Asus P2B | Slot 1 + Pentium II | no power

[AidenX]

Hello!

I would like to repair this old motherboard that i bought weeks ago with some other old stuffs!

Sadly it doesn't work.

I tested the CPU, ram, etc with another slot 1 board and they are all fine.

Probably there is a short on the board, no power at all when connected to the power atx supply.

Also, the power supply works fine with other hardware, but do a strange noise after few seconds that is connected to this board...i think is all ready to explode! XD

Here some measures from the power connector on the board ( + to pin, - to ground):

#PIN / Resistance / Voltage on diode mode

1  OL
2  OL
3  0
4  430      0.438V
5  0
6  430      0.438V
7  0
8  15.47k   2.743V
9  13.67k   2.291V
10 38.02k   2.514V
11 OL
12 590.6k   0.677V
13 0
14 24.49k   2.880V
15 0
16 0
17 0
18 4.937M    OL
19 430       0.438V
20 430       0.438V

The area between ram slots and cpu has the mosfet Q9 and Q10 (and few capacitors near them) with a really low ohm value and the multimeter light red (in continuity mode).

See the photos (in the last one i did some marks), and say me what tests i have to do!

Thanks for the help! ^^

[Dan81]

I'll probably be scolded for this:

Test the MOSFETs with beep mode - if they beep when testing either pin, then they might have gone bad. To be perfectly sure, desolder them. If they beep again, then you're lucky and they're the only ones to be dead, hopefully.

Failing that, you might have a dead southbridge - if the SB gets very hot to the touch, then it clearly went bad.

[AidenX]

Originally Posted by Dan81

I'll probably be scolded for this:

Test the MOSFETs with beep mode - if they beep when testing either pin, then they might have gone bad. To be perfectly sure, desolder them. If they beep again, then you're lucky and they're the only ones to be dead, hopefully.

Failing that, you might have a dead southbridge - if the SB gets very hot to the touch, then it clearly went bad.

Oh....really bad news so

I tested again Q9 and Q10 (without desolder them for now).

Source / Drain / Gate of Q10 doesn't beep....

Q9 is strange...has a different name/mark from all the other mosfets.... and it's normal that both drain and gate go to ground ?!

Anyways again, source to drain or gate has that low resistence of 36 ohms (and why the multimeter light is red ? O.o).

Also, i can't test if the southbridge get hot...as i said i can only left the power on for few seconds....after that the power supply does a strange noise...like it's ready for explode. XD

There is something else that i can do before start to desolder things?

Thank you!

[piernov]

https://www.badcaps.net/forum/showthread.php?t=88758

Additionally, some places will measure low resistance to ground under normal operation. In general, high power low voltage (<2V) power rails will measure less than a few hundred ohms to ground.

[Per Hansson]

Originally Posted by AidenX

The area between ram slots and cpu has the mosfet Q9

Originally Posted by AidenX

Q9 is strange...has a different name/mark from all the other mosfets.... and it's normal that both drain and gate go to ground ?

Q9 is not a Mosfet, it is a Schottky diode
Test it with the multimeters diode test function, you should have a drop of ca 0.2v from pin 3 (anode) to the tab (cathode) and open in the other direction.
See attached datasheet.

[AidenX]

Oh....sorry for the mistake... ^^"

I tested Q9 (still soldered on the board) as you asked:

+ to pin 3 and - to the tab = 0.042V

- to pin 3 and + to the tab = 0.044V

Same values if i test pin 1 with the tab...and again that 36 ohms (with red light) in continuity mode...

Thanks!

[Per Hansson]

Desolder it (or at least lift the legs) to test it out of circuit: it does look shorted with those values.

[AidenX]

Originally Posted by Per Hansson

Desolder it (or at least lift the legs) to test it out of circuit: it does look shorted with those values.

Hello....i did it (disconnected only pin 1 and 3, tab still connected).

+ to pin 3 and - to the tab = 0.115V

- to pin 3 and + to the tab = nothing

pin 1 to tab is nothing in both sides.

But if i test the PCB part of pin 1 / 3 with tab, it's all like before....

The problem is somewhere else right?

Thanks...

[Per Hansson]

Yes, that means the problem is somewhere else.
If you have a high resolution multimeter and sharp probes you can sometimes find the short by looking for the lowest ohms reading.
If not there are some projects where you can build one, called milliohm meter.
You can also try to inject a very low voltage, like 0.5v but at high amperage (a few amps) and see what gets hot, a thermal camera helps in this case...

[AidenX]

I just have a cheap digital multimeter....and i don't have a DC power supply anymore for the voltage injection. I will buy another one soon.

I will also try to ask the seller more details about the story of this board.

Say me if i can do something else in the meanwhile...

[Per Hansson]

From the numbers you provided of your meter it has at least three digits so you can just look for the lowest possible ohm reading on the board...
Would be easier with more digits but maybe still possible...
Here a good example: https://www.youtube.com/watch?v=L4V3BWReZLY

[AidenX]

Ok i will try.

There are many capacitors around the board with about 36 ohms to ground (and the red light)...
This means they are all connected to the same shorted line or something like that?

[piernov]

Again, low resistance reading can be normal on power rails powering main chipset/CPU/GPU/etc. power rails.
Need to compare with a board with a similar architecture to be sure.

[AidenX]

Originally Posted by piernov

Again, low resistance reading can be normal on power rails powering main chipset/CPU/GPU/etc. power rails.
Need to compare with a board with a similar architecture to be sure.

Hello!

