Asrock P4i945GC - component id

[N²x²N]

Issue:

I got this motherboard with a burnt trace and a burnt SOT-23.
On top it has M091 (I think it's a AP2309GN
https://www.alldatasheet.com/datashe.../AP2309GN.html).

This sot-23 connects to the mosfet that then powers the trace that got burnt (se pics).
It had a blob and a broken leg on the S-D path, G was spared but it shorts to both S and D.

Source is from +5VSB; Gate is +5V can't figure from where.

I can't identify, for sure, what this is and my only solution is to either get the .fz file or get lucky and have someone tip me on what it was. (I already posted in the schematics section)

I was suggested to try a 0.5 or 1A fast blow fuse on the S-D path and see what happens.

Thanks for any help!

(I reposted in this section as I mistakenly put it the same schematics request post)

[N²x²N]

Re: Asrock P4i945GC - component id

Pics

[N²x²N]

Re: Asrock P4i945GC - component id

****
I checked a little more and:

the dead SOT23 Drain connects to another sot23 that supplies Vcc to RT8105 (https://pdf1.alldatasheet.com/datash...EK/RT8105.html).

That component is responsible for BOOT and SOFTSTART.

My guess is that the blown SOT23 gives Vcc necessary to turn on the m/b, without it's dead.


The other two components the big mosfet connects to (red line on pics) are:
- https://pdf1.alldatasheet.com/datash...S/74ACT08.html (data I/O)

- https://html.alldatasheet.com/html-p.../st75185c.html (UART to SR-232)

These two, if not powered, maybe just won't allow communication from serial and or other I/O on the back of the board.
With a dedicated graphics card it should display something nonetheless.


Thanks for all the data, it's difficult to repair something if you don't know all that is around it.



Bottom line:
the +5VSB actually powers the soft-on power-on of the motherboard via the SOT23.
the +5V is the Gate allowing +5VSB to go to RT8105 and allow power-on ----> like a "power from ATX plugged, ready to power-on on demand"

"""" I'll post a pic with path. Visually bettet than words.

[N²x²N]

Re: Asrock P4i945GC - component id

pic rt8105

[N²x²N]

Re: Asrock P4i945GC - component id

With further continuity checks, the RT8105 has a signal IN (LGATE) from a parallel circuit that needs to be powered by the component that blew.

I think that Rt8105 is central for power on. ---

[momaka]

Re: Asrock P4i945GC - component id

Many of these older motherboards have a "switch" circuit to toggle how certain components / ICs get their 5V - it can be either from 5VSB or from 5V from the PSU (when the motherboard is turned On.) Generally speaking, most components and ICs will be powered from +5V from the PSU, unless they need to also work when the motherboard is in Stand-by or Soft-Off mode with certain features set (i.e. WOL, WOR, wake by KB/MS, and etc.) Also, when the mobo is in Standby, the RAM will need to have its V_tt line (about half of V_dd, or 0.9 to 1.05V) active in order to keep data stored in it. V_tt is usually generated from a small IC that normally gets is power either from 3.3V line or 3.3V_STB line (which itself is generated from the 5VSB.) However, the 3.3V_STB generator may be behind that switch circuit mentioned above.

Anyways, here's my suggestion on how to proceed:
1) remove all MOSFETs and BJTs connected directly to that burned trace, if possible.
2) draw out a map of what connects to what (in terms of power components) to try to figure out which component or section of the motherboard could be faulty and making the trace short out. Start from the 5VSB line at the ATX connector and follow it further downstream on the motherboard.
3) share what info you find here so we can what you're seeing.

As for the RT8105 - that appears to be your RAM's V_dd buck PWM controller. This is NOT what's responsible for making the motherboard turn on, but rather what controls the two MOSFETs (PQ2 and PQ_?) to generate the V_dd supply (1.8-2V for DDR2) for the RAM.

[N²x²N]

Re: Asrock P4i945GC - component id

Ok, thanks. I'll try to do as directed.

