Has anyone ever had a PIC and/or PCB catch fire?

hokieed

New Member
I had a PICaxe microcontroller on a custom printed circuit board catch fire and fry itself and a nearby board. Have you ever seen this happen ever? It doesn't seem like the kind of failure caused by a gradual overheating with a current overload. It seems more like a large failure caused by a high voltage getting in somehere. Here are some details:

-The board powers a small array of 12v DC LEDs, and a 12v DC linear solenoid. The solenoid is protected with a diode to keep back EMF from getting to the board.

-The board is wired to a 12volt DC power supply. The pic is powered by a voltage regulator and capacitor combination as recommended by the PICaxe manuals.

-This morning, the problem was observed as smoke came from the box containing the boards. None of the electronics worked.

-Upon opening the box, two of the boards were chared, appearently originating from one of the PICs and possibly a large power transistor.

-The boards were working fine for a whole week, running 10 hours per day. Then, this problem suddenly happened. Up untill then, no heat or smoke or anything was observed.

-The burn pattern in the box indicated a rapid flare-up and catastrophic failure in a very short amount of time.

There are other details involved in this that might be the cause, but what I want to know is if this has occured to anyone else.

Have you ever seen a board catch fire from an overheated component on a 12v DC board?

Have you ever burned up a PIC and did it ever get to the point of flames?

Have you ever seen this failure caused by a large voltage accidentally being applied to a PCB? (which is one possible scenario I can imgaine... a high voltage potential leaking across a component, such as the solenoid, and making it to the board)

Just please share your experiences. I'll have the boards in a few days to inspect and take pictures of.

Thanks in advance for any help.

Eddie
 

Mycroft2152

Senior Member
I had an IC driver chip explode when I accidently shorted it to the NIXIE tube (140volt) power suppluy. There was actually a small crater in the center of the chip. :)

Shorted connections can create a lot of heat and ultimately fire. PCBs usually have a fire rating from built in flame retardants.

Since you mentiopned that the boards worked for long time, it may be that they were being run beyond the limits of their safe parameters. This would shorten the life and result in a catastrphic failure.

5 volt devices are not very happy with 12 volts.
 

SD2100

New Member
Could be lots of reasons for the failure, you would have to post circuit diagram also details of the relays and power supply. The answer will be in there somewhere.
 

hippy

Senior Member
I've seen a lot of things - wires smouldering, tracks peeling from PCB's, windowed Eproms lighting up like lighthouses, Leds looking like lasers for a brief period of time ( until their tops melted off ), chips cracked in two and those with holes blown clean through them, transistors with no tops just legs, electrolytic capacitors which have unwound themselves and chucked gunk everywhere and, most impressively, a bank of eight 40A triacs going off like hand grenades.

In all cases it's been down to excessive current or voltage. I hot-wire cut deep into a finger once [unintentionally] holding a single strand of wire against a 1.2V NiCd batttery. Ouch.

Was the power transistor by any chance switching 12V ? It's quite likely for there to be cascade effects in failure. If the transistor blows it can dump 12V through the PICAXE, they in turn melt, short out more things and it all burns until a supply track or wire finally gives up and disconnects the power or a fuse blows. It's often hard to tell what went first but I'm guessing the 12V supply was capable of supplying a fair dose of current.

I believe it is rare for a 'spark' from something to ignite anything, especially at voltages below 50V and although tracks can be arced-over it's not often sustained; the arcing burns the track away and kills the current path. It is always possible the PSU failed and dumped mains through the system. A brief belt of too much voltage or current and there will be a violent reaction, enough to cause short-lived sparks, burning and a slightly longer period while the magic smoke makes it way into the ether.

My suspicion would be directed at the power transistor and a high current 12V supply but it would be interesting to see a circuit diagram and what the photo's of the board damage looks like.
 

Rickharris

Senior Member
Personally I have always liked those faults that were easy to find - you don't get easier than looking for the charred remains.

We have had 08's far too hot to touch - mostly by reversing the power supply - they still work when cool!!.

