Strange Freezing Issue

Toggle sidebar Toggle sidebar
I

imtheman0313

Honorable
Sep 19, 2012
14
0
10,510
Dunno if this is the right section or not, but I've got the strangest problem. First the rig:

ECS GF8200-A
Phenom II x4 940 (downclocked to 2.6Ghz for the mobo)
Plain jane RAM 6GB
5 Hdd
MSI 460GT
450W PSU.

Extra:::::::
500W PSU
2 127MM 3A 260CFM fans
2 120MM 2.4A 200 CFM fans
1 70MM 1.1A fan
2 40MM .7A fans
1 90MM 2A fan

Now my issue is this: Under normal idle or load conditions, no problem. Rig gets hot, no problem. Highest temp I ever see under load with fans off is CPU ~120F, GPU ~160F. So not even close to overheating, no throttling. I monitor the temps anytime I'm not in a fullscreen game, and nothing untowards ever occurs. HOWEVER> When I turn on the SECOND PSU, for some reason I get freezes, the occasional Blue Screen, and HDD errors. Seems strange to me.

Now the second psu is entirely separate from the system, other than case ground. It is bolted, like the primary psu, to the metal case. All the High powered fans I have are powered by the second psu. There's only a couple of low power fans running constantly on the mobo itself.

My question is this. Can it be possible, that when I turn on the second power supply, there is feedback somewhere (I would assume through the case) that interferes with voltage stability for the mobo? IE can switching a high energy device on, on the same ground plane as the mobo and primary psu cause electrical stability problems? If unlikely, what could possibly be causing this strange problem?
 
Solution
The_Tester
have a camera or some other non-computer dependent picture taking device (cell phone) ready to go in the event of an error.

If the second PSU is isolated with the exception of the ground then yeah I would say you have a ground problem on the surface. Provided there are no other events to tie to a crash with the exception of turning on the second power supply. I had a similar situation with my computer case. The fan speed selector would cause a spike/plunge in the SATA power rail it was connected to when I changed to a different voltage. The simple fix for me was to set it to 7V (med) and leave it alone.

First can you give a model number and or show a picture of the second PSU? How are you turning the second power supply on and off...
Sort by date Sort by votes
I

imtheman0313

Honorable
Sep 19, 2012
14
0
10,510
If the post was read, it would be apparent the SECOND PSU IS TOTALLY ISOLATED FROM ALL COMPONENTS. And that it powers ONLY fans. The only thing the two psus have in common is ground (a little electrical humor). I'm wondering if there possibly could exist transients on the ground plane of one psu, that SOMEHOW could affect the ground plane of the primary in such a way as to reduce the potential between the motherboard and it's voltage feed pins.
 
Upvote 0 Downvote
The_Tester

The_Tester

Reputable
Nov 22, 2014
202
1
4,760
have a camera or some other non-computer dependent picture taking device (cell phone) ready to go in the event of an error.

If the second PSU is isolated with the exception of the ground then yeah I would say you have a ground problem on the surface. Provided there are no other events to tie to a crash with the exception of turning on the second power supply. I had a similar situation with my computer case. The fan speed selector would cause a spike/plunge in the SATA power rail it was connected to when I changed to a different voltage. The simple fix for me was to set it to 7V (med) and leave it alone.

First can you give a model number and or show a picture of the second PSU? How are you turning the second power supply on and off exactly (be as detailed as you can)? Has the system behaved this way on it’s own with no interaction from the second PSU that you are aware of? Also are any of the fans metal as far as the frame of the fan/s?

Second once booted up, go into device manager and turn off write caching to ALL disk drives if not done so already (then it's relatively safe to get crashes, not saying that is good but safer…). Restart the computer for it to take effect (then double check that write caching is complete off for all drives).

Third if you have any non-essential startup programs that take a while to get going (Intel RST for example) let them come up and settle down first (or disable them temporarily).
Finally, 100% isolate the fan PSU from the system (literally nothing is physically touching or connected in any way to the system except for power connections to the fans).

TEST:
Repeatedly turn the supply on and off with the system running. First at slow intervals (say once every 5 seconds) then gradually speed up the off/on time till you are rapidly turning it off and on (as fast as you can make it power on/off the fans). If this does not work, just turn it on and use it like normal as if you’re not trying to make something happen a few times
.
If you can’t repeat the problems, electrically ground the second PSU to the case as it would be normally and try the above mentioned again.

