Renesas acquired IDT. This chip does all the clocks.
Highly likely its FS_A/B/C pins are directly wired to CPU socket - in that case it wont produce any clocks without installed cpu. That could be why you couldnt read any.
Clock range suggests something Pentium 4 era. I would say usual "capacitors", but not working warm is the reverse of bad cap symptoms 😐
Pentium 4 and up there are usually diagrams available, mobo model might help.

edit: ok so I started watching video https://www.youtube.com/watch?v=MPFw0hA9r-0 its Atom
edit2: @44:14 pin 1 is PCI0 and 33MHz shows PLL is running and locked fine. This chip generates some 3.3V clocks while the rest are low power differential at 0.8V. Im not sure if you can even measure that without proper differential probe or at least active FET one, with normal probe you might only see some very low voltage noise. What you can check is 0.8V VDD clock supply pins like pin 49 VDDCPU_IO 41 VDDCPUIO.

edit3: 56pin version datasheet https://www.alldatasheet.com/datasheet-pdf/vi … CS9EPRS525.html

I wonder what the U6 with a sticker is, not that I suspect it of anything.

U6 is Winbond.

You found the video 🙂

I have differential probes, do I need to check in between two signals? They have quite some noise at low level but might be ok to see what's going on.

Thanks for the pointers. I've got some ideas from viewers and some from me sleeping on it.

- Check rails with a scope. Someone mentioned 5V was unstable.
- Check Power Good from the LTC monitor
- Check BIOS address/data lines with scope: does the board even try to talk to it?
- LPC header. I understand it's giving me some diagnostics but I'm not sure how to read it. Is that a standard pinout?

Thanks!

tony359 wrote on 2026-03-26, 10:45:

U6 is Winbond.

Winbond what? 😀 it looks like some kind of LPC Super I/O.

tony359 wrote on 2026-03-26, 10:45:

I have differential probes, do I need to check in between two signals? They have quite some noise at low level but might be ok to see what's going on.

those clocks might be so weak and low level that measuring with ground lead clipped somewhere far away will just look like pure noise on the screen

tony359 wrote on 2026-03-26, 10:45:

- Check BIOS address/data lines with scope: does the board even try to talk to it?

bios is either lpc or spi, can be snooped on with logic analyser

tony359 wrote on 2026-03-26, 10:45:

- LPC header. I understand it's giving me some diagnostics but I'm not sure how to read it. Is that a standard pinout?

should be able to read bios post codes over lpc with pico https://github.com/MrGreensWorkshop/RasPiPico … sPostCodeReader but no beeps most likely means not running bios at all when its in fail state

edit: Tons of atom + 945GSE diagrams around, not exactly for this board tho. Somethign ~similar for example compal LA-5141P https://drive.google.com/file/d/13xrTGb03T1Wk … ydrw62A3gx/view https://www.scribd.com/document/126575685/Ace … p-Kav60-Rev-1-0

rasz_pl wrote on 2026-03-26, 11:38:

tony359 wrote on 2026-03-26, 10:45:

U6 is Winbond.

Winbond what? 😀 it looks like some kind of LPC Super I/O.

w8362 if I see right from the video.

tony359 wrote on 2026-03-26, 10:45:

I have differential probes, do I need to check in between two signals? They have quite some noise at low level but might be ok to see what's going on.

those clocks might be so weak and low level that measuring with ground lead clipped somewhere far away will just look like pure noise on the screen

Understood, I'll try with differential probes. But am I supposed to read a CPU clock of some sort somewhere?

tony359 wrote on 2026-03-26, 10:45:

- Check BIOS address/data lines with scope: does the board even try to talk to it?

bios is either lpc or spi, can be snooped on with logic analyser

I think I'll first ascertain if there is any activity whatsoever. If it's completely dead, then there is nothing to read.

tony359 wrote on 2026-03-26, 10:45:

- LPC header. I understand it's giving me some diagnostics but I'm not sure how to read it. Is that a standard pinout?

