Repairing electronic devices with the GA-H110M-S2PH model due to a faulty ceramic cap.
Repairing electronic devices with the GA-H110M-S2PH model due to a faulty ceramic cap.
Hi There, I experienced a motherboard failure a few months ago—it was completely dead and wouldn’t power up at all. I went into a bit of a deep thinking mode and checked the board closely. I noticed a faulty ceramic capacitor on the 12V rail (related to CPU voltage regulators and VRMs). After identifying the issue, I removed it (see photos) and tested it once; it was definitely shorted. Once taken out, the rail resistance dropped from a bad state to around 400-500 ohms now. The board is functioning properly.
I’m wondering if anyone else has encountered this problem or knows what I did to diagnose it? Here’s what I did:
1. Checked the 12V CPU connector—it was dead short.
2. Removed the CPU and verified its health.
3. Used a good multimeter to test all capacitors in the area.
4. Found the bad one by spotting the lowest resistance and looking for visible damage.
5. Replaced it and rechecked all previously problematic components.
6. Tried the board without the faulty capacitor; it worked fine after reboot.
7. Cleared the CMOS, and everything functioned correctly this time.
Regarding the capacitor itself: I don’t have the schematic, but based on what I observed, it was likely there for EMC reasons. It seems worth replacing if you want reliability.
It seems there are four phases, typically similar, allowing you to recognize the same capacitor across different sections. You could desolder one and check it with an LCR meter. The component likely serves to smooth the input voltage. If your PSU has noticeable ripple, this improves stability for the MOSFET. It probably isn't necessary to run the PC; the main risk is a few fluctuations entering the Vcore rail. Unless you're doing significant overclocking, small variations shouldn't cause issues.
Alvin853 has identified the correct approach. The other three phases, located to the right of the 8-pin EPS connector, should be very similar. You might remove one and check its capacitance—it could help confirm the setup. It’s likely a capacitor rated for 25-50V; since your input is only 12V, ceramic types with higher voltage ratings are preferable. A good choice would be 1…10µF, preferably X5R or X7R. The design of the motherboard seems to have altered things: on that specific phase, a larger polymer capacitor (270µF, 16V bulk) was moved farther away due to a hole in the PCB. That part should now be nearer to the MOSFETs, within less than a centimeter or so, which is why you can see them on other phases.
Hey everyone, I’m a bit stuck without a LCR meter. My multimeter won’t work well for low values like 1-10uF, so I ran some tests and found it’s likely being affected by decoupling across GND and 12V. Adding a 10uF capacitor seems reasonable. It’s interesting you mentioned poly caps—probably the reason was either humidity or a noisy power supply that pushed things over the limit. Checking the schematic confirms a 1uF/16V X7R is the right choice, definitely above 12V and ground.