OK I pulled a leg of the resistor, according to the color code it's a 7.5 +/- 5% Ohms, measured it around 7.8 so it's still pretty much in specs. But if it's getting really hot, I suppose it may be receiving too much current?
Would that hint at a capacitor problem anyway?

I've never recapped a mobo before, but I have a good soldering iron and taught myself some soldering skills for the past 2 years, so that's a time as good as any to get going. Do you guys know of any further diagnostic step I may take to confirm a cap issue first?

There aren't that many caps, and they look pretty easy to desolder. So it would be a matter of desoldering all existing caps and check their ratings, and order replacements? What kind of caps/brand/type/whatever should I look for?

Thanks again for any help. I'm not ready to let it go.

You might check the voltage output of the VRM's on the board, to see if they are still in spec. Caps are less likely, but I am no specialist. Perhaps you had an unfortunate short or something else that killed - or hopefully just stunned - the board.

I would measure voltage from input and output of regulators as suggested above. I assume Vcore regulator is located top centre in the picture and output capacitors are close to processor slot. There is another regulator at bottom right corner.
Hear any beep codes if you remove CPU and/or RAM ?
About the slot you removed USB card from, I would check there's no crossed contacts.

I have P2B-F and some PIII CPU's here if you need supporting measurements.

Hey according your find it might be as99127f chip is dead

picture i found so any rights goes @ http://monitor.espec.ws/section5/topic184611.html

Thanks for the suggestions guys. But..unless I'm mistaken, I think I can't test voltages, again the board doesn't power up at all. The power LED on the board lights up, but switching computer on (or bridging ATX PWR pins as I'm currently on the workbench) has absolutely no effect. CPU fan doesn't ever spin, even for a microsecond. So of course that also means no beeps (I have a lone PC speaker attached to the speaker pins). I also have one of those PCI/ISA diagnostic cards, but again no power ever reaches the card so it's useless.

So the situation for me is a braindead mobo 🙁 it's getting some power, as the power led in the middle lights up when the PSU is powered on, but it doesn't react to any attempt to actually start it up. And I don't know if there's anything to troubleshoot at all in such a situation.

ciornyi wrote on 2022-02-26, 16:39:

Hey according your find it might be as99127f chip is dead

picture i found so any rights goes @ http://monitor.espec.ws/section5/topic184611.html

Thank you, went through that thread in google translate, but not sure if that helps as the guy was able to POST, unlike me?

My gut feeling is still about that resistor going very hot being a good hint. I have let the power supply turned on for 30s to 1 min while I was trying to do some reading with my multimeter, and when I touched it, it was EXTREMELY hot. That can't be normal.... but I'm unable to interpret it as a troubleshooting step.

fool wrote on 2022-02-26, 16:19:

About the slot you removed USB card from, I would check there's no crossed contacts.

Carefully checked everything visually and I really so nothing wrong anywhere 🙁 thanks for all the help so far anyway and hope there's still something to do.

i have one of these boards that worked great powered it off one day and then a week or two went by and when i went to use it again it refused to power on. nothing happened to in the time one from when it was working to when it stopped working. it wasn't even plugged in for that time.

i stripped it completely down and tired more then one power supply never could get it to power up again.

Green LED indicates that 5VSB is OK. So does the power supply fan rotate if you attempt power on?
Sounds like power supply protection circuit might be active due to short, some kind of. You could measure resistance from (empty) ATX-connector, wire positions of red-black, yellow-black, orange-black.

According some pictures resistor lower leg is GND plane. Transistor Q1 might be switch for "as99127f" which operates from 5V. Hard to follow just from a picture. Resistor could be for smoke protection if that chip fails, don't know. Looking at those traces near that resistor doesn't look meant for long term high current. Nevertheless that's about 1W resistor and there's some meaning it's rated like that high. Did you try poweing it on when resistor was pulled up? If that "as99127f" or resistor is the cause, mobo should power on resistor leg pulled.

yep psu not starting at all is due to protection, and hot resistor might suggest a short on one of the rails.
This is not due to bad electrolytic caps, they dont die shorted.

rasz_pl wrote on 2022-02-27, 00:11:

yep psu not starting at all is due to protection, and hot resistor might suggest a short on one of the rails.
This is not due to bad electrolytic caps, they dont die shorted.

Ceramic and tantalum caps can fail short. It can also happen with semiconductors. Try probing around with power off and a multimeter in continuity mode. If the short is confirmed, you can start lifting legs and test individual parts.