Not had one melt or burst into flames although have had alkaline battery explode (split the case) in a battery pack last year.
 

demonicpicaxeguy

Senior Member
once upon a time when i lived in malaysia during one of the worst storms i've seen to date, the house we were in quite literally got hit on power lines comming to the house i was 12 at the time and it was me and my father home ,

pretty much every light fixture was blown off the walls leaving black marks everywhere the, fax machine left the table it was on and ended up the floor
and caught fire which proceeded to setfire to some of the office
the fridge also managed to catch fire after it emmited a flash of light from behind it everything else was unplugged

it was probably the biggest bang i had ever heard and after the ears stopped ringing from the initial bang i heard my father yelling "get the ####### fire extinguishers and some tape" which at that point i saw the smoke comming out of the office

my father took one fire extinguisher into the kitchen to put out the fride while i taped down the handle on anther fire extinquisher and threw that into the office and shut the door then we got out the house before we ran out of breathable air
 

kevrus

New Member
It's quite possible that the fault occurred in your power supply and not on the board(s). Have you checked your PSU output voltage since the failure?
 

BeanieBots

Moderator
I've seen a similar array to that described by Hippy. Many caused by external events such as high voltage spikes which cause a cascade event. That is why we bang out the safety warnings to anyone contemplating the use of home made electronics in cars.
Ultimately, such a catestrophic failure is down to the design which has failed to consider a particular event. I used to do such analysis for a living for the purpose of blame involving insurance claims. Nearly always falls on the doorstep of the electronics designer:eek:
Without seeing circuits and board, impossible to tell, but gut feel is that transistor failed short due either to excessive power dissipation or inductive spike not caught by clamp diode. That in turn put 12v into PIC.
Await the pictures!
 

Coyoteboy

Senior Member
Yup, i had a fan controller on my car using a picaxe and a high current driver. It was fused, had a double-smoothed/filtered power supply but I failed to consider the fact that when you put the thing in a box the heatsink would have less airflow that it did out of the box. The heatsink melted the box when the fan got jammed (fortunately it was fire retardant) and while driving along I could smell that nice aciding hot plastic smell, pulled over and opened the bonnet - it didnt catch fire but it did sit and smoulder for a while. Two things I now do with all my in-car projects - use a borderline fuse and make sure any high current switching is done in a soft-start method, and run them in quality metal boxes ensuring they are not packed into carpetted/high plastic content areas.
 

Tom2000

Senior Member
I haven't ever had anything I built catch fire, but I once had a 1/4 amp fuse filament glowing cherry red for a few minutes before it blew. Darndest thing I ever saw.
 

andrew_qld

Senior Member
Likewise, in my industry where we have trasnmitters with up to 15,000 volts floating around I have seen a lot of things, including gecko lizard's and snakes burned to a crisp accross HV rectifiers (at that voltage level it comes to you).

Most of the real carnage I have seen though, icluding very burned PC boards has been on low voltage high current / heat DC applications. Oh, an the occasional 240VAC mains board.

Sounds to me like maybe you got 12V into the pic or the transisitor overheated? I doubt the picaxe would have the carried enough current to cause a fire. It's legs would melt first.
 

moxhamj

New Member
My septic tank pump control system caught fire once. It was/is a commercial product. Luckily the board lives inside the concrete tank so the fire couldn't go anywhere and luckily the methane/oxygen levels were not at the critical explosive ratio.

I saw a current affair TV show a couple of weeks back about houses burnt down due to TV sets catching fire while on standby. It isn't an unusual problem, and would presumably apply to computers too.

I've certainly pulled apart old electronics equipment where components have been overheated and melted nearby plastic. One mode of failure is power supply electrolytic capacitors. They are often rated right at the limit and then they get a bit warm and then a bit of electrolyte evaporates and then the ESR changes and the ripple current goes up, and then the components supplying that component (diodes/transistors) end up overrated and they get hot. I did read once about a TV serviceman who said much of the time a fault can be fixed by mindlessly replacing all the electrolytics. A quick internet search for exploding capacitor recall brings up some interesting stories. It seems there are dodgy components out there.

