Calling all Graphics card fault finding genius's GTX 970 4 GB FAULT

[ag3ntwilson]

Hey Fellow Geeks

I am a bit of trouble with my gtx 970 graphics card

Inno3d GTX 970 OC 4GB GDDR5

Setup - Intel 6600K CPU - MSI Z270 BOARD- 600w Corsair PSU - 16GB Corsair DDR4 -

Tested my RIG with a mates GTX 1080Ti 11gb , with no issues , so its definitely the card not the rig.

the issue i am having is - and what i have done:

NOTE: All Steps are carried out while plugged into PCI Slot of a working motherboard with working Power Supply


Install both 6-pin cables (2x 6pins) *NO POWER - NO BOOT*
Install 6pin cable into only port A (1x6pins) *BOOTS* Message on screen to install other 6 pin plug*
Install 6-pin cable into only port B (1x6pins) *NO POWER - NO BOOT*

so by trial and and error the issue seems to be the 6 pin port B or some capacitor/resistor/ etc etc that links to it. i have included pictures with this post of the board , with a series of parallel capacitors and mosfets.
can anyone make end of this , and inform me of what might be causing this , it powers up and im not ready to give up on this yet ? any ideas chaps ? also there is absolutely no sign anywhere on the board of anything blown burnt or faulty

thanks in advance

IMAGES:
https://ibb.co/ddi1Bw
https://ibb.co/dZA6Bw
https://ibb.co/fuecJb
https://ibb.co/bZZBdb

 

[DRagor]

When it does not boot, how exactly it behaves? No reaction at all, or does it starts to boot and then suddenly stops? If the latter, then most likely some kind of short in GPU circuits connected to port B.
Btw, have you made sure it's not the specific PSU cable or plug that's actually causing the problem, not the GPU port?

 

[rolli59]

Apparently there is a short which really can be anywhere in the PCB so giving up may be the best option.

 

[ag3ntwilson]

When it does not boot, how exactly it behaves? No reaction at all, or does it starts to boot and then suddenly stops? If the latter, then most likely some kind of short in GPU circuits connected to port B.
Btw, have you made sure it's not the specific PSU cable or plug that's actually causing the problem, not the GPU port?

hiya thank you for the prompt reply, it tries , about a quarter of a second it tries to turn on for , board leds light up and fans kick in then straight off again,
i have a multi meter here but i dont know where to start really , what to test , how to test , i was hoping it was a case of replacing the caps or mosfets
oh and yes i have interchanged the wires , not the wires or power supply ,

 

[ag3ntwilson]

Apparently there is a short which really can be anywhere in the PCB so giving up may be the best option.

thanks for reply

i appreciate what your saying , the short could be anywhere, but i was wondering if any visible components linked to PORT B might be able to be desoldered and replaced , the bigger componants not like the micro parts?

not really wanting to give up on this if its going to be something that can be replaced for pennies at a electrical hardware store

 

[rolli59]

Apparently there is a short which really can be anywhere in the PCB so giving up may be the best option.

thanks for reply

i appreciate what your saying , the short could be anywhere, but i was wondering if any visible components linked to PORT B might be able to be desoldered and replaced , the bigger componants not like the micro parts?

not really wanting to give up on this if its going to be something that can be replaced for pennies at a electrical hardware store

Unless you see obvious damage to a component, like blown/bulging capacitor, it is almost impossible to identify the faulty component.

 

[ag3ntwilson]

Apparently there is a short which really can be anywhere in the PCB so giving up may be the best option.

thanks for reply

i appreciate what your saying , the short could be anywhere, but i was wondering if any visible components linked to PORT B might be able to be desoldered and replaced , the bigger componants not like the micro parts?

not really wanting to give up on this if its going to be something that can be replaced for pennies at a electrical hardware store

Unless you see obvious damage to a component, like blown/bulging capacitor, it is almost impossible to identify the faulty component.

ok right i may be onto something if you bear wityh me im going to use classic ms paint to draw on my photo to show what i hope (fingers crossed is the issue , i have been googleing ohm testing and i have set the multimeter up to ohms 200 and have stubled across something , bear with me while i just draw what i have found,
i may be way off but let me know what you think
thanks

 

[ag3ntwilson]

When it does not boot, how exactly it behaves? No reaction at all, or does it starts to boot and then suddenly stops? If the latter, then most likely some kind of short in GPU circuits connected to port B.
Btw, have you made sure it's not the specific PSU cable or plug that's actually causing the problem, not the GPU port?

OK so i may have found the issue (fingers crossed) i hope this makes sense , i'm not even 100% sure it does to me but here goes.

so on each mosfet in the picture there 4 terminals either side
1 2 3 4
| | .| .|
1 2 3 4
Like this

i checked across 1 to 1 - 2 to 2 - 3 to 3 - 4 to 4 on each mosfet
each one reading a steady 23...... not sure what the 23 actually is ....ohms, watts, volts, conkers. whatever its displaying on multi meter

all chips show 23 on every single terminal, however there is one chip (the one with the orange star on it) where terminal 4 to 4 reads 73 , considerably higher than the rest of the chips. could this be the issue ?

@rolli59 @DRagor

https://ibb.co/dOS38b


many thank yous

 

[ag3ntwilson]

Apparently there is a short which really can be anywhere in the PCB so giving up may be the best option.

thanks for reply

i appreciate what your saying , the short could be anywhere, but i was wondering if any visible components linked to PORT B might be able to be desoldered and replaced , the bigger componants not like the micro parts?

not really wanting to give up on this if its going to be something that can be replaced for pennies at a electrical hardware store

Unless you see obvious damage to a component, like blown/bulging capacitor, it is almost impossible to identify the faulty component.

OK so i may have found the issue (fingers crossed) i hope this makes sense , i'm not even 100% sure it does to me but here goes.

so on each mosfet in the picture there 4 terminals either side
1 2 3 4
| | .| .|
1 2 3 4
Like this

i checked across 1 to 1 - 2 to 2 - 3 to 3 - 4 to 4 on each mosfet
each one reading a steady 23...... not sure what the 23 actually is ....ohms, watts, volts, conkers. whatever its displaying on multi meter

all chips show 23 on every single terminal, however there is one chip (the one with the orange star on it) where terminal 4 to 4 reads 73 , considerably higher than the rest of the chips. could this be the issue ?

@rolli59 @DRagor

https://ibb.co/dOS38b


many thank yous

 

[rolli59]

Apparently there is a short which really can be anywhere in the PCB so giving up may be the best option.

thanks for reply

i appreciate what your saying , the short could be anywhere, but i was wondering if any visible components linked to PORT B might be able to be desoldered and replaced , the bigger componants not like the micro parts?

not really wanting to give up on this if its going to be something that can be replaced for pennies at a electrical hardware store

Unless you see obvious damage to a component, like blown/bulging capacitor, it is almost impossible to identify the faulty component.

OK so i may have found the issue (fingers crossed) i hope this makes sense , i'm not even 100% sure it does to me but here goes.

so on each mosfet in the picture there 4 terminals either side
1 2 3 4
| | .| .|
1 2 3 4
Like this

i checked across 1 to 1 - 2 to 2 - 3 to 3 - 4 to 4 on each mosfet
each one reading a steady 23...... not sure what the 23 actually is ....ohms, watts, volts, conkers. whatever its displaying on multi meter

all chips show 23 on every single terminal, however there is one chip (the one with the orange star on it) where terminal 4 to 4 reads 73 , considerably higher than the rest of the chips. could this be the issue ?

@rolli59 @DRagor

https://ibb.co/dOS38b


many thank yous

I simply do not know enough about the power distribution to say.