Is this an Inductor?

[NewCapper]

I have a Video card that I was given by a friend. After looking at it I realized that it had a broken component on it. I belive the component that is broken is and inductor. Can anyone confirm for me that the component is and inductor? Also, what does the component do I believe it is similar to a capacior? I assume this Component is replacable? Where Can I get a replacement?
Any help would be appreciated.

[Krankshaft]

Re: Is this an Inductor?

That indeed is an inductor a toroid inductor to be exact with the core cracked like that you will need to replace it.

You will need to take careful note of how its wound and the number of turns and the gauge of the old wire and wind a new core with new same gauge enameled wire.

Also be sure to take careful measurements of the old core when ordering a new one.

It doesn't have that many turns so it shouldn't be too hard.

Unless you know the value of the inductor in question you can't buy a new one.

[stevo1210]

Re: Is this an Inductor?

It's a thoroid coil so you'll have to try and find a similar one on maybe other unused computer stuff you may have?
Try to glue the center bit of the thoroid coil back toghether and see what happens first? maybe you can get away without having to rewind or find a new one.

Thanks.

[gonzo0815]

Re: Is this an Inductor?

Just take some super glue and put it back together. If there is no missing parts, it may work w/o problem. J&B weld is may be the better glue, but duno if it contains some metallic compound, which would certainly alter the core`s property.
Most of those inductor do have an air gap any way, thus it should not hurt.
But may be, it is an high frequency switcher, then it may not work.
If it is getting too hot or or if the GFX does not work, you have to replace it.
The problem is to figure out, what value and what core material it is. Usually i think you can`t get this too easily.

[PCBONEZ]

Re: Is this an Inductor?

You need to match the inside diameter, outside diameter, and width of the doughnut and the diameter and number of turns of the wire.
-
If you can find/scavenge a doughnut the right size and some wire the right size you can wrap your own.
-
Glue isn't a great idea because they work in part based on the magnetic lines passing through them. Even a crack will screw that up, let alone a crack with glue in there insulating things.
-
You can buy them but looking them up is a pain because of all the dimensions you have to get right.

[gonzo0815]

Re: Is this an Inductor?

Well PCBonez, i hardly agree with you on this matter ;-). A usual toroid core is in reality nothing else then encapsulated iron or ferrite particles. Any particle is surrounded by either some oxide layers or a moulding compound.
This is absolutely necessary to make them useful at higher frequencies in the first place.
The glue will not disturb the magnet field as long as it is an reasonable thin layer (which is easily archived with super glue). And even a thick (metal free) glue would not disturb the magnetic field, but it would certainly change the inductors properties in other areas.

The only concern i have, is that if it is an high frequency VRM, that the loses would increase due to the air gap. This would IMHO lead to some increased heating, but this would not end in an instant fry.
But i think that the cores are simple iron cores, which require an air gap any way.
So i would certainly try it without much worries.
If it works, it was an easy fix, which is fully backed up by at least my basic understanding of toroid cores
If not, i am sure to some degree, that it will cause no further damage, as most of today's VRM controller chips have build in core saturation protection.

May be some of the more theoretically fit people could correct me, if i am totally wrong (which i do not assume ATM...).
At least i would say, it is not a ghetto rep job idea, what i have suggested. It is common practice to try it with super glue first.

[PCBONEZ]

Re: Is this an Inductor?

I stand by what I said. Cracks have significant effects on magnetic fields. If they didn't Eddy Current Technologies used to locate cracks in magnetic metals wouldn't work.
-
http://adsabs.harvard.edu/abs/1998JAP....83.6353S

[Krankshaft]

Re: Is this an Inductor?

Thats what I thought about using glue too.

I once had a ferrite core crack in a variable inductor on one of my older FM transmitters and glued it back together and it didn't work.

Luckily I was able to buy a spare core rather then replace the whole inductor.

[gonzo0815]

Re: Is this an Inductor?

Eddy currents can only spread in uniform metal units. In any ferrite or iron core powder, the particles are isolated from each, to avoid exactly this.
Even at 50Hz most transformers are build with isolated iron plates glued together.
I would say you can`t mix apples and oranges here. A simple metal shape does not have much in common with a ferrite or powdered iron core.
Most fly back transforms (if not any) i have seen where all glued together. This very common and nothing unusual.

Certainly, the introduced air gaps does cause some minor changes, which would increase the loses at higher frequency.
But this is probably not a problem here in a may be 50khz to 100khz buck regulator unit.
So again, i would try it, changes that you can get the same part or even figure out the value and original core material are very rare, unless equipped with expensive test tools.
So i think, trying to get replacement part would certainly pose the greater risk to be fully off the original values.
My advice is just try it, it should not hurt at all.

[linuxguru]

Re: Is this an Inductor?

You can safely glue that toroid with super glue - it will slightly increase the distributed air gap, thus reducing the inductance slightly, but only by a few per cent. There's space for larger caps at two locations, so those can be replaced by 20% higher values to compensate.