If you can repeat the problem then that would be your answer. You got yourself a crap PSU ground (or some other isolating problem in the PSU itself).

If the system behaves this way regardless of the second PSU and the system is known to be in a “working state”, you have some ambiguous problem/s that may not be easy to isolate.

This sounds like it make take some time to confirm and or disprove. Also try to notice and or think up other potential causes while doing this.
 
Upvote 0 Downvote
Solution
I

imtheman0313

Honorable
Sep 19, 2012
14
0
10,510
I have done a few tests already with this thing. I JUST put my rig in a new, larger case, that can now handle 2 PSU. The problem is very new, never had one like it before. It has taken me about a week to narrow down what has been causing the issue. First, I thought it was the HDD: so I ran tests, hddtune, hddregen, minitool surface test, etc. When it happens, theres no extra load on the drives, no ram usage shooting up, and 0-2% cpu usage, so it's definitely not a hardware issue, as far as hardware that controls the system.
None of the fans have metal cases, they are all plastic, that was one of my thoughts as well, but I just rechecked that last night. I am certain it has something to do with the switching on and off of the second PSU, because when I turn it on, everything just stops. It will load once every 15 seconds. The very second I turn off the second PSU, everything returns to normal.
The Fan PSU is switched manually, I have a switch on the front of the case wired to the green wire and ground. All connections to fans are through Molex 4p connectors, then soldered to fan wires, so that I can switch PSUs later if I need to.
Now, I'm looking inside, and there is a possibility that there is a stray wire somewhere, maybe sticking out of the heatshrink. When I put it all together, I spent 3 days setting it up, but I got rushed to finish it because I had some things go out, so I had to switch everything over to the new case quickly. But, I'm pretty sure theres no voltage wires touching ground, for if there were, the PSU would just shut off, for protection.
So I'm back to the transients in the ground plane thing. Next test tonight is to remove the PSU from the metal of the case, and isolate the grounds. If I still get a problem, I'll have to do some exploration. These wires are still a giant mess, honestly.
 
Upvote 0 Downvote
I

imtheman0313

Honorable
Sep 19, 2012
14
0
10,510
Ok, so update. I have checked the ground for transient voltages, at least, to the best of my ability. From what I can tell, there is a 1.2v AC charge present on the ground at all times, my DMM shows 60Hz present. This was checked against the PSU ground AND the mains ground pin. With the primary PSU turned on, this is constant and the two grounds seem to be tied together (albeit through some circuitry).

Now the odd bit. It seems that nothing at all changes when I power up my second PSU. That tiny ACV remains constant, although I did not check for amperage. The only time the voltage drops, is if I touch the metal of the case myself, at which point it will drop to around .2vAC, which is within margin of error for my cheapy DMM (so it's probably really 0).

So, now I'm back to wtf, man. Why would a machine freeze only when turning on a fan? It doesn't cool off immediately, it takes a moment (either way that shouldn't make any difference). I searched and searched, and no stray open wires, nothing grounding by accident. So, again I'm at a loss.

Thank you so much The_Tester for all your help already, although it would seem I'm not done here yet..:)
 
Upvote 0 Downvote
The_Tester

The_Tester

Reputable
Nov 22, 2014
202
1
4,760
Only other thing I could think would be to plug the second PSU in (still completely isolated from the system except for fans) to an ungrounded source (like a cheapo 2 prong cord with that little knob cut off that prevents a 3 prong from being plugged in) and see what happens. Also, try turning the fan psu on just by itself but on a different outlet and or circuit in the dwelling and see what happens. I'm really trying to think of something to try that doesn't involve an oscilloscope or curve tracer lol. May need to break down and try a different PSU :(
 
Upvote 0 Downvote
I

imtheman0313

Honorable
Sep 19, 2012
14
0
10,510
Funny you should say without the ground pin, the surge protector I use doesn't have it (defeats the purpose I know) because the house I'm in was built in the 50s, and the addition hasn't been updated since. I have recently tried replicating the results......aaaannnnd, nothing. It would seem whatever the problem was, it has resolved itself. If it comes back I will return, but for now The_Tester will have provided the best answer. Thank you very much.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom