Electrostatic discharge causing new build to reboot

[OhioDude]

A new system I built for a dear friend reboots every time she or anyone touches the case without touching ground first. Sometimes, if she just shifts in her chair and touches the case it will reboot. I've built quite a few machines over the past 10 years and this is the first one that I've had a problem like this with.

Here's all the pieces/parts:

Gigabyte GA-MA78G-DS3HP Motherboard
Athlon 64 X2 6000 3.1GHz
4MB (2 x 2MB) G.Skill DDR800 RAM
MSI R4850-T2D512 Radeon HD 4850
Western Digital 640GB SATA Hard Drive
LG 22X DVD±R SATA DVD Burner Model GH22LS30
EDIMAX EW-7128G PCI Wireless Card
PC Power & Cooling Silencer PPCS500 500W PSU
AeroCool ExtremEngine 3T Mid Tower Computer Case
Acer X223Wbd 22" LCD
Vista Ultimate 32-bit

She gave me a strict budget to work with and she was active in helping to choose the components. The all-steel AeroCool case is not what I had originally configured, but she picked it out. (She's somewhat of a geek herself... she didn't want a plain old black box.)

I'm just not sure what to do. If I should swap out parts, where do I start? I'd appreciate any comments and suggestions.

 

[RaySun2Be]

Double check the MB mounting, make sure it's not touching any metal where it isn't supposed to be.

Also double check any wiring, to make sure it's not pinched or shorting, especially the wires for the start/reset buttons. Same thing for the external power cord.

 

[petrusbroder]

I agree with RaySun2Be.
There is one thing I would do though: connect a wire to some "ground" (i.e. a water pipe) and fasten it to some screw which is close to to the PSU.
If that cures the problem, then you know it has something to do with a "bad" wire or some unwanted connection somewhere ... especially when you fastened the mobo to the case - have you put washers between the screws and the mobo? I do not know if it is really needed but they are delivered with many cases for some reason, aren't they?
If that does not give a good result I'd check (swap) the PSU.

 

[VeryCharBroiled]

is the outlet shes has it plugged into properly grounded (3 prong). even if it is might be worth testing it with a $5 outlet tester from radio shack. the one with the colored lights that indicate various faults like no ground or reversed hot/neutral.

 

[VirtualLarry]

The case should be grounded, via the PSU, via the earth ground at the outlet. Touching the case should discharge static, but should not reboot the machine.

What happens if you touch the back of the PSU with static, rather than the side or front of the case? Does it also reboot?

That would seem to indicate to me that something isn't grounded properly, and instead, the static charge is discharging backwards, going through the case to the motherboard ground plane, and then back up through the components to the PSU.

 

[bruceb]

I tend to agree about it being improperly grounded. Most likely at the wall outlet, especially if it is an older home or the home was wired with Romex where the ground wire must be manually connected to the Green ground screw on the outlet. If the home has BX / Armored cable, usually the cable itself makes up the ground, but only if a metal box was used and the cable is still in good shape. It can go bad over time. You may also need an electrician to examine the ground connections and bonding at the main panel, ground rod and water main (usually where the water meter or main shutoff is) ... Easy test is take a DMM on AC Volts and test from the Neutral Side of the Outlet (narrow slot, left side above the little ground hole) to the Ground hole .. you should get a reading of ZERO .. any reading above that means a wiring problem.

 

[Insidious]

It might be something as simple as a bad LED, power or reset switch on the case. Once the PC is running, try unplugging the connections from these at the motherboard.

Faulty USB/1934/speaker jacks can also transmit the static shock to your electronics. You might want to try unplugging these too and see if it helps.

(All this assumes the house power and ground phases are correctly wired. That ground check device is a good idea.)

 

[GLeeM]

I have a similar static electricity problem.

Only it doesn't reboot, it just makes the USB keyboard not work (so a reboot is necessary)

It happens when I plug in my USB thumb drive without discharging myself against the side of the case first :spark;

I'll have to try some of these suggestions or get a new computer

 

[OhioDude]

Thanks for all of the suggestions.

I've got plenty of things to try.

 

[imported_wired247]

It could absolutely be the wall outlet itself that is not grounded properly. In fact I would put my money on that.

 

[Markfw]

Originally posted by: wired247
It could absolutely be the wall outlet itself that is not grounded properly. In fact I would put my money on that.

I can say that lately, I get a big shock almost every time I touch my main computer, but no bad effects. However, I did the wiring myself for that addition, and I know its all to code, and was inspected.

So a faulty outlet seems to be the problem. I think the idea of a wire from the case directly to a real earth ground, or the fusebox directly, is a great idea.

