Not getting detected means that it is not talking to the GPU which is expected as the GPU isn't getting power. Since the videocard gives all the voltages besides the Vcore, I don't see the reason being a short somewhere as the whole card would not boot as 12V goes directly to Vcore mosfets. It is likely something related to voltage controller, most likely a resistor similar to what you found on 5V rail. However I didn't see you measuring if the card gets 3.3V from the PCIE slot. PEX rail is far as I know uses 3.3V converted to 1V, but I think it also needs Vcore first, which makes me to believe that it is indeed the voltage controller. It is not BIOS related since you should get boot voltages even if the BIOS is corrupt.
Hi there again. Vintage injection on the main power line that gives power to the CPU MosFets and also check the small transistors that controls that power line. Usually the are on the right side in the board near the 12V line from external 12v.
The controller of that card sense short on the CPU mostfets and sometimes shout down the main power line. That is the reason that im telling you yo try voltage injection. Usually is one of the last MosFets from top so bottom
This is like a soap opera, you watch with suspense building up until the end, only to be told story to be continued. Look forward to episode 2! keep up the good work!
With so much display cards to diagnose and fix, I would suggest to get a good Usb based PCI expansion card. That way you can avoid damaging the motherboard with constant plug in and plug out of video cards. Also it would be easy to lay the card flat on the table while powered up and do the checks under the microscope.
I have one but don't use it. The card is still attached to the power supply cables and can't move it under the microscope without moving the power supply and everything else.
Love those GPU diagnosis and repairs, i still have a Sapphire ATI R9 290 4GB which has a short on the 12v line but i'm still unable to pinpoint the exact problem. thanks for al your knowlegde Alex
I have r9 280x, I have similar problem like customer in this vid, mine works after cleaning but soon after, poof, blue lines on screen and its dead. Where can I learn to diagnose the cards issue ? I have a solid multimeter but I dont know all this about 12v and 3v rail etc. Where can I learn this ? And it seems like most newer cards have similar voltage rails, designs etc ?
If you can get a cheap thermal camera, you might be able to apply power from a current controlled supply to the 12v rail and see where the short is dissipating power.
@@sbrewski27 yes i do. I using my msi rx 470 gaming x 4gb which works perfect but im trying to learn to fix gpu's and other stuff like ps4's and so. My next card will be at least 8gb or more but the prices are really high 🙃
Alex, I'm leaning toward this being a MOSFET. Because if you suspected it were the GPU I think you would have called it quits on this one. Also, you're getting your 12v as well as 5v. Let's see what things look like next time.
hey alex you are now a real pro at gfx-cards. it's unbelievable how much knowledge you've acquired over the years, not only in terms of gfx cards but also electronics in general. great, keep it up and especially with the videos. they are so incredibly informative and helpful. thank you, greets
Been using this exact same card for many years, it's running 24/7 and doing a crap ton of rendering work. Hope it doesn't die on me, I feel like MSI cards were better back then.
1000% agree with you man, and it's not just the cards. I have an old MSI 1080p 120hz monitor i gave to my dad when i upgraded my station to a 1080p165hz ultrawide. That one died, and MSI replaced it with a 1440p144hz monitor which so far has been good. I had to replace thermal pads on my MSI 3090 suprim x since the ones on the VRAM were already starting to slip off when i peeled off the heatsink, and when i removed them to replace, they were very oily underneath. My old 1070 and 1080 ti SLI builds are still going strong after three years each
allways waiting your videos they are lessons like no other for someone on a place where learn this kinds of things it really expensive and dificult, thanks master and waiting to see that card leave better than factory!
i thought you needed the pex rail to turn on vcore. edit i was wrong its probably the voltage controller because that is what turns pex and vcore on. memory i believe has a separate voltage controller which would explain why mem is getting voltage and vcore is not.
