Did my nano project destroy my PCI slot?

Other Hardware General Electronics
1

the hardware relevant to this:
mb: asus M5A 78L-M (am3+ socket)
cpu: amd fx 6300
gpu: GTX960 d2d5
arduino nano

I don't have a schematic, but this all relates to a driver circuit for a large DC motor (90v, 20A). In generally I was rectifying mains, smoothing with large caps (~172vdc), then chopping it up with a mosfet driven by an optoisolator(tlp621). basically a really crude switch mode power supply. A circuit I have built many times, though with smaller components. I' m not gonna waste time with the schematic because the problem was almost certainly in deviation from those schematics (i wired something wrong)

when I went to connected the nano to usb, the moment I touched the plug to the port, my computer instantly lost power. honestly, its happened before(not on this project), the atx quickly shuts off to protect. always a scare, never been a problem. i'm thinking i messed something up with my optical isolation & my 170vDC power supply jumped into my signal lines-->my computer. the thing is, I was still in testing & had not made any connection to the +170. the ground side of that was connected to the ground of 12v power supply powering the gate of the mosfet. actually that ground was common with the arduino power supply too (i was having issues & trying stuff kinda recklessly)

Unlike other times the computer did not come back on, at least the display didn't. suspecting my GPU, I noted the on board fans were still running & if I disconnected the secondary power, the red led indicator on the board lit (indicating the missing power, but also that is was getting power from the slot). I swapped it for a cheepy PNY (tested in another box), still no video. That one has a fan powered by the pci, fan runs, no video. Swapped to the onboard video, viola. terrible image, half the res & only one monitor, but im typing this on it.

No visual damage anywhere. I've tested all my usb ports, hard drives & peripherals. only damage was the arduino(usb chip fried) & apparently the PCI slot? but nothing around the power lines that the surge came in?!? why would that be? and why would the rest of the motherboard be fine?

another detail that may matter: for a couple weeks now, I've had a few randomish display freeze ups / slow running. It was bad enough that I had planned on reinstalling the drivers in the next few days. So MAYBE that slot was failing & then a power surge killed it (and NOTHING else)? I dunno, I'm perplexed.

well thanks for reading my tale & any answers / advice you may have.

ps I haven't posted on here since the format changed, if a mod thinks this is better somewhere else, feel free to move it. you wont hurt my feelings. :wink:

2

Your topic was MOVED to its current forum category as it is more suitable than the original

As to your problem, please post a schematic of your project. The computer should never shut off when you plug a device into a USB port and if it has happened before then that is very worrying

3

thank you for moving it. sorry I got confused with the new(to me) format & i have a hell of a night.

but no I wont be posting a schematic at this time for the reasons I mentioned in my post.

I'm not worried about the power supply turning off, its doing exactly what it is suppose to when I wire something up like an idiot. Never accidentally made contact between +V & the ps case? its the same thing & no biggy.

I'm more interesting in how some current (doesn't matter the source) was able to travel through the usb (only connection to the computer), not damage any of the usb, not damage the power connections in the PCI, but render the pci slot unusable, with no indication from my system?

4

You have the schematic - we don't
You have the actual circuit - we don't

Providing help is going to be difficult

5

That's perfectly OK actually. :grin:

We won't be trying to figure it out either for the reasons explained in the instructions and in so many answers. :+1:

1 Like
6

So you have the MOSFET optically isolated from the Nano, yet the 12V supply for the MOSFET gate has its gnd common with the Nano gnd, which you are trying to isolate!!!
For the 12V to control the gate its gnd will also need to be connected to the MOSFETs Source.

That sounds like the Nano is not isolated from the 170V, which you did not have connected, or the 12V which you do have connected, but for isolation should not be!!

Please a schematic, a picture is worth a thousand words.
If you have made many of these, then you must have a proper schematic to refer to.

Post it as a jpg, either a hand drawn image, or a jpg EXPORT from a CAD.

Please include all power supplies and component labels and pin names.

Thanks.. Tom.. :grinning: :+1: :coffee: :australia:
PS. Sorry, I know its a lot, but you have had a major and possibly expensive failure and you do not want it to happen again.

7

if I had a schematic, I'd post it. I don't, I just built the thing with some scribblings here & there. I wasted a bunch of time on it, I don't wish to spend anymore drawing schematics. and I don't have the circuit anymore to back draw it. and the mistake that caused the problem, simply will not be on there.

for arguments sake, lets assume I touched a 12v to the usb ground. ALL i care about is how a current can run through all my usb & just selectively take out my pci slot.

8

well the 170 & the 12 have to have a common ground & yes I broke my opto isolation in the way you described, but opto isolation shouldn't have been absolutely necessary. Its all moot, I am not interested in what happened on the aruidno end, I am interested in what happened on the computer end. how it did such selective damage.

9

you understand I am not asking for help on the arduino system, right?

I understand the need for & normally would have provided a schematic, but its just not relevant to my questions. Just assume I touched a 12v line to atx ground. How's that cause the problems I described with my computer?

10

Is it possible the short ran from the USB port, through the PCI slot, into the video card, and out the video cable?

11

thank you, thats the sort of discussion I was looking to have.

Your hypothetical is possible, i think, I really don't know enough about the pathways on a motherboard.

though both screens & the GPU are totally functional still (just not with this mb) and the pci slot still supplies power to the card despite it not registering anywhere in the system.

12

Hi,

All the more reason to have a schematic.

Tom... :grinning: :+1: :australia: :coffee:
PS. If you are going to work with and experiment with electronics, you need to get in the habit of drawing circuits and when a project is finished, drawing a legible and informative one.

13

Typically, when requesters wish to limit or control the discussion, it precludes the methodology and solutions that they did not think of. That limits the ability of people to help.

14
  1. You are in Arduino forum but don't want help with Arduino but with computer???
  2. You killed your computer by a fault in your mains voltage circuit. You had similar accident repeatedly. Instead of finding what exactly happened to your PC you should find ways to prevent mains voltage to get where it shouldn't be. You can easily kill yourself next time!
2 Likes
15

  1. yes, thats why its under general electronics. I've also posted on computer specific forums.

  2. its not the same issue I have had before, once or twice i've had a floating ground touch off thee usb & shut down the atx. I didn't kill my computer, I'm typing on it now. PCI slot is the only issue.

  3. I replaced the driver I was building with a simpler system, the next time I try to do it, I will figure out the problem. This project was unplanned, the result of my air compressor dying, I simply haven't the time to rework the project. So I'm not just recklessly moving forward as you seem to think.

16

if this wasn't an impromptu, rush job, I would have done so, as I have many times in the past. But thats not the case.

17

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.