Question on PSU Failure Diagnosis

Associate
Joined
2 Jun 2017
Posts
5
Hi,

Recently upgraded my Mini ITX to asus MB, I5 CPU, hydro AIO, 1060 Mini, and new RAM. The old psu is a 650W, about 4 years old. The rig was really quiet. Last couple of days I noticed the fan noise picked right up when I fired up a game, had freeze, and a couple of reboots without BSOD. Checked the temps and rig fan speeds seemed ok and 30-40C. I checked for dust. Anymore I can do to confirm PSU rather than another hardware issue? Thanks.
 
What PSU is it? Random freezing and reboots can be caused be a number of things including a faulty PSU. Is anything overclocked? An unstable OC can lead to random crashes and reboots. What do the system voltages look like in the BIOS. If they're out of spec, it might point towards the PSU.
 
PSU is: https://www.**********/products/650...onze-sli-crossfire-eps-12v-quiet-fan-atx-v231. Bought it in Mar 2013. Yeah, I wondered about the voltages. I haven't set up an OC, its just on the ASUS turbo which seemed to have tweaked the freq from 3.8 to 4.1. The freezes were when I was on games, though i'd suggest SWTOR and Divinity Original Sin aren't exactly overmatching the 1060. I did note different voltages to the fans when I went into the BIOS but being a newbie not really sure what is out of place. I did wonder whether the louder fan was the PSU rather than the case and MB - they seemed the same speed. Is there a good programme to download to get the figures - including the GPU - to post? Thanks for the prompt reply. JPB.
 
Try HWiNFO64 and run a CPU stress and GPU stress program at the same time.

Can you type out the actual brand/model? Link doesn't work/show it.
 
Hwinfo64 is good for monitoring temps and voltages in windows. There should be a system status page in the BIOS which will give you all the voltages as well. If you've got the Asus Turbo auto-OC turned on, I would suspect that it's not totally stable.

Copied the link into notepad and it gave me the whole starred out link. The PSU in question is a Cooler Master GX RS650 (ACAAD3) unit. Going by the age/product code, it's one of the Enhance built units according to Realhardtechx.
 
The PSU in question is a Cooler Master GX RS650 (ACAAD3) unit. Going by the age/product code, it's one of the Enhance built units according to Realhardtechx.
Finest Chinese garbage capacitors from Sucks'con in secondary without any real airflow because of wire spaghetti and those excessive size heatsinks.
Though not that Matsushita-Panasonics in primary are in much better place.
https://www.kitguru.net/components/...er-master-gx650-bronze-power-supply-review/4/
That PSU would need noisy high speed fan to have some hope of cooling for them and long term reliability...


You would want oscilloscope to measure if ripple is anything near ATX specs anymore, because ripple can wear down PC's components.
 
Ok thanks. I am away with family so I'll post results after the weekend. I've been running the rig since March no probs, so I don't think its directly the OC (its run silent and stable since I upgraded - that's why I noticed the change immediately a couple of days ago). However, to eliminate it I'll turn the turbo off, and do as you have kindly advised me and post the results. I'll also clean out any dust from the PSU fan. Just for info the rest are: Asus ROG STRIX Z270I GAMING; Intel Core i5 7600K; Corsair Hydro Series H60; Zotac GeForce GTX 1060 Mini 6144MB; Team Group Vulcan T-Force 16GB (2x8GB) DDR4 PC4-24000C16 3000MHz Dual Channel Kit. JPB.
 
I've been running the rig since March no probs
That's the way it usually is before problems...

That motherboard has couble BIOS updates so might be worth of try.
Though with possibly dodgy PSU BIUS flashing is more risky than usual maneuver.

Trying another PSU would be the best way to start bracketing what causes problem.
 
Hi

Ran all those programmes. Interestingly the only lock up when I was on the internet not long after starting up. Switched off ASUS Turbo. Ran the CPU stress test, no change to fan noise, max temp 55 degrees, no thermal throttling. GPU (FURMARK) ran fine, 71 FPS, 82 degrees, fans 74% - small increase to noise. Motherboard temps 35-40 degrees. Voltages - I think I was looking at the right figures:

CPU - nominal 1.088, ranged between 1.072 - 1.101, max draw 52W
GPU - 0.625 - 1.050, max draw 120w (averaged 80-90 on stress test)
5V - 5.16 - 5.24V
12V - 11.9 - 12.29

Just doing both tests simultaneously.
 
Did both stress tests. CPU was the same. GPU was a bit stuttering, started at a few FPS then settled at 50-60 FPS but the graph showed temp oscillations (rather than smooth when done earlier & singly), gradually increasing to about 70 degrees. When the CPU finished its test, it returned to the same FPS & 82 degrees. However, not lock up (and fan noise was not excessive). Thanks JPB.
 
Back
Top Bottom