My cousin called me with the same problem on the same card 2 weeks ago. Wasn't until he sent a video that I realized he was plugging the DP cable into the motherboard not the GPU 😂. Nothing was broken
Same with my friend. I've put together pc with i7 2600, 16gb ram and 1650. Tested it, temperatures were ok. I gave it to him, and 2 days later he said that pc wont turn on. At weekend i went and visited him, checked and blamed him(but friendly :D). As he had his TV & monitor plugged into mobo
At 8:47 you said that the pex rail is at 0. So this is probably your issue as it should read between 6 to 12 ohm. You should check the resistor next to it. It's probably a 0 ohm that is bad..
Alex He is one of those people who keeps getting better and better as they work on a particular item, until they solve everything easily. love gpu repairs cant wait for part2
You should be checking if Vcore Pwm controller is getting all of its voltages. The chip is up9511p, it's a qfn-50 chip. Start with a visual inspection around it, then check 5V VCC at pin 12 which is usually connected to a nearby capacitor, and check enable on pin 36 which should read 2.4V. Enable is an output from the controller which gets pulled low by a dual n-channel mosfet if 1.8V is missing on the card. Also check pgood on pin 38, iirc it should be 1.8V. Pgood is an input that gets shorted to ground by the controller istelf if the controller detects a fault or is faulty istelf. So if Pgood is missing, it is advisable to check the gate pins of the high side mosfets for Vcore.
You can check the resistance reading of the PEX rail and if it is unsusually low it is a good indication of a shorted core. That goes for mem and gpu rails as well. But i dont think that's the case because that wouldn't explain how the 5v enable resistor died. So my guess would be that something is wrong with the conversion circuit here as well. I hope you can fix the card. Love your videos keep up the good work!
Hi Alex, I am a tech from Canada. Since the customer indicated the loss of video output was intermittent that's a strong indication the GPU needs to be reflowed or possibly re-balled. The voltage spikes caused by the loosening of the GPU chip made the resistor go into high resistance. I suggest you do a quick reflow to verify the card works and if so re-balled it or do a proper reflow. Will be watching next video to see how it goes. Good luck
Assuming the customer is telling the truth. I wonder if the faulty resistor was previously at a resistance that caused the voltage to drop to a level that was hovering between enabled and not enabled (maybe 3 or 4K) causing the intermittent behaviour. Could the intermittent behaviour caused something else to fail. Could another part on the PCB be faulty and that caused the resistor to go faulty. I go for a random guess of a fault mosfet.
@@pow9606 Hey POW, You have a very good point. However, according to customers story, it worked again when they cleaned the card. I also took that point into consideration because moving around the PCB might have made the GPU chip to reattach barely. Your reasoning in regards to the resistor would make the board behave same manner. A MOSFET would have failed immediately in my experience. I really would like Alex to remeasure the resistance of the replaced resistor to see it still at rated value. If not, we can lean towards your theory. Thx
There should be 12v and 3.3v coming from the PCI slot. Check the components that come from the PCI power. The bios chip doesn't seem to be getting power :)
Not just the components on the bottom of the card can fail. Try checking top side, or the side that does not have fans on. Fixed a couple of cards with failed caps and resistors on top side.
If you were still taking video card repairs, I can't imagine how long your backup would be right now. The market is still pretty bad. I am sure many they are older cards they upgraded from and they figure they can pay the repair cost and recoup the cost and a profit by reselling them.
Maybe there's something in the same circuit as that high reading resistor that has went open circuit before the resistor burnt up. And that's why there's no short!
Excuse me sir, I want to ask. I'm repairing my VGA card, there is a short on one of the MOSFETs, I have replaced it and it is no longer short. but when I test, short comes back on the same mosfet. thanks
My concerns would be customer took it apart to clean. A did they put new thermal compound on it and did the screw it down level. Could be a board fracture or warped. Overall I think gpu has gone. Would be nice to know what you find.
Pls check EN signal which should be a bit above 2 volts, Vin should be 3.3 V , vdd 5 v, if any one or more of them is missing then check the circuitry around most probaly there is a faulty resistor. if all ok change the controller.