Please post the photos when you get them.
 

hokieed

New Member
thanks for the responses. here are two pictures.

The first picure shows the burn pattern and where the boards were. They are mounted vertically, secured with foam tape on a laminated wood surface.

The char pattern picture shows that the main failures happened in the "Track switch" board, the "spectator" board, and the "link lift" boards.

The track switch picture shows how the PIC, on the right of the board, seems to be the msot melted pat of the board, even burning the terminals next to it.
 

Attachments

hokieed

New Member
Two more pictures.

The spectator picture shows the other board that had the most damage. It looks like the transistor got fried and burned everything else up. But it looks like a flame did it. Can a simple heating of the board do this?

The link lift picture shows minor damage. It appear that the power transistors legs warped, but that was a field modification on all boards to fix an output/input pin problem.

I know a circut diagram wil be helpful, so I will work on getting something easier to read than a hand sketch.

Thanks again for your stories and tips.
 

Attachments

papaof2

Senior Member
It looks as though the upper and center leads on the TO220 device are crossed - which is a failure waiting to happen.

What does the wiring side of the board look like? I would suspect that the crossed leads shorted, causing the traces to get very hot, burning the board from the back side.

If that is the failure mode, it could have been prevented by better layout (leads straight) or better construction (a penny's worth of heatshrink tubing on one of the leads).

John
 

hokieed

New Member
well, the center leads look crossed because of a change I had to make in the field. There was at least 1mm of spacing between the legs when I bent them and soldered them back in the board. Until I get the boards in my hand for inspection, I cannot rule out a possible short at that location. But I've never had 12v DC arc over an air gap before. So, I keep investigating.
 

hippy

Senior Member
All boards appear to have suffered the same failure effect in roughly the same area. Although the crossed over legs cry out as a being the culprit I'd be surprised. It looks to me like a classic case of over-current heating things up, somewhere to the left of the TO-220 and possible at the join of its top leg to the board or track near there.

The initial failure could have been a heating/melting transistor/regulator ( whatever the TO-220 is ) and that could cause the body to distort and the legs to short which in turn would dump current into the other boards. The cross-over legs may have contributed to the cascade but are probably not the cause of original failure.
 

BeanieBots

Moderator
I'm with Hippy. Short of having any other evidance other than the pictures.
Initial heating of the T0-220 device causing bending of the crossed legs which then shorted. If the legs were under stress when soldered, then excess could have melted the solder which allowed them to move.
Certainly a common fault on all units so still a possiblity of power supply issues and/or an injection of electrons where there shouldn't have been one.
I'd hassard a guess that a humble fuse could have prevented such catastrophic failure.
 

hippy

Senior Member
Thinking of root causes -

Bulb failure. Going short circuit could have the transistor shorting the power supply through itself. A break in the cable could go short circuit but more likely open. Flexing or trapped cable could cause a shorting fault. Any builders knocked a nail through the cable ?

In-rush current could have put undue strain on the transistor, and it could have been running on its limits anyway.

One thing I suppose we cannot entirely discount is lightning strike or mains power surge.
 
I'd say I was jealous. I let the smoke out of a DS1820 this morning, but all I got was a momentary funny smell and a burned finger.

On the other hand if my projects started catching fire I don't think my wife would let me play with the toys anymore, so I guess it's all good.
 

hax

New Member
-The board powers a small array of 12v DC LEDs, and a 12v DC linear solenoid. The solenoid is protected with a diode to keep back EMF from getting to the board.

-The board is wired to a 12volt DC power supply. The pic is powered by a voltage regulator and capacitor combination as recommended by the PICaxe manuals.

Looks interesting! I love a good investigation.

Just a couple more questions:

What are the specs of the solenoid, and in normal operation, does the solenoid operate momentarily, or constantly?

What are the specs of the power supply? Switchmode or transformer? Current rating?
 
Top