[SOLVED] Question about motherboard inductor

Aug 7, 2019
5
0
10
I know that the answer is probably "obviously yes" but....

I got an old system with a P67 motherboard and managed to get it up and running okay. I was in the process of stripping it down to clean when I noticed that one of the inductors appears to have been chipped; I think it's a ceramic block which covers a copper coil, and a big chunk of the ceramic is gone exposing the coil.

Picture of inductor

The question is whether having an exposed coil like this is a danger to either me or the part? I have another motherboard with broken pins which has an R20 inductor but I'm not confident enough to carry out soldering on a motherboard, so I might be better selling the two motherboards for someone else to do the repair, or would it be okay to just use it with the exposed coil?
 
Solution
Well I've got the motherboard back in and both channels are working. I've run a memory benchmark on each and they're both managing around 18.5GB/sec. One of them is 0.2 faster, and has a latency score of 9.6ns compared to the other's 12.7ns.

Any suggestions for testing it? I've run Prime95 for a little while without any mishaps. It's really just a gaming PC, so is there any harm in just using it? I've messed around with different speeds of memory and really didn't see any major advantage of using 2133MHz over 1333MHz, so I think I might just put my 1600 in and see how it goes. If games don't crash then I think that's probably good enough.

On the off-chance it's not a memory VRM at all you could also test other on-board...
Yeah, that will definitely affect the inductor's performance. I think you'd be 'safe', but voltage stability and noise will suffer, meaning of course your computer might be unstable too. It looks to me like it might be for memory so you may be in luck if you use only the channel that inductor's VRM doesn't power.

Even if it's not a memory VRM, it's certainly not a CPU Vcore VRM. So it could be the device it powers isn't particularly sensitive to voltage instability or noise. All you can really do is power it up and test it out in realistic use-age.

What motherboard is that anyway? a full size image might help figure out how severe the loss of stability on that VRM really is.
 
Last edited:
Aug 7, 2019
5
0
10
Yeah, that will definitely affect the inductor's performance. I think you'd be 'safe', but voltage stability and noise will suffer, meaning of course your computer might be unstable too. It looks to me like it might be for memory so you may be in luck if you use only the channel that inductor's VRM doesn't power.

Even if it's not a memory VRM, it's certainly not a CPU Vcore VRM. So it could be the device it powers isn't particularly sensitive to voltage instability or noise. All you can really do is power it up and test it out in realistic use-age.

What motherboard is that anyway? a full size image might help figure out how severe the loss of stability on that VRM really is.

Thanks for the reply. It's an Asus P8P67 LE. My camera isn't great, nor my lighting, so a picture from elsewhere might give better detail, but here is a pic.

I bought it as a "for parts" system and managed to coax it into life with the 2600k that was in it. At first I thought that it was just dirty as the whole system is stinking, but then I noticed the wire showing.

I did try different DIMM slots before I got it to post, and did have the memory led on at one point. Once I've fumigated the case I can put it back in and can try to identify if it only boots with one channel.
 
Thanks for the reply. It's an Asus P8P67 LE. My camera isn't great, nor my lighting, so a picture from elsewhere might give better detail, but here is a pic.
...
can't tell much because all the pic's I can find are low-res. But it appears to me theat other VRM circuit, next to your broken inductor, is NOT the same type as the one where your inductor is...the inductors are different sizes, for instance. So that strongly suggests it's not like one VRM was for one channel, the other for the other channel. So now I'm not sure what that VRM is powering: it could be memory, but the fuzzy pictures don't really help to find the other.

That board looks high-end enough to not skimp with one VRM for all 4 DIMM's, but then I don't know Intel boards very well either.

At any rate, all you can really do is try to get it to work. Just be flexible to moving DIMMs about to see if it helps. The important thing is it's not going to be disastrous, just unstable. The loss of the ferrite material surrounding the inductor coil means the inductor values have shifted dramatically and so the VRM won't perform well...high ripple, incorrect switching frequency, unstable voltage. The device it powers may not have sufficient tolerance to handle it; memory and CPU's don't.
 
Aug 7, 2019
5
0
10
can't tell much because all the pic's I can find are low-res. But it appears to me theat other VRM circuit, next to your broken inductor, is NOT the same type as the one where your inductor is...the inductors are different sizes, for instance. So that strongly suggests it's not like one VRM was for one channel, the other for the other channel. So now I'm not sure what that VRM is powering: it could be memory, but the fuzzy pictures don't really help to find the other.

That board looks high-end enough to not skimp with one VRM for all 4 DIMM's, but then I don't know Intel boards very well either.

At any rate, all you can really do is try to get it to work. Just be flexible to moving DIMMs about to see if it helps. The important thing is it's not going to be disastrous, just unstable. The loss of the ferrite material surrounding the inductor coil means the inductor values have shifted dramatically and so the VRM won't perform well...high ripple, incorrect switching frequency, unstable voltage. The device it powers may not have sufficient tolerance to handle it; memory and CPU's don't.
Okay, so do you think my idea of taking an R20 from another board would work? It might be easier to just sell it on like that if so
 
Okay, so do you think my idea of taking an R20 from another board would work? It might be easier to just sell it on like that if so
I think those inductors have heavy gauge leads that are through-hole mounted. It should work OK, the only problem is the heavy lead gauge and (probable) 2 oz copper power and ground planes means it will take a lot of heat so be careful not to damage the board. I'd cut away the bad inductor with diagonal cutters then unsolder the remaining lead stubs to allow the least dwell time of the iron on the board, just enough to wet the solder and remove them.

Also: are you sure that's an R20 Inductor? I really don't know how you can tell with the top completely gone!
 
Last edited:
Aug 7, 2019
5
0
10
I think those inductors have heavy gauge leads that are through-hole mounted. It should work OK, the only problem is the heavy lead gauge and (probable) 2 oz copper power and ground planes means it will take a lot of heat so be careful not to damage the board.

Also: are you sure that's an R20 Inductor? I really don't know how you can tell with the top completely gone!
It's an R20 by looking at various pics of the same motherboard on eBay, and the same position on a different but similar ASUS mobo, and all the ones that size seem to be the same.

Don't worry I won't be doing it, I would offer it for people who know what they're doing and with the caveat that they need to confirm that it's an R20. And it'll depend on the results of the testing which I'll do first.
 
Aug 7, 2019
5
0
10
Well I've got the motherboard back in and both channels are working. I've run a memory benchmark on each and they're both managing around 18.5GB/sec. One of them is 0.2 faster, and has a latency score of 9.6ns compared to the other's 12.7ns.

Any suggestions for testing it? I've run Prime95 for a little while without any mishaps. It's really just a gaming PC, so is there any harm in just using it? I've messed around with different speeds of memory and really didn't see any major advantage of using 2133MHz over 1333MHz, so I think I might just put my 1600 in and see how it goes. If games don't crash then I think that's probably good enough.
 
Well I've got the motherboard back in and both channels are working. I've run a memory benchmark on each and they're both managing around 18.5GB/sec. One of them is 0.2 faster, and has a latency score of 9.6ns compared to the other's 12.7ns.

Any suggestions for testing it? I've run Prime95 for a little while without any mishaps. It's really just a gaming PC, so is there any harm in just using it? I've messed around with different speeds of memory and really didn't see any major advantage of using 2133MHz over 1333MHz, so I think I might just put my 1600 in and see how it goes. If games don't crash then I think that's probably good enough.

On the off-chance it's not a memory VRM at all you could also test other on-board services provided: like LAN, audio, WiFi and the various features of the PCH, e.g., SATA ports and USB ports.

Other than that I can't really think of anything else to do.
 
Solution