what exactly you supplied, part number or link ?
 
#12
It could have been a faulty capacitor but it seems a little too coincidental that it was fine until the customer got it. I don't recall anyone ever having reporting a faulty electrolytic on any of our boards.

I think it's going to be hard to figure out what's gone wrong. The 100uF 16V cap is on the 5V regulated side of the circuit so should never have encountered anything more than 5V, no matter what supply the user was using.

Possibilities are the regulator has been fitted the wrong way round, the R9 link has been left in place, the 12V voltage has been applied to V1 rather than V2. Depending on how it was used in development and tested it is possible there could have been an issue lurking which only manifested itself once in the customers hands.

It's also potentially possible the customer did something which they are not admitting to.
Regulator - no, that's fine. R9 link is cut off, 12v has been applied to V2. Was tested with a 12v transformer. Hmmm, Agreed, I've used quite a few of these boards and never had any problems with them before and the capacitor is an unusual one to fail and since the board comes assembled there's usually not a lot to mess up.

Yes, usually it is the customer and she is a total novice but I shipped it assembled so there wasn't a lot she could have done. Her son was helping but no idea what they could have done. It's been shipped back to me assembled and I can't see anything obvious so not really sure what's happened.
 

premelec

Senior Member
#14
Have you determined that the hot capacitor is faulty? If you don't already have one of the cheap LCR testers available on ebay you might get one - they are handy though not always accurate.... You can add small tantalum capacitors - say 10uF easily to a circuit board on the foil side. Since you added the regulator it would be good to check its data sheet for capacitor requirements...
 
#15
Have you determined that the hot capacitor is faulty? If you don't already have one of the cheap LCR testers available on ebay you might get one - they are handy though not always accurate.... You can add small tantalum capacitors - say 10uF easily to a circuit board on the foil side. Since you added the regulator it would be good to check its data sheet for capacitor requirements...
I'm pretty sure Picaxe would have included them if they were required, but I checked the data sheet and no external components are required for standard operation. Like I said, I've been using this board with the 5v regulator for lots of different projects. If it was a problem with that I think it would have become apparent by now.

I haven't had time to check the capacitor yet. Can I check it on the board? Didn't want to keep making it hot. It's not bulging yet.
 

hippy

Technical Support
Staff member
#16
I haven't had time to check the capacitor yet. Can I check it on the board? Didn't want to keep making it hot. It's not bulging yet.
Best advice is to remove the capacitor as soon as possible, do not allow it to cause more damage.

I doubt it is possible to test the capacitor on the board because it's across the 5V and 0V rails and everything else on the board or connected to it would likely affect any attempt at testing.

I would guess it is not only the capacitor which is damaged but IC's as well. To get red hot it has to have a large amount of current going through it. I wouldn't have thought that possible without the regulator having also failed. If that has failed it might have failed in a manner which passes the 12V through to the 5V which will have damaged the PICAXE and other IC's.
 

techElder

Well-known member
#18
A capacitor is not going to pass a constant DC current unless it has a “short” between the plates. Period.

12 volts probably won’t damage a 16 volt capacitor unless it is overrated by the Chinese.

I agree that the regulator should also be changed, but I believe it has some over current protection.
 
#19
A capacitor is not going to pass a constant DC current unless it has a “short” between the plates. Period.

12 volts probably won’t damage a 16 volt capacitor unless it is overrated by the Chinese.

I agree that the regulator should also be changed, but I believe it has some over current protection.
It does have short circuit and over current protection and there's thermal shutdown protection.

There's no short between the capacitor. I'm not sure I understand what you're trying to say.
 

The bear

Senior Member
#20
Capacitors can and do go open circuit, or leaky, passing current through them. The leaky versions then get hot and bulge, leaking electrolyte.

High voltage capacitors (300--450v) in TV's were prone to exploding, messy enough to write-off the TV. So be careful with your 'hotty'.

Please keep us updated.

Regards, bear..
 
Last edited:
#21
Is it possible that this capacitor was accidentally fitted the wrong way around?

Electrolytics will get pretty hot if connected the wrong way around, to the extent that they may well blow.
 
#22
It would seem unlikely because it is a pre-assembled board by Picaxe and if it had been that way, we would have had the problem during testing.

Either something has gone wrong over time or the customer has done something to short it.
 

techElder

Well-known member
#23
Capacitors go bad. Change it out and forget about it. It may be the only thing wrong, but change out anything that has carried excess current, too.
 

hippy

Technical Support
Staff member
#24
I agree that the regulator should also be changed, but I believe it has some over current protection.
Indeed. Which intrigues me as to why it's allowing the cap to get red hot and not shutting down. That makes me suspect the regulator has also been lost to whatever adversity befell the board. But I could be wrong.