My brother thank u this is not related to the video i was too buys repairing this hash board for that had a broken cap. job looks great i found my soldering iron, soldering iron with the sucktion adaptor, heat gun with a small tip. All cuz of your video i have learned how to replair cap and how they can blow. worked from the back of the board and the doner board one just fell out. Anyways take care my friend. I need some of that kill flux u have.
dont you think that this resistor died because of a proper reason ? there is some thing around it caused it to die , its not a filter or 0 ohm resistor that would normally die in case of power problem , so you should suspect a chip being shorted around it ,
Yeah, it's time to inject voltage and look for the hot spots. I did see where moisture corroding certain parts could be a problem and that should show up on the heat camera. Then, I see that if a graphics card is run on a game NOT using the right driver could be the problem as well And of course the ever present over heating of a fried component but if not visually evident then go for the voltage inject and look for the hot spot. Oh, and what about pulling the GPU and cleaning the old paste off and then resetting it with some fresh past to deter any heat problems. One more idea is dirty board causing problems. An alcohol dip maybe?
if you look on thermal the resistor you replaced will be heating up, the chip it feeds will likely be faulty, if you leave power on the card long enough the resistor. will fail again
Since the customer said I clean it and it worked for a day, I assume the GPU chip has disconecting. And when the card has a separator solder, it means it overheats while working. So that resistance burns out due to overheating. The best way to test it, put just a little heat on GPU, if the card works you have to re-ball it
@@golamrabby5335 Either you know it from schematics, or you can take as good that, in the same productive process, capacitors with same color and same size should have same value. Taking this as a fact, you could measure (off circuit) another capacitor with same characteristics installed in the same board.
I have a 3080 when I turn on my PC graphics card cpu Everythings is running but VGA light on motherboard stays on and no display i changed ram psu and everything but display
Thermalcam probobly will show a short in some of VCORE mosfett. Thats my guess. I hope that it doesn't show heat spot on GPU-DIE. It is possible because self degradation of thermal paste and bad heat spread. If is THAT case the GPU is completely dead 😢 Hope GPU is not fried and this graphics card can be resurrected! We will see in the next video!
Not getting detected means that it is not talking to the GPU which is expected as the GPU isn't getting power.
Since the videocard gives all the voltages besides the Vcore, I don't see the reason being a short somewhere as the whole card would not boot as 12V goes directly to Vcore mosfets.
It is likely something related to voltage controller, most likely a resistor similar to what you found on 5V rail.
However I didn't see you measuring if the card gets 3.3V from the PCIE slot. PEX rail is far as I know uses 3.3V converted to 1V, but I think it also needs Vcore first, which makes me to believe that it is indeed the voltage controller.
It is not BIOS related since you should get boot voltages even if the BIOS is corrupt.
We have 3.3v and 12V on PCIE and Bios is getting all the right voltages. I will do more measurements later today.
Hi there again. Vintage injection on the main power line that gives power to the CPU MosFets and also check the small transistors that controls that power line. Usually the are on the right side in the board near the 12V line from external 12v.
The controller of that card sense short on the CPU mostfets and sometimes shout down the main power line. That is the reason that im telling you yo try voltage injection. Usually is one of the last MosFets from top so bottom
Do we have power enable? Just concern with the chip near error resistor
Is it also possible that the person blew or knocked of a componet when they where cleaning the card or static built up by the blower
This is like a soap opera, you watch with suspense building up until the end, only to be told story to be continued. Look forward to episode 2! keep up the good work!
GPUs are tougher to diagnose and repair. I am not giving up! I want to continue learning. Keep up the good work. Thx!!!!!!
With so much display cards to diagnose and fix, I would suggest to get a good Usb based PCI expansion card. That way you can avoid damaging the motherboard with constant plug in and plug out of video cards. Also it would be easy to lay the card flat on the table while powered up and do the checks under the microscope.
I have one but don't use it. The card is still attached to the power supply cables and can't move it under the microscope without moving the power supply and everything else.
Hum! Necessity is the mother of invention. Thinking.
@@NorthridgeFix Thats some thing you should be able to work around with cable extensions.
Agree! I am learning from you need to be able to evaluate both side of the card under the microscope. Off course without creating a new short.
