Help identifying component on computer motherboard

[keystone]

Hello everyone.  I'm looking at an old ECS PT800CE computer motherboard that won't power up.  Specifically I'm looking at the voltage regulator circuit area.  There seems to be a short somewhere.  I have limited troubleshooting experience but I do have a dmm and know how to use it.  One component in question is I believe a mosfet but not sure.  It has an ON Semiconductor logo on it.  I went to their website and looked for a part number but couldn't decipher the numbers on the face of the component.  There are 3 of these in the same area on the motherboard.  Any help identifying these would be appreciated.  Thank you for looking.

[retiredfeline]

Hmm, I put in T60N02R into a web search and it was immediately identified as a power mosfet by ON Semi.

[sean0118]

Yeah retiredfeline is right, the 437 is just a datecode, NTD60N02R is the part you want.

https://www.onsemi.com/pdf/datasheet/ntd60n02r-d.pdf

[cybermaus]

Shorts are quite often capacitors. And hard to pin-point using a DMM, because everything they touch appears as short as well.
Check if the FET measures as diode on some and open on other pins, or indeed it is shorted on some pins

But even if it does appear shorted, while it could be the FET, also suspect any capacitors or diodes linked to it.
Oh, and the nearby transformer also will measure as short, even though perfectly fine.

Barring better equipment like thermal cam, or tricks like seeing where alcohol evaporates first, you may need to desolder stuff and measure them out of circuit.
Always a lot of trouble, as you can do accidental damage while desoldering.

[keystone]

Yeah, that first number threw me.  The O's and zeros give me a hard time too.  Thanks for all the responses.  I think I will remove some stuff and check them out more closely.  I already think I have one shorted mosfet.  This mosfet is discontinued but there is a substitute so that's good.  I have a couple of other chips I need to investigate what they are and how I can check them and if there are any replacements available.  This is a learning experience for me.  Thank you for your responses again.

[xavier60]

On my working Asus P4P800S-X MB, the low side VRM transistors appear to all be short circuited. This is because they are JFETs rather than enhancement mode MOSFETs which the high side transistors are.

[keystone]

On my working Asus P4P800S-X MB, the low side VRM transistors appear to all be short circuited. This is because they are JFETs rather than enhancement mode MOSFETs which the high side transistors are.

Now I don't know anything about high side or low side.  I've read something about that but don't know about it.  Also about the enhancement mode mosfets.  I'll have to do more research on those topics.  I guess I'll just do a search for those things.  Things sure have changed from when I took an electronics class is high school.  We learned about vacuum tubes!  Transistors were the new technology then.  Thanks for your input.

[xavier60]

On my working Asus P4P800S-X MB, the low side VRM transistors appear to all be short circuited. This is because they are JFETs rather than enhancement mode MOSFETs which the high side transistors are.

Now I don't know anything about high side or low side.  I've read something about that but don't know about it.  Also about the enhancement mode mosfets.  I'll have to do more research on those topics.  I guess I'll just do a search for those things.  Things sure have changed from when I took an electronics class is high school.  We learned about vacuum tubes!  Transistors were the new technology then.  Thanks for your input.

The use of the JFETs is very unusual. I mention it because the readings will be very misleading to anyone who is unaware.
Like a vacuum tube, JFETs are On with zero Gate bias, needing negative bias to turn them Off. It's unlikely that your MB has them.

[rob77]

both high and low side switches of a buck converter on any motherboard are definitely mosfets

[xavier60]

both high and low side switches of a buck converter on any motherboard are definitely mosfets

Except when they are JFETs.

[rob77]

both high and low side switches of a buck converter on any motherboard are definitely mosfets

Except when they are JFETs.

can you name an example controller used on motherboards ? power JFETs are ususaly SiC - much more pricey than MOSFETS.. and considering the OP's case it's a motherboard from times when SiC was not leaving R&D labs yet.

but anyways.. i never encountered JFETs in vcore regulators (apparently was not repairing new enough stuff)... so please share the controller IC and JFET part numbers for such a regulator... and probably a product where it's used, i'm curious what benefits it might have.

