I have an Aorus B450 M motherboard, and since a couple of weeks ago, I hear permanent static/crackling noise coming from the rear headphone jack, and also a buzzing sound when I move the mouse or hold down any keyboard key. Investigating, here are some of my findings:
- when I reduce mouse polling rate from 1000Hz to 250Hz, the static and buzz sound are almost completely eliminated
- changing Windows power plan to Power Saving, which greatly lowers CPU clock, also reduces noise considerably
- I can also hear the noise when touching the TRS connector onto any metallic parts like I/O shield, mobo pins and unpainted screws
- there's no noise when plugging the headphones in frontal jacks or a USB DAC (FiiO K1)
- I don't get shocked when touching any metal parts from the case or devices plugged into it
- changing outlet makes no difference
Based on these symptoms, as a layman, I'd say it's like the motherboard is not grounded and/or filtering electrical noise from parts and peripherals. Is this issue caused by mains improper grounding? Is it caused by a faulty PSU? I haven't tested another PSU yet or took the PC to another place, because I'm afraid it's a fault from the mobo, which unfortunately isn't in warranty period anymore.
If it's the mobo the culprit, is there any economically feasible repair, or I'd better buy another mobo, USB sound card or live with the noise? Is there something I can investigate with just a multimetre, or do I need a scope and other tools? Thanks in advance.
majority or all of the grounding is done through the cable from the main power supply cable between motherboard and PSU from which practically is grounded through the mains AC. Could ensure it's connected?
You mean the ATX 20+4 pins cable? The PC wouldn't even turn on without it.
Is it with headphones or speakers? If speakers, is the amplifier / active speaker grounded by its mains cable?
3. I can also hear the noise when touching the TRS connector onto any metallic parts like I/O shield, mobo pins and unpainted screws
Again, headphone, speakers? Which part of the TRS is touching?
I've only tested with headphones, but speakers might be affected as well. TRS tip is touching any metallic part like I/O shield edges, rear DVI-D jack threads, motherboard pins like S/PDIF connector, GPU screws on backplate etc. pretty much every part made from metal.
So how are you able to make this noise by touching TRS to the case?
For headphones to work, both the the tip or ring AND the sleeve need to make contact with something.
One possible source of such noise is some metal junk causing unwanted contact between audio socket ground and the chassis, I/O shield, etc. These need to be separated, because motherboard (and audio) ground is at a slightly higher potential than chassis ground due to all the current flowing from the mobo to the PSU (Ohm's law).
I don't know how, but I can hear the exact same crackling noise and hum when touching the TRS connector onto any metal part. I think I've read something about that, IIRC one "fix" was putting a piece of cardboard between the edge of the mobo and the case, it's worth a try. I'll see if I can spare some time this week to unscrew the motherboard from the case, last case I'll see if someone can lend me a PSU. Thanks for the help.
Generally, what you hear is the voltage between the chassis and motherboard or between different parts of the motherboard. It's not zero because of current flowing through motherboard ground traces and those black ATX cables and it varies because the current varies with CPU load.
Ideally, the headphone jack should have its own ground trace going straight to the audio chip and connected to the motherboard's ground plane right there and isolated from all other "grounds" along the way. In such case, only a short circuit between audio jack and the chassis or some other nearby connector could cause such problem.
Alternatively, maybe they connect the audio jack to the motherboard's ground plane, I don't know. Then any connection between any exposed ground in this area and the chassis could cause issues and improving isolation may help.
I have difficulty believing that you can touch only the tip of the jack and hear anything in the headphones. Headphones need current to sound. The current needs to flow from somewhere through the driver to somewhere, and the two ends of the driver are connected to different terminals of the jack.
Are your outlets grounded/earthed?
Thanks for the explanation, but like I said, I do hear noise just by touching the tip, and not the ring or sleeve, I know it's hard to believe, but I wish I were lying.
Anyway, it doesn't matter now because a disaster happened. I unscrewed the motherboard from the case and I/O shield, and turned it on. I started Windows fine, but after a few seconds, the PC shut down. Tried to swith it on again, the DRAM diagnostic led lit. Then I reinstalled the mobo onto the case and turned the PC on again, and the fans leds and CPU led started flashing very quickly, then almost immediately I saw smoke coming from the top part of the board, I guess from the top VRMs, and there was a burn smell, I shut it down as fast as I could.
I inspected the motherboard and CPU die, I can't see any burn mark on any components or traces. Unfortunately I don't know anyone who has an AM4 motherboard willing to lend me for testing, so I could buy a new mobo or CPU, it sucks. Man I'm so depressed with this, BS I should have at least tested while the mobo was still in warranty, just don't know what to do
That sucks.
But what the hell has happened?
Is there a chance that the ATX ground connection is so dodgy that the motherboard was returning a significant part of its ground current through the I/O shield and something went wrong when this was removed?
(That could still be a problem with a new board. BTW, is the PSU a good one?)
I don't know what happened. The PSU is a Cooler Master MWE V2 Bronze PSU. I took some pictures, visually it seems fine both on both component and solder side:
That smear on VRMs is not a burn mark, but fingerprints I believe. Tomorrow I'll go to a computer shop and test another PSU, still have some hope it hasn't burnt either the motherboard or processor.
Well, smoke is never a good thing. I would continue searching until something is found.
It may have originated somewhere lower, before flowing upwards, as smoke likes to do.
I don't think there is much combustible material in a CPU except for the bottom PCB, so if the PCB isn't charred then the CPU probably wasn't the thing that smoked. Are the power connectors and their cables alright?
The case and motherboard were laying horizontally, so I don't think the smoke came from below. All power connectors, both from the PSU and mobo are apparently fine at visual inspection, I checked everything with a flashlight.
Turned on the PSU with the metal clip on black and green pins trick, it powers on as expected. Is there any diagnostics I can do with just a DMM? Like measuring +12V on EPS connector that feeds the CPU?
"I inspected the motherboard and CPU die"
What did you mean ? can one really conclude such quickly just due to dismaying smoke or smell?
Must sit back, calm down and reproduce chronological steps of the disaster to be definitively reevaluated and written down
Nevermind, I tested another PSU with the mobo out of the case, and the faulty component was ignited a few seconds after turning on the PC (using a screwdriver on power pins) with flame coming from it, thanks God it wasn't installed in the case. Checking the second photo I posted, I could now see burn marks from the solder and/or terminals. This is the result of the component that underwent combustion:
Before this, I turned on my PSU with 6x 120mm fans, a PWM fan hub and a 3.5" hard disk as loads, using a metal clip. Then I checked all ATX (20+4), EPS (4+4 CPU) and GPU (4+2) connectors pins, all of them were outputting voltages within tolerances. Don't know if it means anything, and probably not accurate because it wasn't powering other parts like CPU, GPU and motherboard itself, but I felt like checking anyway.
Anyway, still don't know if it's a motherboard itself or PSU that caused this, too depressed to investigate further. Anyway, thanks for everyone who took their time to help me.
Yeah, MOSFETs are usually adjacent to the VRMs. Do you believe it was the component itself that was faulty?
a driver fault partially opening both mosfets at the same time would screw with main 12v power until one of the mosfets gives up