PC monitor killed graphics card?

[kawlby]

  Hi! First time poster here but a long time lurker and an even longer viewer of Dave's videos on Youtube. This is such a great community it really made me take an interest in electronics and grab my first soldering iron and so far so good.

  A few months ago my 4 year old GTX 580 died, I wasn't really surprised because the card was showing signs of being near it's life span (artifacting etc.) and I even baked it once when it stopped working completely and managed to get another 5 months use from it. Being that I started college and didn't have time ti play video games anymore I decided not to spend money on a new graphics card and just stuck my old 8800GT into the PC as it was more than enough for my needs.

  Today my 8800GT too has stopped working it seems, this struck me as odd as even though the card was old (bought it back in early 2008)  it was never used a lot. Not wanting to come to terms that 2 GPUs died on me in the past three months I decide to try and find out what might be the reason behind this, so I power cycle my PC (unplug, press the power button on of for a minute plug back in), try another cable from the PSU, try a different PCI-e slot, everything to no avail. At one point I unplug the DVI cable from the back of the PC and while plugging it in to a different DVI port (the 8800GT has two) I notice a very small spark.

  The spark appeared between the metal "casing" of the DVI port on the GPU and the same metal casing on the DVI cable from the monitor, which if I am correct both serves as ground contacts between the GPU and the monitor. I have no idea why there would be any voltage between the monitor ground and the PC/GPU ground, I have a safety extension cord where both my monitor and PC are plugged in so I wouldn't think it would be the problem. I've just recently redone the grounding in my house as it is pretty old but even before I did all of my other sensitive electrical equipment worked fine. I'm not sure if even this small sparking between the monitor and gpu ground contacts could be the reason for the graphics card dying.

[wraper]

I've just recently redone the grounding in my house as it is pretty old but even before I did all of my other sensitive electrical equipment worked fine.

Dunno what you have done with grounding but spark means that at least one of the devices is not grounded. Also, if it didn't die in the moment when connected, this is not the reason of fail. 8800GT GPU have a design defect BTW, when thermally cycled bumps under the silicon die crack and it fails.

[Ice-Tea]

Unplug everything. Measure resistance between earth connection on the power plug and the connector metal shell.

[SaabFAN]

Many pieces of equipment, especially monitors, aren't grounded on their own.
This is due to the fact that the entire case is made out of plastic and therefore the user cannot be exposed to mains voltage even in case there's a catastrophic failure in the PSU.
This means that the shield of the Monitor-Cable becomes the ground-lead (might be a good thing btw. - no ground-loops possible that way) and when you connect it to the PC, there's a spark.

I don't know if this can destroy a graphics-card, because the tiny amounts of current that flow when the charges equalize go through the chassis. If the connectors of the graphics-card do not provide a direct path to the chassis however, as it was the case with my HDMI-Capture Card that died, I think it can cause damage.
Because you connected via DVI, which are always grounded to the chassis via the screws, I think the most likely explanation is that the chip itself died. The early revisions of the 8000-Series GPUs are unfortnuately prone to internal chip-failure.

Edit: Btw. I measured about 30V difference between a non-grounded Xbox One and my PC-Case. The current-flow did only show up on the µA-Range and was about 2 or 3µA.

[wraper]

Many pieces of equipment, especially monitors, aren't grounded on their own.
This is due to the fact that the entire case is made out of plastic and therefore the user cannot be exposed to mains voltage even in case there's a catastrophic failure in the PSU.

PC monitors are grounded in 95+ % of cases, those with external power supplies might be not grounded sometimes. And there are exposed connectors, to which you can attach exposed cables. User easily can be exposed to the mains voltage if it is not grounded and such PSU failure occurs.
On the other hand, Apple laptops are made of metal and not grounded. Therefore can give you a tingle because of the PSU noise suppression capacitors connected between primary and secondary side.

[botcrusher]

The only ungrounded ones I've seen are led backlit flatpanels that have wallwarts or bricks. Either case, the voltage going into them is pretty small.

Also, the dvi ground is connected (or should be) directly to system ground. You should see much bigger issues is ground was a problem. Also, large ESD tends to trip at least some ATX supplies (I've accidentally zapped one with my socked foot before on several occasions, don't ask.)
If it isn't esd, take a meter to it and see what kind of voltage is present on the shield.

[SaabFAN]

PC monitors are grounded in 95+ % of cases, those with external power supplies might be not grounded sometimes. And there are exposed connectors, to which you can attach exposed cables. User easily can be exposed to the mains voltage if it is not grounded and such PSU failure occurs.
On the other hand, Apple laptops are made of metal and not grounded. Therefore can give you a tingle because of the PSU noise suppression capacitors connected between primary and secondary side.

I have such a monitor and also killed a small board yesterday because my Signal-Generator has the same stupid setup: External PSU, no grounding - Caused a little spark -> BGA2869 dead -.-
It's probably some stupid little hole in the regulations that allows manufacturers to save 10 cents on an extra wire and a different plug.
Btw. If you're working on an iPad or a Macbook that is connected to the charger and have an antistatic wrist-strap on, there's a nasty stinging sensation where the stap is connected to the skin (everything properly connected according to VDE-Regulations).
I could continue to rant about this, but I think I stop by saying: I don't like PSUs built like this!^^

[wraper]

I have such a monitor and also killed a small board yesterday because my Signal-Generator has the same stupid setup: External PSU, no grounding - Caused a little spark -> BGA2869 dead -.-
It's probably some stupid little hole in the regulations that allows manufacturers to save 10 cents on an extra wire and a different plug.
Btw. If you're working on an iPad or a Macbook that is connected to the charger and have an antistatic wrist-strap on, there's a nasty stinging sensation where the stap is connected to the skin (everything properly connected according to VDE-Regulations).
I could continue to rant about this, but I think I stop by saying: I don't like PSUs built like this!^^

There are downsides with grounded devices too. While double insulated devices can give you a tingle, they have reinforced insulation and are safe to use everywhere. There are a lot of old houses which do not have any ground in the outlets. Devices which require grounding, cannot be considered safe when used in such places. Also most of the portable power tools are not grounded for a reason.
In your case you have done hot plugging, I guess.