We know there's an issue with the capacitor or it wouldn't get hot. We don't know why or what caused that, or what other damage has been done. Removing the capacitor will at least allow other parts of the circuit to be tested.

I would disconnect all connections and remove the IC's, check the 5V as it currently is, and then after the capacitor has been removed. If it's more than 5V, give or take the usual bit, then the regulator has probably also gone.
 
#25
I have used quite a few CHI030 boards and they are great.
One time the supply accidentally got connected in reverse (it happens!) and the entire board survived except for the regulator. I knew it had happened but your customer may not have realised, or hoped it would be OK?
I too would change the regulator (only pence anyway).
. . . and fit some form of reverse polarity protection, if not already?
 
#26
I have used quite a few CHI030 boards and they are great.
One time the supply accidentally got connected in reverse (it happens!) and the entire board survived except for the regulator. I knew it had happened but your customer may not have realised, or hoped it would be OK?
I too would change the regulator (only pence anyway).
. . . and fit some form of reverse polarity protection, if not already?
Thanks Dartmoor. I did supply it with the power wires already connected so all they needed to do was plug in the power supply jack plug into the adaptor, however, it is possible they changed this. I also have warning in the instructions not to connect it in reverse, but who knows! I do usually use a cheap 1N4001 diode to protect but it depends if there's suitable spare connections on the board. I'll see if I can add one.
 
#27
Indeed. Which intrigues me as to why it's allowing the cap to get red hot and not shutting down. That makes me suspect the regulator has also been lost to whatever adversity befell the board. But I could be wrong.

We know there's an issue with the capacitor or it wouldn't get hot. We don't know why or what caused that, or what other damage has been done. Removing the capacitor will at least allow other parts of the circuit to be tested.

I would disconnect all connections and remove the IC's, check the 5V as it currently is, and then after the capacitor has been removed. If it's more than 5V, give or take the usual bit, then the regulator has probably also gone.
Thanks Hippy. Will give that a go.
 
#28
I've done some investigating and it looks like it's the MP3 player which is causing the problem - it's the Picaxe SPE035 module. http://www.picaxestore.com/index.php/en_gb/picaxe/add-on-modules/spe035.html

I took everything off and just supplied power and it was all fine. Capacitor and regulator both normal and reading 5v so both are still doing what they should.

I then started plugging stuff back in 1 by 1, least likely to be the issue first and as soon as I re-added the MP3 module the capacitor got hot within seconds. The light was also on on the Mp3 player.

I tried a different Mp3 module and the capacitor was fine and the light didn't come on. I assume something has happened to the mp3 module that's somehow caused a short circuit but it all looks okay.

picaxe18m2andMP3.jpg

No idea what's happened to cause this as it was working fine when I had it and tested it but at least the cause has been found.

EDIT: Tried seeing if a light would work with the different MP3 player and when nothing happened I then noticed that the 5v regulator was getting very hot. It is reading a consistent 5v so it is working but something funny is going on as it shouldn't really be getting red hot should it?
 
Last edited:

hippy

Technical Support
Staff member
#29
No idea what's happened to cause this as it was working fine when I had it and tested it but at least the cause has been found.
I have no idea why it is happening but there seems no obvious reason adding something across 5V and 0V should cause a capacitor which is also across 5V and 0V to get hot. Unless perhaps that something was back-feeding into the 5V rail.

If there were a short in the MP3 module or it was drawing too much current it would be that getting hot not the capacitor.
 
#30
I just tried seeing if a light would work with the different MP3 player and when nothing happened I then noticed that the 5v regulator was getting very hot. It is reading a consistent 5v so it is working but something funny is going on as it shouldn't really be getting red hot should it?

I was going to measure current through the regulator but the speaker started emitting noise so I quickly disconnected.
 

hippy

Technical Support
Staff member
#31
something funny is going on as it shouldn't really be getting red hot should it?
There is definitely something funny going on.

We would not recommend using the boards, CHI030A or SPE035, nor the MP3 Player module, which have been problematic, until you can figure out what has gone wrong or you may simply end up damaging and destroying other boards and components.
 
#32
There is definitely something funny going on.

We would not recommend using the boards, CHI030A or SPE035, nor the MP3 Player module, which have been problematic, until you can figure out what has gone wrong or you may simply end up damaging and destroying other boards and components.
Okay, agreed, but this is way beyond my level of expertise. I have checked the MP3 player is only receiving 5v, which it is, but where do I even start?