Hahaha take look closely to the table .
No space for new tools hehehe
Then suggesting a cable hahaha
That so funny
Really love the video Card repairs - they are the most fun to watch - ty for the great videos :)
Love those GPU diagnosis and repairs, i still have a Sapphire ATI R9 290 4GB which has a short on the 12v line but i'm still unable to pinpoint the exact problem. thanks for al your knowlegde Alex
The 1070 would be a nice upgrade assuming you dont have another working card.
I have r9 280x, I have similar problem like customer in this vid, mine works after cleaning but soon after, poof, blue lines on screen and its dead.
Where can I learn to diagnose the cards issue ?
I have a solid multimeter but I dont know all this about 12v and 3v rail etc.
Where can I learn this ?
And it seems like most newer cards have similar voltage rails, designs etc ?
If you can get a cheap thermal camera, you might be able to apply power from a current controlled supply to the 12v rail and see where the short is dissipating power.
@@sbrewski27 yes i do. I using my msi rx 470 gaming x 4gb which works perfect but im trying to learn to fix gpu's and other stuff like ps4's and so. My next card will be at least 8gb or more but the prices are really high 🙃
@@nicholassage2076 there are more youtubers that really go in depth to all functions on a GPU so search and you will find
Alex, I'm leaning toward this being a MOSFET. Because if you suspected it were the GPU I think you would have called it quits on this one. Also, you're getting your 12v as well as 5v. Let's see what things look like next time.
hey alex you are now a real pro at gfx-cards. it's unbelievable how much knowledge you've acquired over the years, not only in terms of gfx cards but also electronics in general. great, keep it up and especially with the videos. they are so incredibly informative and helpful. thank you, greets
Been using this exact same card for many years, it's running 24/7 and doing a crap ton of rendering work. Hope it doesn't die on me, I feel like MSI cards were better back then.
1000% agree with you man, and it's not just the cards. I have an old MSI 1080p 120hz monitor i gave to my dad when i upgraded my station to a 1080p165hz ultrawide. That one died, and MSI replaced it with a 1440p144hz monitor which so far has been good. I had to replace thermal pads on my MSI 3090 suprim x since the ones on the VRAM were already starting to slip off when i peeled off the heatsink, and when i removed them to replace, they were very oily underneath.
My old 1070 and 1080 ti SLI builds are still going strong after three years each
@@GenDrag1 If I remember correctly MSI's old CEO died some years ago, maybe that's related to the decline of their reputation since then.
My 1070 just died last week after being on 24/7 for 6 years. My 660ti is still going strong after 11 years... I wonder if it's worth getting repaired.
allways waiting your videos they are lessons like no other for someone on a place where learn this kinds of things it really expensive and dificult, thanks master and waiting to see that card leave better than factory!
i thought you needed the pex rail to turn on vcore.
edit i was wrong its probably the voltage controller because that is what turns pex and vcore on. memory i believe has a separate voltage controller which would explain why mem is getting voltage and vcore is not.
My cousin called me with the same problem on the same card 2 weeks ago. Wasn't until he sent a video that I realized he was plugging the DP cable into the motherboard not the GPU 😂. Nothing was broken
Same with my friend. I've put together pc with i7 2600, 16gb ram and 1650. Tested it, temperatures were ok. I gave it to him, and 2 days later he said that pc wont turn on. At weekend i went and visited him, checked and blamed him(but friendly :D). As he had his TV & monitor plugged into mobo
🤓🤓🤓🤓🤓
@@mmllmmll22 🤓🤓🤓🤓🤓
At 8:47 you said that the pex rail is at 0.
So this is probably your issue as it should read between 6 to 12 ohm. You should check the resistor next to it. It's probably a 0 ohm that is bad..
Alex He is one of those people who keeps getting better and better as they work on a particular item, until they solve everything easily. love gpu repairs cant wait for part2
the suspense is killing me.. looking forward to the 2nd part of this repair attempt
I love the GPU videos the most even if they are a no fix a lot of the time.