[rob77]

both high and low side switches of a buck converter on any motherboard are definitely mosfets

Except when they are JFETs.

ok i call it bullcrap.... SiC is for high voltage and high frequency... and even if it was not, the lowest Rdson JFET/mosfet cascode i found has 27mOhm Rdson, it's a giant d2pak and costs more than 40$ in single piece quantitties...

so i stand behind my statement that ALL vcore regulators on ALL motherboards are using MOSFETs  

most probably you don't know what are you talking about and your "low side measures shorted" was the low impedance/resistance of the cpu/gpu core itself.. DC resistances below 50 Ohms are pretty common.. if you measure with diode-mode (usually at 1mA) then few ohms will show as dead short unless you have at least a 20000 count multimeter and even then it's just the last digit or last 2 in best case.

[xavier60]

With the CPU removed, I measure 0.8mΩ to ground.
I have just scoped the Gate drive waveform.
Do you have an Asus P4P800S-X MB there to test?

Ill see if I can get a clear snap of the part number.

[xavier60]

Nothing special about the driver, HIP6602. The drive is capacitively coupled into the Gates.
I couldn't find any mention of the JFET's part number when I last searched many years ago but got a hit today.
https://datasheetspdf.com/pdf-file/626050/Lovoltech/LD1010D/1
The fault turned out to be a failed HIP6602.
The MB is still in service.
There is some possibility that other model P4 main-boards also use JFETs.

[rob77]

it seems to be a oddball board then that lovoltech seems to be long gone.. datasheetarchive has 14 entries for lovoltech...  nowadays there are virtually no power JFETs apart from the silicon carbide ones...

btw.. can you post a photo of that mobo , just the vcore area with all related circuitry would be enough + the exact model and revision number indicated on the PCB silkscreen ? downloaded a boardview for P4P800 , and it's 3 phase vcore with ADP3180 controller, ADP3418 gate drivers and  13N03 high and double 06N03 low side mosfets..

btw.. i'm really curious how a JFET is driven with a HIP6602 mosfet driver.

[xavier60]

Yes, the Asus model is P4P800S-X rev 1.01.
The Vcore converter is biphase using HIP6302 controller and HIP6602 driver.
For drive level shifting, there is just a capacitor with parallel 1.2K resistor between the driver's output and the JFETs' Gates.
I can see where I had removed 2 of the JFETs before I realized that they were not shorted MOSFETs.

[rob77]

Yes, the Asus model is P4P800S-X rev 1.01.
The Vcore converter is biphase using HIP6302 controller and HIP6602 driver.
For drive level shifting, there is just a capacitor with parallel 1.2K resistor between the driver's output and the JFETs' Gates.
I can see where I had removed 2 of the JFETs before I realized that they were not shorted MOSFETs.

interesting.. so basically they just charge a capacitor through the gate's forward biased diode during the "high" period (jfet is open anyways..) and by switching to low it grounds the capacitor and it creates the negative potential to switch off the jfet.. basically a charge pump to get negative voltage.
and shoot-trough at startup seems to be prevented because the high side driver don't have a charge built up yet on it's bootstrap capacitor during the first cycle (jfet is open, so opening the high side during 1st cycle would shoot-through).

[rob77]

Yes, the Asus model is P4P800S-X rev 1.01.

looks like it was specific to revision 01. the boardview files i have are for A02 and A03 are both 3phase vcore with mosfets.

so i stand corrected , some boards did have jfets indeed , even if it was short-lived and kind of niche but it did exist    

and of course i owe you an apology , sorry for being an arrogant dickhead not believing you.
but at least i learned my lesson here

[keystone]

That is exactly what my ECS PT800 CE motherboard has in the vrm!  This motherboard is from around 2004 or so.  Maybe I should check that controller chip.

[xavier60]

That is exactly what my ECS PT800 CE motherboard has in the vrm!  This motherboard is from around 2004 or so.  Maybe I should check that controller chip.

Wow! that is a coincidence.
My MB diagnostic skills are limited. What have you checked so far, like the Standby 5+?
Are you testing the bare board on the bench or still in the PC case?

And this whole time, part of a JFETs number is visible in your first post. There is some suspicious looking solder on its Source pin?