Any ideas?
 

premelec

Senior Member
#33
Is the regulator physically near the capacitor and perhaps heating the capacitor before you noticed it? It may be there is simply overcurrent through the regulator due to some fault in an attached unit. You can measure current by putting a small [10 ohm..] resistor in series with the 5v line to external units and measuring the voltage across it. Situations like this boil down to eliminating one item at a time. Perhaps a picture would help us... Does the regulator have a heat sink? What currents are you expecting in normal operation?
 

hippy

Technical Support
Staff member
#34
Okay, agreed, but this is way beyond my level of expertise. I have checked the MP3 player is only receiving 5v, which it is, but where do I even start?

Any ideas?
Depends on what you want to do.

If you just want to get something back into the hands of your customer it would seem the best way forward is to take another set of boards and components, get them working, send them out.

If you want to find out what went wrong and what's now broken or of unknown status on what's been returned that's going to be a lot more difficult.

As is repairing the boards so you know they are reliable enough if sent back out.

First thing is to work out if you have the skills and tools to test everything and whether the effort of doing all that is worth it or it's better to mark it up as a bad experience, throw that in the bin or put it to one side, and just move on.

The most bizarre thing is that capacitor getting hot, and in the circumstances it does. The first thing perhaps is to investigate that, mentally figure out what failure modes could plausibly exists for that. Then set about determining if they do exist or not.
 
#35
Is the regulator physically near the capacitor and perhaps heating the capacitor before you noticed it? It may be there is simply overcurrent through the regulator due to some fault in an attached unit. You can measure current by putting a small [10 ohm..] resistor in series with the 5v line to external units and measuring the voltage across it. Situations like this boil down to eliminating one item at a time. Perhaps a picture would help us... Does the regulator have a heat sink? What currents are you expecting in normal operation?
I put a photo in my post on page 3.

No, it doesn't have a heat sink. I'm confident the capacitor was heating up. They are next to each other but the capacitor was extremely hot. I doubt the regulator would have passed that much heat through.
 
#36
Depends on what you want to do.

If you just want to get something back into the hands of your customer it would seem the best way forward is to take another set of boards and components, get them working, send them out.

If you want to find out what went wrong and what's now broken or of unknown status on what's been returned that's going to be a lot more difficult.

As is repairing the boards so you know they are reliable enough if sent back out.

First thing is to work out if you have the skills and tools to test everything and whether the effort of doing all that is worth it or it's better to mark it up as a bad experience, throw that in the bin or put it to one side, and just move on.

The most bizarre thing is that capacitor getting hot, and in the circumstances it does. The first thing perhaps is to investigate that, mentally figure out what failure modes could plausibly exists for that. Then set about determining if they do exist or not.
I don't really have enough experience to work out why the capacitor is getting hot or what might cause it.

It seems the easiest would be just to do a new set of boards and test again. But it was all working fine before I sent them the first time. That's my only concern.
 

RexLan

Senior Member
#37
OK ... as several others have said and if it were me I would remove the cap and replace it as a start. Does not matter it if looks OK and if it appears to be correct. Replace it with a new one and see if it gets hot too. The capacitor is the suspect for me.

Better to understand than to swap components like car dealers.
 

hippy

Technical Support
Staff member
#38
It seems the easiest would be just to do a new set of boards and test again.
That is what I would do.

But it was all working fine before I sent them the first time. That's my only concern.
I think we can all understand concerns that the same might happen again. Whatever has happened shouldn't have happened but it is currently impossible to say what did happen, and it is incredibly difficult to apportion blame for the outcome.

About the only thing you can do is to have trust in your design, testing and instructions, your customer that they won't do anything wrong or something they shouldn't, and that our boards and components won't fail unexpectedly.
 
#39
That is what I would do.



I think we can all understand concerns that the same might happen again. Whatever has happened shouldn't have happened but it is currently impossible to say what did happen, and it is incredibly difficult to apportion blame for the outcome.

About the only thing you can do is to have trust in your design, testing and instructions, your customer that they won't do anything wrong or something they shouldn't, and that our boards and components won't fail unexpectedly.
Okay, thanks Hippy. Will give it a go and hope it's all okay this time.
 

rq3

Senior Member
#40
I just tried seeing if a light would work with the different MP3 player and when nothing happened I then noticed that the 5v regulator was getting very hot. It is reading a consistent 5v so it is working but something funny is going on as it shouldn't really be getting red hot should it?

I was going to measure current through the regulator but the speaker started emitting noise so I quickly disconnected.
Do you have an oscilloscope? If the capacitor isn't being DC voltage stressed, and isn't installed backwards, the MP3 unit may be inducing a severe radio frequency (RF) oscillation, either in itself or the voltage regulator. That could also make the decoupling cap overheat very quickly. There should be a 0.001 uF ceramic cap in parallel with the electrolytic to swamp that kind of thing, but different regulators are very sensitive to capacitive loading, and to the equivalent series resistance (ESR) of the decoupling caps. The exact device data sheet is the bible.
 
Top