Mmh now that's a hopeful thought that didn't cross my mind. The short protection makes sense, instead of a critical component that is dead preventing any power on.

"Try probing around with power off and a multimeter in continuity mode" - happy to do that, I just wouldn't know where to start and how to somewhat be efficient and organized in my probing? Anyone would be willing to give me some pointers or a few instructions I can follow with my limited knowledge? Thanks again.

My issue is that I wouldn't really know what I'm doing, and may be getting continuity that is expected. I guess I need to test different voltage rails that shouldn't short, but no idea where and how.

fool wrote on 2022-02-26, 19:36:

Did you try poweing it on when resistor was pulled up? If that "as99127f" or resistor is the cause, mobo should power on resistor leg pulled.

Yes I tried with the ground leg out, and no difference. Mobo does not start.

Here comes what I would do.
Put multimeter to resistance mode, auto range or 1k should be fine. Touch 3.3V pin with red cable and GND point with black one. Resistance should start from low value and rise over to hundreds ohms and more, that just fine (resistance increases because multimeter is slowly charging capacitance in the circuit). Swap red and black wires around and repeat last step (black to 3.3V pin and red to GND point). If you get some value, close to zero, something is wrong.

Do the same for 5V and 12V pins.

If everything looks OK we can move to Vcore regulator in the green box.

edit: corrected mirror connector pin

Thanks man, much appreciated.

I believe the orientation of the connector is wrong on your picture and color-coding though...this should be more like it, right?

If that is correct, I am only reading a resistance with my multimeter on the 5V pins (all of them) and ground, and nothing on any 3.3 pins or the 12V pin and ground...not what you expected it seems?

Boohyaka wrote on 2022-02-27, 14:28:

Thanks man, much appreciated. […]
Show full quote

Thanks man, much appreciated.

I believe the orientation of the connector is wrong on your picture and color-coding though...this should be more like it, right?

24-pinP1.jpgatxpins.png

If that is correct, I am only reading a resistance with my multimeter on the 5V pins (all of them) and ground, and nothing on any 3.3 pins or the 12V pin and ground...not what you expected it seems?

nope fools scheme is correct anyway you measure for the short not voltage

ciornyi wrote on 2022-02-27, 14:49:

nope fools scheme is correct anyway you measure for the short not voltage

It wasn't, as he now corrected his to look like mine.

So yep now that it's confirmed, just tried measuring resistance again: all 5V pins I'm getting something on the multimeter (including pin 4 and 6), but nothing happens on any of the 3.3V nor the 12V pin, multimeter doesn't move at all.

As I have my other Gigabyte (working) board next to it, I tried the same and got the exact same result. 5V pins read something, nothing on 3.3 and 12V pins.

Another observation (maybe useless and/or irrelevant, again I'm in unknown territories to me but trying to put some method and logic into it): when measuring 5V and ground with my multimeter set to 2K:

- ASUS (non-working) measurement steadily rises and stops around .615
- Gigabyte (working) measurement steadily rises and stops around 1.180

If that is correct, I am only reading a resistance with my multimeter on the 5V pins (all of them) and ground, and nothing on any 3.3 pins or the 12V pin and ground...not what you expected it seems?

Well, sometimes I should look into mirror 😀 picture corrected.

It depens what do you mean with "nothing"? Is it zero or infinite/over range?
I get something like this: 12V=18kohm, 5V=1kohm, 3.3V=100kohm. I measured on ESD-bag which probably decreases resistances some amount as it's conductive material.

About Vcore regulator. Those two transistors are switching syncronously between 5V and 0V planes to Vcore output to generate correct Vcore value monitored by feedback. Shorted transistor would conduct 5V straight to Vcore or Vcore to ground. There are also other scenarios but those are the two what would possibly generate short.
I would do similar short check and measure resistance between [Vcore output <-> GND] and [Vcore output <-> 5V input].
You can also switch multimeter to diode mode and go through those legs of both transistors, all three of them. Thats 6+6 measurements. Diode mode measures the threshold voltage what's needed for current to start flowing in the circuit. I would assume to get readings like over 0.3... to something. Zero reading would be short circuit.

fool wrote on 2022-02-27, 15:54:

Well, sometimes I should look into mirror 😀 picture corrected.

No worries, but I was really confused by ciornyi's message 😁
I appreciate all the help, seriously.

It depens what do you mean with "nothing"? Is it zero or infinite/over range?

Let me show you, easier. First picture is when I set my MM to Ohms 2K. displays "1. " by default.
Then second picture when I try to measure a 3.3V pin, in this example. Basically, nothing happens on the display. Same for other 3.3V pins, same for the single 12V pin, and same on my other Slot1 working Gigabyte motherboard.

