Could You Fix This GTX 1080 - part 2. Graphics Card Not Detected

The first part of this repair attempt was published a couple of weeks ago and prompted a lot of comments, discussion and speculation. So in part two I explore all the various suggestions made in an attempt to make this graphics card detect. Wanna know? You just have to press PLAY.

I work in collaboration with:
The Electronics Channel (with Carlos and Detlef)

Gran Canaria Uncovered

For All Your PCB needs: free $5 discount coupon

Equipment used in my videos. These are affiliate links, you pay the normal price and I make a small commission.

TEST METERS
ANENG AN8009 MULITIMETER
KAIWEETS HT118E MULTIMETER
VC480C+ MILLIOHM METER
MESR-100 ESR METER
XC6013L CAPACITOR METER

TM-902C TEMPERATURE METER
LCR-T4 COMPONENT ANALYZER
FNB58 USB ANALYZER
PCI POST ANALYZER
TL460S PLUS PCI_E ANALYZER
TOOLTOP ET120MC2 SCOPE
FNIRSI 1014D SCOPE
NPS3010W 30V 10A PSU
T12 STATION WITH M8 9501 HANDLE
M8 9501 HANDLE
YIHUA 982 Soldering Station C210/C245
FNIRSI HS-02 PORTABLE SOLDERING C210/C245
QUICK 861DW
PROS'KIT SS-331H

INFIRAY P2 PRO IR CAMERA + MACRO LENS
TOOLTOP ET13S THERMAL CAMERA METER
ET13S MACRO LENS
OPTICAL MICROSCOPE
TOMLOV TM4K AF FLEX
TL866 II+ PROGRAMMER
CH341 PROGRAMMER
NC-559-ASM FLUX

DESOLDER BRAID (I use size 8045)
HX-T100 SOLDER (0.6MM)
ESD-11 TWEEZERS

If you would like to support this channel
You can send donations
You can subscribe to Patreon
You can click Join to become a channel member

Thank you
Richard

[Alvin]

I already commented on the first video: Just because a signal is not connected to the motherboard on the riser cable, doesn't mean the signal isn't used, it could be generated on the riser board itself and only be required by the GPU, not the mainboard, so it doesn't have to go back to the slot on the mainboard. According to a quick Google search, both SMB rails should be pulled up to 1.8V. But I doubt that's the issue here (I think it might be the GPU doing the pull-up, then it wouldn't matter anyway), the reset signal is more concerning (I think wake is optional, shouldn't be required to start the card). Maybe there's a broken trace or a short to a different power rail somewhere?

[krzysztofbednarek]

Hello from Norwich . Good job . I like it . Thank You . I love the oscilloscope . I have too

[wherami]

Wow talk about a rabbit hole. That was a wild ride

[andymouse]

I hope we get a part 3 ! when you was saying " we have a clock" you had the timebase right up (fast) so what is that on your scope dial ? and what is the clock supposed to be ? more info on timebase and V/Div would be great for us watching ! with a timebase that high on the scope most of what could be seen looked like high frequency rubbish to me Lol! it would it have been useful to measure that clock just to be sure or have you seen it a million times ? again It would be useful to know a value. When you talked about TX/RX on those pins is that just a UART ? I am literally learning about GPU's from scratch so please forgive dumb questions !...cheers.

[marcellipovsky]

Be careful Richard, there seem to be rabbit hole in the board ;-)

[ORRAMIXDJ]

Hi Richard thanks for the video. You can check the 33 ohm resistor between bios and gpu desoldering the bios chip and the pull up resistor, then you can measure the resistance value or the voltage drop in diode mode trough the pin 1 pad and ground. If you get open line the probability that the resistor is faulty is high. In order to found that resistor I suggest to put at the top of the multimeter probe a little brush made by some layer of desoldering wick, then you can check in continuity mode which one resistor is touching the resistors in the back of the gpu (usually it is placed there or near the bios chip).
Hoping can help!

[KBUIF]

Hi Rich, I don't see a link to the PCI-e x1 to x16 mining adapter in your list of equipment. Maybe be worth adding a link?!

[Dutch_off_grid_homesteading]

Heya, oke I think I get it, but still a bit blurry for a newbee on a difficult problem as this

[ralphj]

Lost track of which card you were on, but the 0.8V (PCI reset?) is suspicious. Perhaps lift any series resistors etc and / or put finger on to see if floating. You could lift the enable pin of the BIOS IC and pull it permanently low, but stand by with an extinguisher, new pants etc (or trousers, depending on country). Ideally need some status LEDs on the reset, wake etc signals (on the adaptor board) and a storage scope

[Xorat]

Hey Richard i do have a off topic question. I saw a video where someone was talking about tools for repair business and they insisted that you need a linear power supply for repair work and that a switching one would be bad. they didn't give any reasoning. I couldn't think of a reason why switching one would be inferior to a linear one. any idea ?
best regards!

[theshemullet]

Thanks. Have you any advice on a cheap oscilloscope to buy, say around €200.