should be able to read bios post codes over lpc with pico https://github.com/MrGreensWorkshop/RasPiPico … sPostCodeReader but no beeps most likely means not running bios at all when its in fail state

Very cool, thank you!

edit: Tons of atom + 945GSE diagrams around, not exactly for this board tho. Somethign ~similar for example compal LA-5141P https://drive.google.com/file/d/13xrTGb03T1Wk … ydrw62A3gx/view https://www.scribd.com/document/126575685/Ace … p-Kav60-Rev-1-0

Thanks for the links!

tony359 wrote on 2026-03-26, 14:50:

rasz_pl wrote on 2026-03-26, 11:38:

those clocks might be so weak and low level that measuring with ground lead clipped somewhere far away will just look like pure noise on the screen

Understood, I'll try with differential probes. But am I supposed to read a CPU clock of some sort somewhere?

56pin version datasheet has proper pinout for cput/cpuc outputs, but since you can see PCI clock then fsb is most likely also ok

in case of laptops there is KBC chip, a microcontroller, managing everything. Controls keyboard, monitoring fans/temps/voltages/charging, powergood, sleep and reset. Here I dont know, I was hoping whatever is under the sticker would be that. Winbond w8362 <- last letter missing, maybe 7? W83627 is "LPC /IO"
https://www.alldatasheet.com/datasheet-pdf/vi … /W83627UHG.html
is not as smart as laptop KBCs - doesnt seem to have own firmware so would require BIOS to run first? but does have tons of monitoring suggesting fixed hardware management block running independent of main CPU:
PSOUT# 67 Panel Switch Output. This signal is used to wake-up the system from S3/S5 state.
RSMRST# 75 Resume reset signal output.
PSON# 72 Power Supply on-off Output.
PWROK 71 This pin generates the PWROK signal while 5VCC comes in.
OVT# 95 Over temperature Shutdown Output. This pin indicates the temperature is over the temperature limit. (Default after LRESET#)
CPUTIN 103 AIN Temperature sensor 2 input. It is used for CPU temperature sensing.
SYSTIN 104 AIN Temperature sensor 1 input. It is used for system temperature sensing.
KBRST 59 O12 Keyboard reset. This pin is high after system reset. (KBC P20)
LRESET# 30 IN Reset signal. It can be connected to the PCIRST# signal on the host.
CLKIN 19 IN System clock input, either 24MHz or 48MHz. The actual frequency must be specified in register. The default value is 48MHz.
PCICLK 21 IN PCI-clock 33-MHz input

I would check those pins, I havent verified if pins are for same pinout version like your chip.

I found a better picture 🙂

W83627DHG-A

https://www.alldatasheet.com/datasheet-pdf/do … /W83627DHG.html

I didn't realise it has a PowerOK pin as well.
I was looking for a reset OUTPUT but it only has a reset INPUT, right?

So the main voltage regulator IC has a power good signal, the Winbond has a power good signal. However the general Reset line seems to work as it goes LOW even when the board refuses to start.

If this was the c2000 issue, then I'd have to add a 100Ohm pull-up resistor between LPC-ClockOut and 3.3v, right? It's something I can try to see if it helps?

I need to make a list of things to check!
Thanks for now.

tony359 wrote on 2026-03-27, 10:19:

I was looking for a reset OUTPUT but it only has a reset INPUT, right?

it has RSMRST KBRST outputs, wheter they are used and how I dont know

tony359 wrote on 2026-03-27, 10:19:

If this was the c2000 issue, then I'd have to add a 100Ohm pull-up resistor between LPC-ClockOut and 3.3v, right? It's something I can try to see if it helps?

that would be PLC clockout of C2000 chip which you dont have here 😀 LPC is driven by chipset that is not known for going bad
winbond has some temp sensing and OVT. I would compare all the pins I listed in both good and bad states, but start with bios SPI to check if machine even attempts to start in bad state.