burnt toroid?

[chooken]

looking at the board there seems to be a few caps that have died something shocking, they are identifyable. however i have also noticed that the toroid (VRM?) is burnt and seems to be cooking and producing a smell when the machine is on. it will still boot but it incredibly unstable and dies once it has warmed up.

the location of the toroid is right next to 2x 1500uF 10v caps. it has 5 twists of wire excluding its feet that go through the pcb. the board is a MS-6351.

i assume the toroid needs replacing, i just have no idea what to replace it with.

any help would be greatly appreciated as i have 15 boards in this condition.

[Chris1992]

Re: burnt toroid?

Can't help you much except to say to never turn on boards with anything burnt on them. You'll usually end up cooking something expensive.

[kc8adu]

Re: burnt toroid?

just get a similar one from a scrap board.
or if you can find the same gauge magnet wire rewind it.
oh and recap it to prevent smoking the new coil :-)

[chooken]

Re: burnt toroid?

Definately recapping the board, also wanted to replace the toroid. Do i just take off the wire, go for a walk down to my local electronics store and ask for wire of the same gauge as what ive got? The round (magnet) looks slightly toasted, its ok to use still, just as long as i wrap it with new wire yes?

Thanks a heap :-)

[AK0R]

Re: burnt toroid?

The round (magnet) looks slightly toasted, its ok to use still, just as long as i wrap it with new wire yes?

The toroid core is a simple iron core which, coupled with the wire wrapped around it, provides more inductance than an air-core toroid. The toroid core is *NOT* magnetized (therefore, it's not a magnet) and *should* be unaffected by overheating. It would help to take a picture before you start, if you're going to rewind it yourself. The wire gauge, the number of turns, the spacing, the direction each winding is wrapped on the toroid, and the connection point of each lead are all important.

[PeteS in CA]

Re: burnt toroid?

I'm new, but been lurking for a week or so. My background (which is relevant in this thread) is in P/Ss (PSUs) rather than in systems, mostly in engrg, with some mfg, repair, and tech support. I'm not working for a P/S company, a components company, nor a computer company - no conflict of interest.

I don't think chooken is going to enjoy what follows, but there is a significant chance that the toroid core has been damaged. The core is almost certainly a powdered iron core, which can age significantly at core temps above ~110-120C. The aging affects the core's magnetic properties, leading to a runaway failure. If you want some technoid info, go here http://www.micrometals.com/thermalaging_index.html . This company is probably the premier mfr of powdered iron cores.

Unfortunately, to replace the inductor, you would either have to obtain a replacement part from the MB mfr or get the vendor and P/N for the core and wind one yourself. It is very possible, even likely that the failure of the MB caps and the overheating of the inductor were mutually reinforcing.

PeteS in CA

[kc8adu]

Re: burnt toroid?

i have seen toroid cores that were too cooked to reuse.
if his is shot he can pm me.
he can send a padded sase and i can match one up.
i strip boards with a heatgun before i pitch them.
i service stuff that cannot be easily or cheaply replaced(industrial stuff,cnc hardware,ect) and i have to keep my junk box well stocked.

[PeteS in CA]

Re: burnt toroid?

I'd be careful in using salvaged parts. You basically need the same type/size of core - inductance:turns ratio and loss characteristics - as the original part. Otherwise you could have problems such as oscillation, extra stress on the caps that could destroy good caps, or the inductor might fail again in a few hundred or a couple thousand hours.

The inductor and the caps are part of an L-C filter that takes a square wave and smooths it into a DC voltage. In doing this, the caps conduct a relatively high magnitude of ripple current - if you know the ripple current and the ESR of the cap, I**2 x ESR = the power dissipated by the cap. At the same time, the inductor is charging and discharging (never fully) with every cycle. Magnetically, this means that the core is flipping back and forth. The core's reluctance (I think that's the right word) causes a loss, i.e. heat, in the core. If a powdered iron core gets hot enough long enough, its magnetic properties deteriorate into runaway destruction.

That's a very long-winded way of saying that if the MB maker uses cheapo caps (or gets a bad batch), uses a cheapo inductor core, or under-designs the inductor and caps network, the result will be unpleasant.

[willawake]

Re: burnt toroid?

welcome pete!

i see you know electronics. if you ever have time to write something for the FAQ section then we would be most pleased.

[PeteS in CA]

Re: burnt toroid?

Sorry for taking so long to respond. I'll work on that.