An Inductor must have blown up on me

[Lord Nicoll]

I was wondering why something wasn't working right, and when I performed a detailed analysis of the card in question, an inductor was quite badly damaged. Do these randomly blow up on GPUs? I'm pretty sure this is the display driver circuit, I have used a lot of SMD ferrite core inductors and never seen one fail like that, this card never left it's server so I have no idea how it could gave been damaged any other way. 

[simson0606]

looks like it has gotten a bit too toasty

[Lord Nicoll]

5 minutes ago, simson0606 said:

looks like it has gotten a bit too toasty

Maybe, I have found the rest of it, oddly on the other side of the server, so either I broke it taking it out (I don't think so) or it had blown and when I flipped it around bits fell onto the server (it's now littered in tiny conductive ferrite particles. 

[BrinkGG]

8 minutes ago, Lord Nicoll said:

I was wondering why something wasn't working right, and when I performed a detailed analysis of the card in question, an inductor was quite badly damaged. Do these randomly blow up on GPUs? I'm pretty sure this is the display driver circuit, I have used a lot of SMD ferrite core inductors and never seen one fail like that, this card never left it's server so I have no idea how it could gave been damaged any other way. 

 

You can see a lot of heat damage between the solder points as well, so It looks like it overheated and failed to me. 

[Lord Nicoll]

Just now, Brink2Three said:

You can see a lot of heat damage between the solder points as well, so It looks like it overheated and failed to me. 

it was always like that, it's an AMD Firepro V7900, so single slot, it wasn't a cool running card but it wasn't hot either (in the 70°C range)

 

[DragonTamer1]

Probably a micro defect in the powdered ferric core that got worse over time due to heat.

 

Inductors don't usually fail unless you shove outrageous amount of current through it (that one would need more than 8A DC based on its wire gauge).

[Lord Nicoll]

3 hours ago, DragonTamer1 said:

Probably a micro defect in the powdered ferric core that got worse over time due to heat.

 

Inductors don't usually fail unless you shove outrageous amount of current through it (that one would need more than 8A DC based on its wire gauge).

well, I must have broken it myself, because the card still works, meaning it must be SDI or CrossFire Pro related, since I use neither of those. I de-soldered the inductor and just cleaned up the pads. 

[DragonTamer1]

2 minutes ago, Lord Nicoll said:

well, I must have broken it myself, because the card still works, meaning it must be SDI or CrossFire Pro related, since I use neither of those. I de-soldered the inductor and just cleaned up the pads. 

You weren't working around it frequently were you?

 

Are you trying to replace it?

[Lord Nicoll]

2 minutes ago, DragonTamer1 said:

You weren't working around it frequently were you?

 

Are you trying to replace it?

I pulled the FirePro V7900 out of a HP ProLiant DL160 G6 rack server that's ESXi file system had seemed to have become corrupt, I was testing all the cards in the server (there's only two and only one is a normal PCIe device) to test them on their own, and I must have knocked the top part of the card off the PSU (where I found the inductor fragments). I don't remember even hitting that area, I rested the card down but it was still in the PCIe frame, so it's possible the inductor just randomly broke at some point and fell out the PSU at that point, but unlikely.