Melted torroid

[soundguy]

hi folks-

I powered up a piece of gear today that had been working just fine with no problem for about a year. After the bad smell came, I pulled it from the rack to discover a super hot case and upon popping the top, the power torroid had melted. While I know this is a broad question, what are some of the things that can lead to this happening? This box has been on for hours and hours at a time, so I think if I wired it wrong, this would have happened a real long time ago. Im a bit powerless to trouble shoot this, so any possible avenues to look down would be greatly appreciated. Do power transformers just decide to die? It was a brand new one from digikey when I installed it.

thanks!

dave

 

[radiance]

Scary stuff. Where the mains fused?

 

[soundguy]

yes. oddly enough there doesnt seem to be any damage even in the power supply and the fuse didnt fail...

dave

 

[radiance]

Beats me...and scares me.
Either the fuse or the power supply should have been damaged if you ask me. Was it a heavy fuse? A melting toroid will draw some current...

 

[soundguy]

note to self: dont let the mounting bolt on your torroid touch the bottom and top of the metal chasis...

further proof that I actually am a complete idiot.

dave

 

[kato]

[quote author="soundguy"]note to self: dont let the mounting bolt on your torroid touch the bottom and top of the metal chasis...

further proof that I actually am a complete idiot.
[/quote]

Maybe I'm the complete idiot but why would this matter?
The mounting bolt isn't live? Is it?

 

[radiance]

@Soundguy: Looks like you found out what happened in there right?
Check out this thread. Someone shielded his toroid with a tin can upside down.

 

[Bjorn Zetterlund]

I've done that too Dave; was silly enough to clamp a shielding plate to the top and the bottom of the mounting bolt once. Luckily I caught it in about 20 seconds and it didn't fry...

Kato, the chassis and bolt make a shorted turn on the transformer if the bolt touches the chassis at the top and bottom. The transformer will do its best to push current through the chassis and bolt. A bit like shorting all the secondaries, but much worse!

Bjorn

 

[blakbeltjonez]

[quote author="soundguy"]note to self: dont let the mounting bolt on your torroid touch the bottom and top of the metal chasis...

further proof that I actually am a complete idiot.

dave[/quote]

that's strange, i think Joe Malone mentioned that very same "don't!" to someone else on DIY here just yesterday - and now you found out why!

 

[CJ]

A good thing to do is short the secondaries of the transformer briefly before you install it and measure the current draw. That way you can fuse it so it dosen't catch fire. I use to put humongous fuses in my guitar amps because I hated popping fuses, I would rather deal with a cooked transformer than be interupted.

Then one day I accidently left an amp on all day and when I came back from work, the place was filled with smoke and henious varnish fumes. The tubes went which took down the x-former.

 

[bcarso]

Ahh the phantom shorted turn effect! When I started the thread I thought hmmm sounds like a shorted turn...but i figured internally not the old trick of the mounting details.

Oh well, you join the company of many illustrious people!

 

[kato]

Would it make sense to mount the toroid using a non-conductive bolt?

 

[alk509]

[quote author="kato"]Would it make sense to mount the toroid using a non-conductive bolt?[/quote]

Yeah, or just make sure there's no way the top of the bolt will ever touch the chassis by either leaving enough clearance, or insulating the head of the bolt with something.

Peace,
Al.

 

[soundguy]

after today, a nylon bolt that is strong enough would be the way to go. I had been working on something and put it on top of the rack without screwing it in, and it forced the top of the chasis down on the mounting bolt. dumb luck. I had electrical tape on the inside of the lid but the pressure pushed the bolt through. dumb... My new project for next week is installing new shorter or nylon bolts into everything, this was a really stupid thing I dont want to repeat, live and learn. Not much of a worry with a double space box, but a single space, there isnt much clearance above the torroid and the lid...

learn from my woe.

dave

 

[jrmintz]

I try to cut the end of the bolt flush with the top of the nut, which ought to be roughly at the top of the transformer. It doesn't prevent shorting, but it makes it a little bit harder. I don't think a nylon bolt would be strong enough - one good jar and it would shear off with a heavy transformer.

 

[Family Hoof]

Sorry to hear about the toroid, Dave. On the bright side, at least you're raising awareness.

I've been trying to get ahold of some non-flamable neoprene foam and adhesive or something similar for insulating the insides of rack cases underneath or above anything that can get shorted. Used to be easy until the rubber store on Canal Street closed (or did it?). Now I have to order it online like everything else! Anybody done something similar or have tips?

 

[CJ]

Would a brass bolt be ok?

 

[soundguy]

Im not so upset this happened to be honest, it was a really really really good lesson and pretty much the first close call Ive had with power, so my head is really turned around in the right direction. This has prompted me to go back and double safety check everyting Ive built, which is something that I shoudl have done a while ago but of course kept putting off.

dave

 

[BYacey]

Use "Goop" glue to glue the torroid down to the chassis and use a nylon bolt. The bolt will never see any sheer stress unless the glue bond breaks. Nylon tends to strech and give before breaking anyways, unless it's old and brittle.

 

[PRR]

> Would a brass bolt be ok?

No. Worse.

The situation is: he had a conductive path up through the hole in the toroid, over, around and down, and back to the hole. You will notice this is the same path as the wires wrapped around the toroid, just not as tight to the core.

The AC-induced magnetic field urges electrons around such loops. As long as the loop is open, no electrons flow, no problem. If the loop contains something to slow the electrons, like a load (input of a power supply), a reasonable amount of power flows.

But he had a fat bolt and metal panels. The bolt is similar to the fat wire on your car's battery, which flows 100+Amps easily. Except the bolt is a lot shorter, lower resistance. The panels may be thin but wide, pretty good conductors.

The resistance to current flow could be as low as 0.001Ω. The induced voltage could be 0.1V per turn (typical of many power transformers). And 0.1V/0.001Ω is about 100 Amps.

Actually, this is so much more than the design load that the current will be limited by primary resistance: most of the power appears as heat in the primary winding.

Hmmm... too many unknowns to guess what the actual current in the transformer was. Evidently it was less than wall-socket fusing, and less than equipment fuse.

To make-up some numbers: say a 25VA transformer, and loaded to 25VA. Transformer efficiency is 20%. At 25VA output, there is about 0.20*25VA= 5 Watts of heat in the transformer. A 25VA transformer does not have 25 Watts of heat: most of the power flows-through to the load.

Or in ohms: say it was 120V 0.21A load, 576Ω load, 20% efficiency, the transformer inefficiency acts like a 144Ω series resistor. Now put a near-short on the transformer. We have 120V in 144Ω, 0.83Amps, 120V*0.83A= 100 Watts, all in the transformer! Which is designed to run with about 5 watts of heat in it. It cooks. OTOH, the 0.83 Amps won't blow the house-fuse, or even a "reasonable" 1A in-box fuse. (Cap-input power supplies often have to be over-fused to survive start-up surges.)

> non-flamable neoprene foam and adhesive

They sell a tough rubber sheet for gripping stuck jar lids. ShoeGoo will stick it to anything.

> anything that can get shorted.

Here's a tip I learned the stinky way. Silicone Seal or similar bathtub caulk will NOT stand-off 600 Volts. It works for a few seconds, then leaks, the leaky part carbonizes, leakage increases, it gets hot, and pretty soon it burns. I use it all the time for lesser voltages, but not around big tubes.