Motherboard issue - all voltages OK but no display

[gurj1979]

I've got a couple of Intel and AMD boards which turns on fine, all voltages at all coils are present, but there is no display.
I guess this like Code 00 issue.
I've tried a couple of good CPU's and working Graphics Card.
PCH has been replaced too. On one Intel board I have even replaced the CPU Socket.
The CPU's don't get hot, but gets a bit warm.
All RAMs are good.
BIOS is good on all.
PSU is Antec 1000W and I use this for on more powerful boards and 2080Ti cards, so this isn't the issue.
What's the issue when motherboards have this issue ?
Must be some sort of common issue on all boards which causes this ?
Any sort of signals missing / required by the CPU?

[gurj1979]

Re: Motherboard issue - all voltages OK but no display

You've just stated all the obvious stuff, most what I've already mentioned.

I've tried BIOS updates.
I'm aware of chipsets being faulty, and I've replaced these already. And on some I have also replaced the SIO and reprogrammed.

How can I test if a capacitor has failed (I' cant find any short circuit on any of the caps around the CPU Sockets). I don't want to inject voltage to it because I don't want to damage anything.

[DynaxSC]

Re: Motherboard issue - all voltages OK but no display

In such a case the root cause analysis is not easy and can take really a lot of time, even several days. Code 00 means CPU does not even start.

To find the root cause without hoping for luck you need to have a scope and look for a power sequence diagram and a boardview at minimum, having a schematics can help also a lot. Then you need to trace state by state what is happening in the diagram, and where the sequence is interrupted, which signals are incorrect. Then you need to find the reason for it, it's not easy as this is a process in time, there are a lot of dependencies, and requires sometimes a massive brain CPU power to understand what is the real root cause of the problem. With this method I found once a broken track on the motherboard, it was a small scratch, I would'nt expect to give an broken connection. Powersequence diagrams are quite similar for one MOBO vendor/model line, or even for similar CPU/Chipset generations, so you can use often a similar one, if you do not find one directly for the model you have.

Useful hints:

Does the CPU VRM generate VCORE, do you have VCC_SA and VCC_IO present ?

Do you have all RAM voltages present (3,3V, 2.5V, 1.2V, 0.6V) ? With RAM often the 0,6V VTT_DDR is missing due to a small VRM or lin. regulator bad, mostly the 0.6V is generated from the 1.2V. VCORE may not come even if CPU VRM is EN-abled if 0,6V not present.

Sometimes also some low resistance resistors tend to decalibrate or even blow under load with time, especially serial 0-2,2 ohms resistors in VRM's, check all of them for correct values. Capacitors go very seldom broken.

Check Socket Occupied signal on SIO, name will be something like" SOC_OCC (some boards, but not all, will not start if this signal is inactive, ie. CPU not present, or signal interrupted) - you can test this activating the signal artificially on the SIO.

Check LAN chip and all USB ports for shorts, they are exposed to electrostatic spikes/surges, and often go broken (all ports are protected by EDS devices, check them for shorts, for USB usually the ESD devices are able to protect the PCH or USB controller and go short first before PCH/USB controller are demaged, but not always). A short on USB/LAN can impact capability of PCH/LAN chip to communicate/work properly, however I guess this probably won't be the reason for CPU not to start.

Check DMI bus between CPU socket and PCH for shorts, open lines (in reversed diode mode, reading must be ca 0.3V). For older LGA'a (eg. 1151) you can buy a dedicated tester on Aliexpress, up from Z490 you need to measure it very carefully pin by pin in the CPU socket.

Check other communication buses on the motherboard, eg the one between PCH and RAM, CPU VRM controller and CPU, and others (PCH). Shorts or open lines here may also prevent board from starting.

Check if all clock signals are present. If you buy the DMI CPU tester on Alexpressi, then you'll get also a second small pcb fitting into the cpu socket allowing an easy measuring the main signals/powers and clocks in the cpu socket - but you may need to fool the SIO with the socket occupied signal to start the board.

Check common control signals for working/changing state, eg. SLP_S3#, RSM_RST#, PLT_RST#, PWR_BTN in and out at the SIO, etc.

Check if RAM's are correctly connected to CPU (all signals) - here u can use a tester from Aliexpress too, otherwise its mass of time to check all signals manually.

Take a scope, and check whether the PCH starts to read BIOS after power on, best to do it on pin 2 (data line). This should indicate that PCH is working (RTC present).

Read out the BIOS with a programmer, eg CH-341, and check if the image is healthy and the a suitable one for the dedicated board, eg. by opening it with a tool called CoffeTime 0.99, or another BIOS editor - especially if you don't know the history of the board.

Inspect the board carefully for missing, ripped off components, especially for small ones, use a 10x magnifier glass.

Check also this thread:
https://www.badcaps.net/forum/showthread.php?t=44779

You can also look into this chinese site: https://www.xwfix.com/
The english language there is dramatic (i.e. "chinglish") but just read without stopping and try to understand, with time you will get it out, at least the most of it. They have some impressive repairs described, which show how complicated it can be to find such kind of issues with motherboards, and not only.

The order of the measures I described is maybe not optimal, so start with the for you esiest and fastest to do measures, and go into the power sequence analysis as the last and most difficult part.

Also all these experience I described is coming only from the Intel platform, so it may in some cases not apply to AMD, as I do not have lot of experience with this platform.