Is it possible that no power to DDR prevents the board from powering on?

Another user suggested to try putting a fast blow fuse 0.5A between +5VSB of the burnt component and the mosfets that power the circuit RT8105 is a part of. Should I? The porblem is that I'd still need to figure out what the smd that blew has to be replaced with.

Well, learning stuff again

[N²x²N]

Re: Asrock P4i945GC - component id

Partial.

I discovered that the other +5 to blown comp. comes from +5VSB too, but it goes through a cap and a resistor and then to the other leg (of the side that has 2 legs).

Basically the blown component gets +5 to both from +5VSB.

The burnt trace goes to the +5VSB/+5V 3-pin header selector. Indeed powering PS2 and USB (pin continuity with pin +5VSB from ATX)


BTW: there are no dead shorts on ATX nor anywhere else. Burnt trace could have been because of +V from USB?

The square component that gets Vcc from +5VSB is L6713A (https://pdf1.alldatasheet.com/datash...CS/L6713A.html)

[momaka]

Re: Asrock P4i945GC - component id

Is it possible that no power to DDR prevents the board from powering on?

No.

It will only prevent it from booting, but not from powering On.

Power to the DDR modules (V_dd... 1.8-2.1V) comes on only after the power button is pressed on the motherboard.

The active gate outputs of the RAM (V_dd) use about 10 Watts, which can be a bit too much for the 5VSB to handle alone, so RAM V_dd is usually powered either from 3.3V or 5V rail of the PSU. In this case, it looks like 5V.

Another user suggested to try putting a fast blow fuse 0.5A between +5VSB of the burnt component and the mosfets that power the circuit RT8105 is a part of. Should I?

No, that probably won't help yo troubleshoot much, if anything.
Also, most (decent) PSUs are fairly well protected on the 5VSB line, so if something is shorted, the 5VSB on the PSU will crowbar and shut down. So fuse is unnecessary.

Use a multimeter set to current measuring mode instead - preferably on the 5 / 10 / 20 Amp setting. Connect it between the 5VSB wire on the ATX PSU and the 5VSB wire on the ATX connector. Then, just monitor the current. If it goes much above 1.5-2 Amps, there may be something wrong or shorted.

BTW: there are no dead shorts on ATX nor anywhere else. Burnt trace could have been because of +V from USB?

Yes.

Check the USB ports for shorted, pushed, or bent pins. Same with PS/2. If none, remove jumper from PS2_USB_PWR1 jumper and verify that pin 1 has connection / low-resistance to 5V pins on ATX connector. Next, verify that pin 3 has connection / low-resistance to 5VSB pins on ATX connector. After this, verify that pins 1 and 3 DO NOT have connection / low-resistance to each other. Also verify that 5VSB on ATX connector does NOT have low resistance to 5V pins on ATX connector.

Once these are verified, put jumper back on PS2_USB_PWR1 on pins 2-3 (power for KB, mouse, and USB from 5VSB.) Now if you have a PS/2 keyboard, connect it to the PS/2 connector. Next, connect PSU to the motherboard and give it power. You should see Caps Lock light on keyboard toggle on and off when you repeatedly press the Caps Lock key. If you plug something in the USB port(s), it should get power too.

The square component that gets Vcc from +5VSB is L6713A (https://pdf1.alldatasheet.com/datash...CS/L6713A.html)

That's your CPU VRM controller - i.e. what generates V_core for the CPU.
Like RAM V_dd supply, the motherboard can start even if V_core is not present. It's just the motherboard won't boot / POST.
So probably don't worry worry about this one for now, it's unlikely to have gone bad and shorted.

Instead, have a look at this picture you posted again, just as a reference:
https://www.badcaps.net/forum/attach...8&d=1680908718

On this picture, there is a SOT-223 device between the two PCI slots. I suspect this is your 3.3V STB (3.3V standby) generator / linear regulator. On one pin, you should see 5VSB voltage, and on another, you should see 3.3V. This 3.3V should be present all the time when the motherboard is plugged in to the PSU and the PSU is connected to the wall.

