GTX TITAN FE Died

J

jasonexe

n00b
Joined
Mar 17, 2021
Messages
20
Hello,

So my GTX TITAN FE recently died and I’m hoping that someone with more experience can help me. The first thing i did with the gpu when it died was checking for components that were physically damaged but I didn’t find anything. After that i decided to measure resistance on each rail and i noticed that i have a straight short when measuring at the inductor, close to the buck converter (APW7142). The 7, 8th pin from the IC is supposed to output around 3.3v and it passes through that inductor that I measured earlier. The normal resistance there should be around 2k ohm so there’s definitely something wrong. I decided to remove the chip and check resistance again but no luck.. I don’t know much about this pcb and have no clue which components use the 3.3v.

I’m hoping that someone could help me with this, Any help would be appreciated.

Kind regards,
C24E4FCB-396E-4FA3-9F5A-172DA6ADF271.jpeg
 
So, you removed the APW7142, but you still have a dead short to ground on its output?
 
Your original post makes it sound like you have more than one of these cards on hand. If that's the case, can you check on a known-good one?

Your next step is probably to start looking up data sheets and checking what calls for 3.3V. I'm surprised there would be a dedicated VRM for 3.3V, given that 3.3V is available from the power supply, but nvidia has been known to do things like that.

You could also voltage injection with isopropanol, freeze spray or a thermal camera - start with a low voltage and work your way up.

This may be a borderline insulting question, but... you do have access to a hot air machine, right? You won't get good results trying to use a heat gun or a soldering iron. You need a real hot air station with temperature control.
 
RazorWind said:
Your original post makes it sound like you have more than one of these cards on hand. If that's the case, can you check on a known-good one?

Your next step is probably to start looking up data sheets and checking what calls for 3.3V. I'm surprised there would be a dedicated VRM for 3.3V, given that 3.3V is available from the power supply, but nvidia has been known to do things like that.

You could also voltage injection with isopropanol, freeze spray or a thermal camera - start with a low voltage and work your way up.

This may be a borderline insulting question, but... you do have access to a hot air machine, right? You won't get good results trying to use a heat gun or a soldering iron. You need a real hot air station with temperature control.
Click to expand...
I have a GTX 980ti laying here thats how i know some things about it. The pcb is really similar. I also seen tutorials about repairing a gtx titan. I checked the data sheet from the ic but I don’t exactly know what to look for, Im still new to this. Also, i dont have a bench power supply or thermal camera to check for shorted components. I do have a hot air station for smd soldering etc. So that’s not a problem.
 
I have a reference 980 Ti here, and it's definitely different from the photo you posted. I wouldn't rely too heavily on it being that similar.

You're going to need to look for datasheets for the other ICs on the board, and look for which ones call for 3.3V. Then, find the pin where they're supposed to have 3.3V and check for a short to ground. Keep in mind that there are likely to be a bunch of capacitors on that rail, and one of them could be shorted too.

Your best bet here will be voltage injection, though. Keep in mind that short circuits can occur due to defects internal to the board, and in those case, you won't find them without the use of X-ray equipment.
 
RazorWind said:
I have a reference 980 Ti here, and it's definitely different from the photo you posted. I wouldn't rely too heavily on it being that similar.

You're going to need to look for datasheets for the other ICs on the board, and look for which ones call for 3.3V. Then, find the pin where they're supposed to have 3.3V and check for a short to ground. Keep in mind that there are likely to be a bunch of capacitors on that rail, and one of them could be shorted too.

Your best bet here will be voltage injection, though. Keep in mind that short circuits can occur due to defects internal to the board, and in those case, you won't find them without the use of X-ray equipment.
Click to expand...
I will take a look at it tonight. I will let you know if i have more information about it.
 
The GTX Titan is Kepler, which means it would be analogous to a GTX 780 as seen below.

GTX Titan:
1616076376211.png


GTX 780:
1616076460952.png
 
You must log in or register to reply here.
Back
Top