Hoping for some informed advice here. I plugged in a new device, via a new USB3 controller card (PCIe) that has an extra power feed independent of the PCIe slot. Everything worked. A day later, on switching my PC on I got a bad smell and hastily switched everything off again. What I found is pictured below. Basically the insulation has pretty much burned of one of the red wires, exposing the conductor. Regardless of what caused the fault, should the PSU overcurrent protection not have prevented this? Am I risking something else by continuing to use that PSU?

 

 

This is why I favour multi-rail PSU's. Your Corsair is a single rail PSU.

 

Far better chance OCP would have tripped with a multi-rail.

 

Your single rail is rated at about 50 amps I guess. A multi-rail would have been running of one of the sub rails with an OCP trip of about 25 amps.

 

What was the device you were powering? Any shorts anywhere?

What was the device you were powering? Any shorts anywhere?

 

It was a Tobii EyeX eye tracker. I don't know if anything shorted, I am sending it back for analysis. The USB3 board appears unscathed so I would guess it's a fault in the EyeX.

 

Power supplies - we live and learn

Certainly could be the EyeX

 

The red wire by the way is the 5 Volt wire. That 5 volt wire was certainly getting more juice than it could handle.

 

It's not the first time OCP has failed to trip in these circumstances either. The PSU itself may not be faulty, as said, the OCP trip point on a single rail PSU is high, high enough so that a cable can melt before it trips. 600 watts or above, multi-rail is the way to go in my opinion.

 

It's worth testing the PSU with a PSU tester or multi-meter. Personally, I always have an Antec PSU tester available for such occasions.

 

Was there any scorching on the pins, or melting of the actual connectors themselves? That's the more common failure, loose connector arching. I had this happen to me with an ATX extender. Was arching out at the connector.

Was there any scorching on the pins, or melting of the actual connectors themselves? That's the more common failure, loose connector arching. I had this happen to me with an ATX extender. Was arching out at the connector.

 

No, as far as I can see everything else is pristine, including the EyeX and the USB cable (at least on the outside). The EyeX, BTW, requires USB3 and comes with a hefty cable, so I think the normal power draw is quite high.

 

The power supply seems to be still working fine - I have a multimeter but not sure what kind of test to do. Next time I will for sure look at multi-rail.

I have a multimeter but not sure what kind of test to do. Next time I will for sure look at multi-rail.

 

 

I use an Antec PSU tester. Very good device. Don't think they make them now, there are others on Amazon though.

 

 

http://www.amazon.co.uk/DIGIFLEX-Power-Supply-Tester-Diagnose-Black/dp/B004IO5CT8

I use an Antec PSU tester. Very good device. Don't think they make them now, there are others on Amazon though.

 

 

http://www.amazon.co.uk/DIGIFLEX-Power-Supply-Tester-Diagnose-Black/dp/B004IO5CT8

 

Ta, will check it out. Seems cheap enough.

Let us know if you discover the cause Mark. It's worth knowing in case anyone else has a similar issue. 

Well, I would steer clear of these cards. Tobii claims the EyeX isn't faulty and I have since discovered another couple of recent Amazon reviews highlighting the same problem, including another PC fire.

 

Computer fire

Technical critique

 

Thanks for letting us know Mark. Worth remembering.