I understood this, and I also read the guide when I tried to fix the notebook motherboard here on the forum.
I had found various low values just near the CPU/GPU etc, even from 1-3 or 0.x ohms and in that case they were normal as you mentioned.

On that occasion, however, it had never happened that the multimeter red indicator lights up due to a low resistance, which is why I asked several times if it was normal or not! XD

These days I have been looking for the manual of my multimeter where it is specified that this just happens when (in continuity mode) the resistance is between 30 and 60 ohms.
I hadn't noticed it yet, so I'm sorry for the misunderstanding. ^^"

Anyway, i only have another Slot 1 board, from a different manufacturer (QDI) but with the same 440BX chipset. See photos below.
It doesn't have the mosfet and that "special diode" like the Asus, but i tested the capacitors near the same area.

There isn't anyone with a so low resistance (and the red light on!), most of them have about 240 ohms to ground.

Only few have a less value of 80-90 ohms (like some of the asus capacitors).

I also compared the ATX Power supply connector of the two boards.

They are 20-pin connector, so i numbered them using the scheme here: https://en.wikipedia.org/wiki/ATX#Power_supply.
PIN 11-12-23-24 are missing for this reason.

             Asus P2B                 |   QDI working board

 #PIN | Resistance | v. diode mode

1         .0L           .0L           |  241.3        0.249V  
2         .0L           .0L           |  241.3        0.249V
3           0                         |      0
4         430        0.438V           |  452.5        0.449V   
5           0                         |      0
6         430        0.438V           |  452.5        0.449V
7           0                         |      0
8       15.47k       2.743V           |  10.28k       1.166V
9       13.67k       2.291V           |     74k       1.542V
10      38.02k       2.514V           |   54.4k       1.826V

13        .0L           .0L           |  241.3        0.249V
14      590.6k       0.677V           |  48.68        0.578V
15          0                         |      0
16      24.49k       2.880V           |   78.7k       1.785V
17          0                         |      0
18          0                         |      0
19          0                         |      0
20      4.937M          .0L           |  32.51k       2.936V
21        430        0.438V           |  452.5        0.449V
22        430        0.438V           |  452.5        0.449V

  ___________________________________________________________

  PIN 1-2-13     NOT linked           |  PIN 1-2-13    linked
  PIN 4-6-21-22      linked           |  PIN 4-6-21-22 linked

The +3.3V are totally missing in the ASUS board. I read that some early atx board can not have the +3.3V, so maybe is normal...
The +5V seems fine.
I'm not sure about the other lines, the values are quite different (+12V etc..).


Last thing, i got some information from the seller.

The motherboard was working until a power loss on the +12V.
The board was ON when ONLY the yellow power supply cable (connected to pin 10) came out of the plastic connector (wtf?) and the pc turned off.
After that the board was (and still is) dead.

That's all, i hope there is something useful this time, and again thanks for the help!

[Per Hansson]

Well simply based on what the previous user said and the fact that you say your power supply does not switch on it means we suspect a dead short somewhere.
So what you can try based on what he said is to disconnect the +12v wire and see if the system will then power on, if it does you know the short is on +12v.

[AidenX]

Originally Posted by Per Hansson

Well simply based on what the previous user said and the fact that you say your power supply does not switch on it means we suspect a dead short somewhere.
So what you can try based on what he said is to disconnect the +12v wire and see if the system will then power on, if it does you know the short is on +12v.

So i tested the board as you said, without the +12v wire on PIN #10, and the power supply turns ON.

But the voltage on PIN #10 of the motherboard (not connected to the power supply!) isn't 0V....but 3.84V.

It's not normal...right?

[Dan81]

Probably a short down the 12v line that in turn pulls the voltage down to 3.84v, I guess.

I'd trace the whole 12v line of the mainboard and check for shorts, or what ever is pulling the line down.

[Per Hansson]

Does the board actually work without the +12v?
I mean did you have a monitor hooked up also and a PC speaker to listen for any beep codes?

Amazingly I found a boardview for this board, attached to this post.
Use Openboardview to view it: https://github.com/OpenBoardView/OpenBoardView

You can select to view the 12v net, I would start at Q13 near the edge of the board where the front panel connectors are.
12v seems to be used for all chassis fans, D14 seems to be missing?
After that I'd look at:
U6 hardware monitor Winbond W83781D
U2 & U3 RS232 line drivers Holtek HT6571

It is not a perfect match, if you look at your board there should be a matching designation somewhere.
If you Google that maybe you can find a better match but really these seem good enough.

If you can't get anywhere you could hook up a current limiting resistor in series with the 12v rail, lets say 20ohm would limit you to 0.6A.
That way the PSU will no longer shut down and you can look for what component (other than the power resistor) gets hot.

[AidenX]

Originally Posted by Per Hansson

Does the board actually work without the +12v?
I mean did you have a monitor hooked up also and a PC speaker to listen for any beep codes?

Negative. I tried the board with only a fan connected (CPU_FAN) and the speaker (without cpu, ram, etc).
The fan didn't work and 0 beep.

Originally Posted by Per Hansson

D14 seems to be missing?

In my board revision (1.02) D14 and D15 (with the CHA_FAN between them) are both missing.

I will see the boardview later! Again thanks for the help and all the info!

[Per Hansson]

Originally Posted by AidenX

Negative. I tried the board with only a fan connected (CPU_FAN) and the speaker (without cpu, ram, etc).
The fan didn't work and 0 beep.

You're not going to get anything without a CPU installed.
If you leave the memory out the system will stop very soon at memory init, better to have that installed too.
It would be interesting to see if the board does work without the +12v.
None of the devices connected to +12v seemed critical IMO.