I get something like this: 12V=18kohm, 5V=1kohm, 3.3V=100kohm. I measured on ESD-bag which probably decreases resistances some amount as it's conductive material.

Weird..I'm on an cutting mat (that is isolated I suppose?), just added an ESD bag under the motherboards now and I still don't read anything for 3.3V and 12V, on both of my motherboards.
So that would mean I'm getting:

(V): (ASUS/GIGABYTE)
12V: null/null
5V: 0.6kohm/1.2kohm
3.3V: null/null

I also try setting the multimeter to 200 instead of 2K, but I am still not reading anything on any of the two boards...

About Vcore regulator. Those two transistors are switching syncronously between 5V and 0V planes to Vcore output to generate correct Vcore value monitored by feedback. Shorted transistor would conduct 5V straight to Vcore or Vcore to ground. There are also other scenarios but those are the two what would possibly generate short.
I would do similar short check and measure resistance between [Vcore output <-> GND] and [Vcore output <-> 5V input].
You can also switch multimeter to diode mode and go through those legs of both transistors, all three of them. Thats 6+6 measurements. Diode mode measures the threshold voltage what's needed for current to start flowing in the circuit. I would assume to get readings like over 0.3... to something. Zero reading would be short circuit.

Diode is same setting on multimeter as continuity, right? the "->|-" symbol?

Ok as my board is slightly different than your picture I just want to share this first to make sure we're talking about the same things, as I have many points that "continuity-beep" with ground.
Hope it's somewhat clear...I've numbered legs 1.1, 1.2, 1.3 for the top regulator and it's "backside" 1.4, and same logic for the bottom regulator but with 2.X instead.

So 1.1, 1.2, 1.4 and 2.1 beep with ground, while the others don't.

With that out of the way, would you be able to list the measurements you'd like me to do with that same nomenclature?

(V): (ASUS/GIGABYTE) 12V: null/null 5V: 0.6kohm/1.2kohm 3.3V: null/null […]
Show full quote

(V): (ASUS/GIGABYTE)
12V: null/null
5V: 0.6kohm/1.2kohm
3.3V: null/null

That looks like null is "over range" and results seems to be OK. Your looks like general garage level MM and results may vary when comparing.

Yes, layout is a bit different. Looks like the bottom transistor is rotated 90 degrees, overall it's the same circuit.

I meant about diode mode to measure between two transistor legs. So that makes 3 "pairs of legs" to measure in both polarities. That one pin in the middle, the short one, is same as large tab. Don't get me wrong, there's nothing wrong in the way you did it.

1.2 1.4 2.1 are connected together and that should make the output plane of Vcore. What does is show on the MM screen while measuring 1.4 to GND continuity and what if polarity is turned around? You can measure also resistance between Vcore<->GND. Would be better to remove the CPU first.

Red on 1.4 to Black on GND: MM beeps, and reads 006
Black on 1.4 to Red on GND: MM beeps, and reads 007

Those last 2 measurements have to be done with the CPU still connected, right?

Now removed CPU, and measuring resistance between VCORE and GND (so basically I can measure again between 1.4 and GND as well?):

Red on 1.4 to Black on GND: quickly rose up to 1.008 kohms, then kept rising but extremely slowly
Black on 1.4 to Red on GND: quickly rose up to 0.985 kohms, then kept rising but extremely slowly

Hope I'm following your instructions correctly!

Perhaps it is frequency bias or simply misplaced fear, but it seems like I am seeing more P3B-F's stop working lately. This kind of troubleshooting however is very useful to me and undoubtedly to others down the line, so thanks guys for documenting your steps and findings. I had the caps on hand, so I started a recap on my 1.04 board (not even top notch Rubycon/Sanyo caps last forever, I suppose). It's straightfoward, and not a ton of caps overall, as was mentioned earlier in this thread.

Nichicon solids will be beyond overkill but easy to get, and the board will probably outlast me now.

Here is what I am using in case it helps anyone - again I had these available from another project. Don't take this as gospel.
- For the Teapo 16V 100uF: Nichicon RS81C101MDN1
- For the shorter 8mm x ~12.5mm Rubycon 6.3V 1000uF: Nichicon RNU0J102MDN1
- For the taller 8mm x ~20mm Rubycon 6.3V 1000uF: Nichicon RNL1C102MDS1
- For the Sanyo 10mm 6.3V 1500uF: Nichicon RNU0J152MDN1