If this SOT-223 device does not have 5VSB voltage, then your 3.3V STB generator might be the small transistor / regulator / MOSFET just to the right of the CLRCMOS1 jumper (right by the PCI-E connector end.) Check that one, and see if one pin has 5VSB and another 3.3V.

Whichever the case, you should see at least one such similar device generating 3.3V when there is 5VSB sent to the 5VSB pin on the ATX connector. I don't think I've seen a motherboard from that era not have a 3.3V STB voltage. Most use it for the audio, LAN chip, and SIO / LPC chip. A dead / shorted / light-struck LAN chip could be pulling the 3.3V STB line low, so check for that too. LAN IC should be the one to the left of the "U-TRON" coupler, right behind the LAN / USB ports stack. See if it's getting overly-hot or not.

If not, just repair the burned trace and try powering on the motherboard. Feel free to install something like a 2-3 Amps fuse across the burned trace, just so that it doesn't burn again, should there still be a fault.

[N²x²N]

Re: Asrock P4i945GC - component id

1) 3.3V Mosfet between PCI slots:
- +5VSB on one leg (left) ** +3.3 on (center) ** +2.08 on (right)

2) the ST75185C, close to USB ports, has a few data lines shorting to ground.
I'll desolder this one and see what happens.

3) I already repaired the burnt trace, it wasn't powering: before reapair, after repair. No change.

4) I tried all A measures but I only got -0.23 on some pins/legs.

Thanks again fro your help!

[N²x²N]

Re: Asrock P4i945GC - component id

I really can't understand the purpose of that sot-23 that died.

Update: +5V goes to Mosfet that follows to PS2 PWR select header


I found that the +12V rails of the 4-pin header has a D-S short to Ground.
I'll desolder one at a time to check.
https://pdf1.alldatasheet.com/datash...TD60N3LH5.html

[N²x²N]

Re: Asrock P4i945GC - component id

Desoldered the mosfets and caps but still shorting.
Maybe it's normal.

Ps2 and USB power is clear, no shorts. Light goes on but won't stay on. If I plug/unplug it'll light on again.

[N²x²N]

Re: Asrock P4i945GC - component id

From the bottom of my knowledge,

- if this SOT-23 has 2 +5VSB inputs: 1 PSU and 1 from a resistor path (RP)
- and one is a Gate

couldn't it be just that, a huge G? Only use is to detect if there is +5VSB

What if I just solder a copper wire from +5VSB PSU to output?

[N²x²N]

Re: Asrock P4i945GC - component id

If anyone reading this has this motherboard, could you please attach a picture of the Q748 sot-23, or the markings on top of it?

I believe that different markings are used for this component.

[N²x²N]

Re: Asrock P4i945GC - component id

If the datasheet shows that the AP2309GN,
https://pdf1.alldatasheet.com/datash.../AP2309GN.html,

has a diode, can I solder a diode from +5VSB to Drain?

Or just a bad idea of the uneducated?

[N²x²N]

Re: Asrock P4i945GC - component id

I soldered a copper wire from +5VSB to Drain of the sot-23.

1) board powered on when I probed the + of the OSC that goes to SB
2) it then allows RBT and PWR to work
3) removing the probe it may or may not work again
4) voltage is irregular, going from 0.04 to 0.28, if above 0.1 it'll power on
5) keyboard works and it POSTed + entered BIOS; if probe on OSC removed it'll freeze and if put back on OSC it'll start working again
6) by watching some YT videos I learnt the the OSC is essential for PWR and afterwards

Maybe a cap gone bad or the OSC has issues?
Value is 32768, tomorrow I'll put another from a working board.

[N²x²N]

Re: Asrock P4i945GC - component id

OSC wasn't the issue.
Board not powering on anymore.

No idea why it did, rtc lead isn't shorting anymore.

If anyone has anything to suggest....

Thanks for reading.