[SOLVED] Help with troubleshooting a dead Zotac 3060 Ti GA104-200-A1 ?

Feb 7, 2022
3
0
10
0
So, due to Zotac's lack of proper product warranty (based on age from serial nr etc) I'm now on my own in trying to get my belowed 3060 Ti to work agian.
Card was bought in June 2021, but with a pre-built PC, which I again bought 2nd hand in August. The PC I (unlucky for me) sold on after taking out the card. If I had it I could of course get it on warranty by handing the whole PC in where I bought it, but that ain't possible, so here I am...

I've taken apart the card now and started measuring with the multimeter on the different "power stages", however I'm still very confused about what is what,

I'm attacing a couple of pics. Is it correct that the 1 and 2 (orange circles) are power controllers, and if so, for what?

I've only been able to measure 12v and 3.3v here and there around the board, the 5volt doesn't seem to be present at all.

Also, what I believe to be the resistance between earth and the core I measure to 1.2-1.3 Ohm. I hope this means that the core is likely not dead?

Would appreciate if anyone can give me any clues how to get on with this.

It also struck me why there is no voltage on the two silver/pink coils on the left row, are these probably 5v coils, the other silver/black ones measure 12v on the right side as illustrated.





Thanks in advance for any relevant input, or a redirection to a different forum where there might be a bigger chance of getting any answers.

Regards

Kristian
 
Last edited:

Ralston18

Titan
Moderator
Unfortunately answers (full disclosure) may be hard to come by....

What all was happening (or not happening for that matter) that caused you to remove the GPU? Overheating, sparks, smoke, no signal errors, static screens, crashes, error codes?

Has the GPU been tested in another known working computer? Which may, fair warning, put the test computer at risk if the card was damaged via the previous host computer....

= = = =

You need the applicable board schematic and troubleshooting documents that list the expected voltages and resistance between any given points.

And to identify specific components such as controllers. (I could not read the markings on the orange circled components.)

And in some cases, the measurements may require a test plug or jumper of some sort be installed as part of the troubleshooting process.

The cards are not really designed and made to be repairable. And even a bad component is found, de-soldering it and soldering in a replacement (if a replacement can be found) is likely to cause damage.

Some of the chips appear a bit grainy or bubbly on top - possible sign of over-heating and the chip may be internally damaged.

Just my thoughts on the matter....
 

Ralston18

Titan
Moderator
Unfortunately answers (full disclosure) may be hard to come by....

What all was happening (or not happening for that matter) that caused you to remove the GPU? Overheating, sparks, smoke, no signal errors, static screens, crashes, error codes?

Has the GPU been tested in another known working computer? Which may, fair warning, put the test computer at risk if the card was damaged via the previous host computer....

= = = =

You need the applicable board schematic and troubleshooting documents that list the expected voltages and resistance between any given points.

And to identify specific components such as controllers. (I could not read the markings on the orange circled components.)

And in some cases, the measurements may require a test plug or jumper of some sort be installed as part of the troubleshooting process.

The cards are not really designed and made to be repairable. And even a bad component is found, de-soldering it and soldering in a replacement (if a replacement can be found) is likely to cause damage.

Some of the chips appear a bit grainy or bubbly on top - possible sign of over-heating and the chip may be internally damaged.

Just my thoughts on the matter....
 
Feb 7, 2022
3
0
10
0
Just noticed both images were of the front of the card.

It is not my image though, I just used it to start to try to figure out what components are what, and to see where there is voltage or not.

I can't really see anything that look burned on my card, it is just not turning on.

I tried it in several other computers. Fans doesn't move or anything when it is powered on.

It was (god forbid) mounted in a computer along with a 1660 S, using it as a heater for cloth drying (during mining). Card just died during one night, computer had rebooted and only the 1660 S were now visible in Device Manager.

They might have been running a bit on the hot side as it wasn't the spaciest case they were in, but I did add two 140mm fans that were blowing air straight onto them, and they kinda were operating like this for 4-5 months before it died.

I don't think I'll be able to find any schematics, so I probably just need to try to figure out a bit more of the start-up logic on the card myself to try to see if it possibly can be fixed or not.
 
Feb 7, 2022
3
0
10
0
Continuing my kinda monologue here.

I actually found a breach on the board now, and it seem to me that the broken copper line connects where I drew the blue line.
Maybe it really is as simple as soldering an piece of wire between those two components?

I suppose I need to measure some of the lines for shorts afterwards though before I try to "plug it in" if I decide to try to solder it.

 
Aug 11, 2022
1
0
10
0
Continuing my kinda monologue here.

I actually found a breach on the board now, and it seem to me that the broken copper line connects where I drew the blue line.
Maybe it really is as simple as soldering an piece of wire between those two components?

I suppose I need to measure some of the lines for shorts afterwards though before I try to "plug it in" if I decide to try to solder it.

Hello.. I have the same line burn from the same card. followed by a short circuit, did you solve the problem.. Thanks..
 

ASK THE COMMUNITY