ASUS IPIBL-LB High side MOSFET shorted - what's next?

[Stefan Payne]

Hi

I've got an ASUS IPIBL-LB (out of an HP/Compaq thing) and there was a high side MOSFET shorted. I did somethings to eliminate it, was on the primary side of the VRM so I started removing the FETs. The first one I've removed was a bingo.

So without it short was gone and Board runs so far.

So question is what to do next?

a) replace the FET (I don't have and I probably have to order it oversees)
b) leave it as is
c) remove the rest of that phase (and what should I remove)

What do you think I should do??



Pictures attached...

[cpt.charlie]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

a) replace the FET (I don't have and I probably have to order it oversees)

I would take any other donor board and use one of its high side mosfet, if the new one fails, likely you have a problem with the VRM controller or mosfet driver.

b) leave it as is

If you use a low TDP processor this should work

c) remove the rest of that phase (and what should I remove)

In this case just remove the low side fets.

[Stefan Payne]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

I would take any other donor board and use one of its high side mosfet, if the new one fails, likely you have a problem with the VRM controller or mosfet driver.

Ah, that's what Louis Rossmann said about replacing VRMs...
That if they fail probably the controller for that also is fried...

Somehow I forgot that...

I've removed the low side FETs and also the Controller...

[momaka]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

So question is what to do next?

a) replace the FET (I don't have and I probably have to order it oversees)
b) leave it as is
c) remove the rest of that phase (and what should I remove)

Do either a) or c).

If going with a), you don't need an exact replacement MOSFET. As cpt.charlie noted, one from a donor board will do.

If going with c), some controllers may have the option to choose between driving a 2-phase, 3-phase, and 4-phase VRM, so check the controller's datasheet. If there is such an option, you might just need to change/add/remove a few SMD resistors to set the controller properly. Once set, it will run the remaining two phases a bit more smoothly, thus lowering the ripple on the output.

b) is not a good idea. Taking out the upper FET but not the lower one may cause serious damage (unless you have two upper FETs in parallel and you only removed one. Same goes for the lower FETs). Reason being is that if the FET driver does not realize that and pulses the lower FETs to turn ON, then a current will be drawn through the inductor from CPU V_core. Once the lower FETs are turned off, the inductor will try to keep that current going and force it through whatever other path it can. Typically, most FET drivers have phase detection though the input side of the inductor. So current is most likely to get forced through there. With no other return path that can definitely damage the driver IC (if it wasn't already).

Thus, NEVER run a synchronous buck regulator without the upper or lower MOSFETs.

Ah, that's what Louis Rossmann said about replacing VRMs...
That if they fail probably the controller for that also is fried...

I've rarely seen that happen.

From about ten MOSFET repairs I've done (mix of motherboards and video cards), only on two had the controllers/drivers damaged.

*edit*
If you don't mind a ghetto repair, I see there is enough space in the area to use a TO-252 MOSFET in place of that QFN FET. May need to run a free-standing wire to the gate, but it is doable.

[Stefan Payne]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

Well, now I've removed the lower side FETs. Haven't tested the Board again though.

And after thinkin about it, wouldn't removing the coil also be a decent choice, instead of the FETs?
Well, either way, the Board should work again with the Pentium Dual Core thingy...


Oh and the best thing:
After removing all that shit I've realized that I could have had a replacement from a GTX7900. Though something here also is shorted...

[momaka]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

And after thinkin about it, wouldn't removing the coil also be a decent choice, instead of the FETs?

Well, if you have shorted FETs, then they still need to be removed first. After that, it's a matter of choice if you want to remove the inductor or the FETs in the other phase. Either one will do.

I typically prefer to remove the FETs, as the coil by itself won't cause issues.

That said, again, see which controller you have. It might be just a matter of playing with a few SMD resistors to tell the PWM controller that there are less phases available (so the VRM will run smoother).

[cpt.charlie]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

Remember, it's easier to remove the fets than the coil.

[Stefan Payne]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

Depends on the Gear
With Hot Air: Absolutely.

Without Hot Air: Not so much.

You can remove the Coil sensible or destructive. In the last case all you need is a sidecutter.

[momaka]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

I remove all my MOSFETs without hot air.

The method I use is two irons: my 70 Watt temperature-controlled station (T12 tips, so excellent temperature regulation) and my general-purpose 30 Watt Radio Shack iron.

First I use the 70 Watt iron to add fresh solder to the tab and the legs of the MOSFET(s) I want to desolder. Then while heating the MOSFET tab with the 70 Watt iron, I use the 30 Watt iron to alternate between heating each leg on the MOSFET. Within a few seconds, I slide the MOSFET away, and that's it.

I find this to be easier and quicker than hot air, especially if you don't have a pre-heater (I do - the kitchen stove... but I rather not use it if I don't need to).

I also find removing MOSFETs easier than the toroid inductors, as those sometimes are placed in very tight holes on the PCB and hard to put back in.

As for using side cutters to remove the inductors - you don't always get easy access to the wires, and if it's a closed inductor (the newer ones with wire completely encased in the ferrite body), then that makes it impossible to use any kind of cutters to remove the inductor.

[a15995]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

Hi

I've got an ASUS IPIBL-LB (out of an HP/Compaq thing) and there was a high side MOSFET shorted. I did somethings to eliminate it, was on the primary side of the VRM so I started removing the FETs. The first one I've removed was a bingo.

So without it short was gone and Board runs so far.

So question is what to do next?

a) replace the FET (I don't have and I probably have to order it oversees)
b) leave it as is
c) remove the rest of that phase (and what should I remove)

What do you think I should do??



Pictures attached...

Hello Stefan!

I know this is an old thread but I think I have the same problem with a blown MOSFET. Can you post a picture of what you desoldered and which part number you replaced it with (specs)?

I have a computer (with IPIBL-LB) that will only turn on, if the 4-pin CPU power connector is disconnected (white socket in your pictures). I have changed everything from mobo to CPU and PSU with no luck.

Thanks,

/Søren

[Per Hansson]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

Read the thread again carefully and I think it will be quite obvious.
You can see what he removed in the pictures he posted, see here: https://www.badcaps.net/forum/attach...8&d=1500410288

But the probability that the same phase is shorted on yours is only 33% for obvious reasons
Sometimes it is possible to identify the shorted phase with an ohms check:
For example the "real" shorted one might show 0.00ohm while the others 0.01ohm...
Obviously good sharp probes and a good meter are a must...

[a15995]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

Read the thread again carefully and I think it will be quite obvious.
You can see what he removed in the pictures he posted, see here: https://www.badcaps.net/forum/attach...8&d=1500410288

But the probability that the same phase is shorted on yours is only 33% for obvious reasons
Sometimes it is possible to identify the shorted phase with an ohms check:
For example the "real" shorted one might show 0.00ohm while the others 0.01ohm...
Obviously good sharp probes and a good meter are a must...

Ahh, I see it now by comparing with a reference board. So, no visible signs of short...?

[Per Hansson]

Re: ASUS IPIBL-LB High side MOSFET shorted - what's next?

There might be, just carefully inspect it with a magnifying glass and a light in different directions.
But usually the overload protection is quick enough to turn everything off before smoke appears...