 

[OhioDude]

Interesting development. The wall outlet tests fine. Also, they ran a ground wire from the case to a water pipe and the same thing continues to happen. As long as Michelle grounds herself first, or at least lowers her electrical potential by touching some metal object first, like the big metal chair she sits in, everything is fine.

So now we know it has to be something in the PC itself.

I doubt that it's a direct short between the motherboard and the case. I've had a wandering screw work it's way underneath a motherboard during a build and the system wouldn't even boot - fans started briefly then everything stopped. Located and removed the screw, and everything was fine. I would think a direct short would cause more severe symptoms similar to that. But I suppose it would depend on exactly what was being shorted to the case...

I think what I'll try next is have her bring the system in and I'll pull the motherboard out of the case and run it on a cardboard box to see if touching the case and/or PSU causes it to reboot then. If it does, I would tend to think that it would have to do with the power/reset buttons, case audio ports, USB/1394 as Sid suggested, or the PSU. If that alleviates the rebooting symptom, then I'll... I'll... Not sure what I'll do. Possibly try putting some insulating washers between the MB mounts and the MB as PB suggested.

Sound logical?

Thx again, ppl!

 

[Insidious]

I had a sonata II case (Antec) that loved to re-boot when I plugged in a USB memory stick. As long as I touched the shroud on the drive and the case (chrome face plate around the jacks) at the same time with my finger before actually inserting it... life was good. Never did figure that one out... Sold the case :laugh:

-Sid

 

[imported_wired247]

it could be that touching the case causes an arc across the reset switch. try unplugging the reset switch from the mobo.

that would be directly related to what's going on in the case since the reset button is purely a feature of the case.

 

[OhioDude]

Originally posted by: wired247
it could be that touching the case causes an arc across the reset switch. try unplugging the reset switch from the mobo.

that would be directly related to what's going on in the case since the reset button is purely a feature of the case.

Ray was thinking along the same lines, too. Certainly is worth a try. Before I remove the MB from the case, I'll disconnect the reset switch leads from the MB header and give that theory a spin.

Thx.

 

[Insidious]

You can do that with virtually all of the case access connectors. Get the PC running and then you can remove all of the case jumpers (LEDs and switches) from the motherboard. You can also disconnect any usb ports that are cabled to the motherboard for 'quick n dirty testing.

-Sid

 

[VirtualLarry]

I've wondered why case designers don't have an exposed piece of metal showing on the front of the case, to use to ground yourself, before, say, you plug in a USB flash drive. Specifically to prevent things like this.

 

[Drsignguy]

The above Ideas and suggestions are all in line to a problem that can be solved. Nice work.

For starters, Static is most likely caused by low humidity in the home. Somewhat an easy fix there and would suggest to her by looking into that also.

I would look at the backplate where your connections are and see if all the tabs are in the right place. I have had, and seen before of a little grounding tab getting into a usb connection and causing problems. Just another added place to check

 

[Alyx]

I've had a mobo short on the case if it didn't have a washer. I ended up having to pull as many of the stands (the things the mobo screws go into) then put them back one at a time till I found which was doing it. If it doesn't short when the mobo is sitting on a static bag then that could be the problem.

Also, you can turn on a mobo without the front panel stuff plugged in. The mobo manual should have info on which space is which, then short the two power spaces real quickly with a screw driver.

 

[soccerballtux]

This could be a motherboard design problem. If the traces to/from the RAM and CPU/bus aren't within a certain length tolerance, then that's extra time the CPU ends up waiting around for all the bits requested to come in. As it's waiting, an ESD can bump a falling signal back up past 90%, or a rising signal back below 10%. In mild cases this causes the computer lock up; the extreme cases (extremely poor design) it reboots entirely.

Gigabyte is a pretty respectable manufacturer though; have you found any other users with similar issues while googling around?
I never have any ES to D, perhaps see if you can move the computer away. Replace her chair with something non-synthetic in material (leather or cloth would be great if it's nylon/plastic).
You shouldn't have to do that for this to not be a problem though. I'd talk to Gigabyte and see if they're interested in testing it-> figuring out why it's doing this.

Best of luck.

 

[Assimilator1]

I'd virtually gaurentee they'd ask you to test the case/PSU 1st before they got interested.

I also think it would be very unlikely to be a mbrd design issue, as they would of probably come across any issues like that in testing. Though if all else fails it'd certainly be worth considering.

OD
I take it the PSU is screwed to a metal part of the case chassis? & the case cover is screwed to a metal chassis?
Reset button sounds like a good suggestion .