[PCBONEZ]

Re: Is this an Inductor?

An O has no air gap around the loop.
With a crack you have a C which does.
In this case you have more like ( ) with two air gaps.

The stacked plates in an xfmer are like stacking O upon O upon O.
The space between them is irrelevant to the situation here.
The path that is broken is in the other direction. - Around the loop, not across it.

If the continuous shape wasn't important they would use a bar | or a C instead of a doughnut O or manufacture the O in 2 halves and glue them to start with.
It would be easier and less expensive to make.

.

[linuxguru]

Re: Is this an Inductor?

The binder/adhesive that is used to bind the powdered iron or ferrite particles constitutes the distributed air gap. Sticking the toroid with super glue just increases it slightly.

[PCBONEZ]

Re: Is this an Inductor?

I follow what YOU are saying linux but I think gonzo is lost.
-
I still don't agree.
If a glued break was acceptable it would be less expensive to make them that way to start with,,, and they don't.

[kikkoman]

Re: Is this an Inductor?

I agree with linuxguru, too.
if it's glued properly, magnetic flux doesn't change much.
since this is not a transformer, the flux AROUND the core doesn't matter much anyway.
the reason why toroids are used here is the lower EMI. (many other designs use shielded cylindrical pot cores, but afaik they're a bit pricier.)

the only concern i can think of is mechanical stability. there is quite some force in that core, so super glue might be not enough. epoxy is probably the better choice.

@PCBONEZ: cores (coils) with a glued break aren't cheaper to manufacture. it means a lot more steps in the process.

FYI: there even are ring band transformers with a split core (i think they're called split-band cores).
they're used in cases where the machines can't handle the number of windings required.
in this case they wind a core, cut it in half, polish the facing ends and insert it into the pre-wound coils (and glue it in place).
electrical properties don't seem to change at all. the only disadvantage is the higher mfr cost of the core.

[PCBONEZ]

Re: Is this an Inductor?

Seems it to me it would be easier/cheaper to wind 2 halves then glue together then it would be to wind wire through a doughnut hole x many times. Of course they are probably made in China and an assembly machine consisting of the fingers of pre-teen Chinese girls probably comes very cheap.

I might glue one on a $15 boom box or something that's not important but never on anything that matters.

A LOOOOONG time ago (1990's) I did try to glue one on an old 486 'practice' board I'd given to my son to learn soldering on.
-
Super Glue did not hold/stick/stay on that one.
In fact trying to use it made the gap bigger because chunks of the core material were lost or dislocated by the failed glue and it wouldn't fit firmly together after the first gluing attempt.
-
Quite frankly desoldering and replacing them is FAR easier and less time consuming than fiddling around trying to position one for glue to set anyway.
.

[linuxguru]

Re: Is this an Inductor?

Perhaps JB Weld or any other two-component epoxy will ensure a stronger bond, compared to super glue. It can't hurt to try one of them.

[berniedd]

Re: Is this an Inductor?

Nope, gluing the pieces together will result in a LARGE change in inductance from the original value. Trust me on this.

[arneson]

Re: Is this an Inductor?

I agree based on common sense, that the ferrite is ruined and glue will not fix this, not at all.
The good news is that any good junk box of power supply parts should have a part just like this.
Even more so if you've been scraping monitors.

[gonzo0815]

Re: Is this an Inductor?

Well, i would not second that. Grabbing any ferrite core from your junk box, but whining over a little air gab in an glued core that is from my understanding the same roulette game.
There are many different ferrite and iron powder cores on the market, so that gluing together would probably be the most accurate rebuild then any else suggested here.

In the end it is both an try and error thing, as you simply can not identify the original material.
So fore me, this discussion is really a little academic.
Most cores from ATX PSU`s are powdered iron cores with distributed air gab. If we assume that this would be the correct replacement, then the small additional gab caused by super glue would not hurt to much.

But if you accidentally grab the wrong ferrite from your junk box, then you have way bigger problems, then reduced inductance and wider air gap.

If you put in an real ferrite core, made for frequencies way in the MHZ range and your Step down regulator was designed for 50 -100khz, the core will simply saturate, ending in a near complete loss of inductance.

If it is the other way round, losses would be astronomical high in the iron powder core, which would certainly lead to some thermal problem.
So my first try would certainly be super glue. If that would not work, then i would looking into the junk box for some replacements.
But i would certainly determine the switching frequency first and may be order an appropriate core.

[arneson]

Re: Is this an Inductor?

I know this is rudimentary:
http://hyperphysics.phy-astr.gsu.edu...ic/indtor.html
But imagine this broken into pieces, now imagine it glued back together.
I don't have a good enough imagination to believe I could put it back in and make it work.
I do have a huge collection, and if a toroid with coil matches up, color, size, guage and winding count?
I'm in there.