[keystone]

Xavier60, I probably have a more limited skill at diagnosing motherboards than you.  I don't know how to check the standby +5v, where to find it.  I am testing the bare board out of the pc case.  That solder joint is from me trying to reheat the solder to see if there was a bad solder joint at any of the pins.  There are 3 of these jfets and 3 mosfets in the vrm.  I guess that makes it a 3 phase system.  Where would I find the mosfet driver and controller?  Maybe that is what my problem is with the board.  Something seems to be shorted.  When I plug in the 4 pin "+12v connector, I get a big arc and the ps shuts down.  I unsoldered the 3 electrolytic caps there and they checked okay with vom.  They are Osd brand, 6.3 volt, 3300 mfd.  I need to do more checking but not sure where to look.  I sure wish I had someone to physically show me how to check this out and how to hook up a scope to check things.  It's hard for me to just read about it and then I have questions.  Glad you got your motherboard going again Xavier60.

[xavier60]

Xavier60, I probably have a more limited skill at diagnosing motherboards than you.  I don't know how to check the standby +5v, where to find it.  I am testing the bare board out of the pc case.  That solder joint is from me trying to reheat the solder to see if there was a bad solder joint at any of the pins.  There are 3 of these jfets and 3 mosfets in the vrm.  I guess that makes it a 3 phase system.  Where would I find the mosfet driver and controller?  Maybe that is what my problem is with the board.  Something seems to be shorted.  When I plug in the 4 pin "+12v connector, I get a big arc and the ps shuts down.  I unsoldered the 3 electrolytic caps there and they checked okay with vom.  They are Osd brand, 6.3 volt, 3300 mfd.  I need to do more checking but not sure where to look.  I sure wish I had someone to physically show me how to check this out and how to hook up a scope to check things.  It's hard for me to just read about it and then I have questions.  Glad you got your motherboard going again Xavier60.

The IC near the 12V connector could be the controller/driver. Get its part number.
What you have said implies that MB's PWRSW pins have been momentarily bridged causing the MB to turn on the PSU's main rails including the 12V as that's the only way there can be an arc when it's plugged in. Not a good idea to do that BTW. Even accidentally shorting the wrong MB control panel pins  can be fatal. I use a graphite pencil although this can cause its own problems if to much graphite falls onto the board.
Does the CPU fan run?

Google "atx power supply pinout"

[rob77]

When I plug in the 4 pin "+12v connector, I get a big arc and the ps shuts down.  I unsoldered the 3 electrolytic caps there and they checked okay with vom.  They are Osd brand, 6.3 volt, 3300 mfd.  I need to do more checking but not sure where to look.  I sure wish I had someone to physically show me how to check this out and how to hook up a scope to check things.  It's hard for me to just read about it and then I have questions.  Glad you got your motherboard going again Xavier60.

most likely your VRM is shorted that's why the sparks and shutdown of the psu, check the high side mosfets Q44, Q46 on your picture... measure the resistance between drain (the big tab) against ground. if it's close to short then one of the high side mosfets (or more of them ) are shot through. also check the lowside jfets gate-source should look like a diode, also check the small capacitors on the jfet gates of those are shorted the driver can't create negative voltage to close the jfet.

if the vrm is shorted try desolder the high side mosfets one-by-one and check if the short is gone.. if all mosfets are gone and short is still there , then desolder the ceramic smd capacitors. there are faster methods but for someone with not enough skills this is the easiest. and of course you will need a hot air station (the cheapo chinese is good enough) to desolder the mosfets, there is no other safe way.

btw.. VSB (+5V always on is thh purple wire coming from ATX supply... and poweron is the green one - needs to be shorted to ground to power on the PSU), but your first problem to solve is definitely the shorted VRM.

[keystone]

The chip next to the 4 pin 12 volt connector is an rs232 driver/receiver.  I'm referring to U5 in the pic attached.  I'll do more testing on this tomorrow as it is late here now.  Thank you both for the guidance and help with this.  I do have a cheap chinese hot air setup.  I'll try that to unsolder the mosfets.  I tried a solder iron but that didn't work.  I really appreciate your help guys.

[sean0118]

I do have a cheap chinese hot air setup.  I'll try that to unsolder the mosfets.  I tried a solder iron but that didn't work.  I really appreciate your help guys.

Try adding some leaded solder if you haven't already, it makes it much easier to de-solder anything lead free.