I have never had a problem with mine from new for years. Still works.
Thank you! saved my GPU, same issue but the resistor replacement solved it!
You should be checking if Vcore Pwm controller is getting all of its voltages.
The chip is up9511p, it's a qfn-50 chip. Start with a visual inspection around it, then check 5V VCC at pin 12 which is usually connected to a nearby capacitor, and check enable on pin 36 which should read 2.4V. Enable is an output from the controller which gets pulled low by a dual n-channel mosfet if 1.8V is missing on the card.
Also check pgood on pin 38, iirc it should be 1.8V. Pgood is an input that gets shorted to ground by the controller istelf if the controller detects a fault or is faulty istelf.
So if Pgood is missing, it is advisable to check the gate pins of the high side mosfets for Vcore.
Master of his Craft, Love these videos !
Look forward to episode 2
You can check the resistance reading of the PEX rail and if it is unsusually low it is a good indication of a shorted core. That goes for mem and gpu rails as well. But i dont think that's the case because that wouldn't explain how the 5v enable resistor died. So my guess would be that something is wrong with the conversion circuit here as well. I hope you can fix the card. Love your videos keep up the good work!
Hi Alex,
I am a tech from Canada. Since the customer indicated the loss of video output was intermittent that's a strong indication the GPU needs to be reflowed or possibly re-balled. The voltage spikes caused by the loosening of the GPU chip made the resistor go into high resistance. I suggest you do a quick reflow to verify the card works and if so re-balled it or do a proper reflow. Will be watching next video to see how it goes. Good luck
Assuming the customer is telling the truth. I wonder if the faulty resistor was previously at a resistance that caused the voltage to drop to a level that was hovering between enabled and not enabled (maybe 3 or 4K) causing the intermittent behaviour. Could the intermittent behaviour caused something else to fail. Could another part on the PCB be faulty and that caused the resistor to go faulty. I go for a random guess of a fault mosfet.
@@pow9606 Hey POW,
You have a very good point. However, according to customers story, it worked again when they cleaned the card. I also took that point into consideration because moving around the PCB might have made the GPU chip to reattach barely. Your reasoning in regards to the resistor would make the board behave same manner. A MOSFET would have failed immediately in my experience. I really would like Alex to remeasure the resistance of the replaced resistor to see it still at rated value. If not, we can lean towards your theory. Thx
yo opino lo mismo
There should be 12v and 3.3v coming from the PCI slot. Check the components that come from the PCI power. The bios chip doesn't seem to be getting power :)
Not just the components on the bottom of the card can fail. Try checking top side, or the side that does not have fans on. Fixed a couple of cards with failed caps and resistors on top side.
Yay! Another GPU video! Bless my bro'
thank you for doing more gpu repairs, even if they are not successful we learn a lot with each video.
No follow up video for this one? :'(
love your videos, stay motivated!
Any ideas what happened to the second part of this video?
It went to the 9th Dimension ;)
Thank you for the invaluable info, i wish you document GPU's based on your experience if you happen to get some free time
If you were still taking video card repairs, I can't imagine how long your backup would be right now. The market is still pretty bad. I am sure many they are older cards they upgraded from and they figure they can pay the repair cost and recoup the cost and a profit by reselling them.
That card was released in 2016. I'm amazed people still use it.
not surprising at all, 10 series cards are still very popular
@NorthridgeFix Hello! i wanna buy EVGA gtx 1070 SC with acx 3.0 it used without thermal pads for gaming what can you say about it ?
Keep up don't give up
Hi sir, Does it have to be in large quantities to buy mosfet and other parts?
Did you repair it?? i would like to see that video if you did :)
Maybe there's something in the same circuit as that high reading resistor that has went open circuit before the resistor burnt up. And that's why there's no short!
please what model of motherboard you are using for test ?
Thank you from kurdistan of iraq...
Excuse me sir, I want to ask. I'm repairing my VGA card, there is a short on one of the MOSFETs, I have replaced it and it is no longer short. but when I test, short comes back on the same mosfet. thanks
Hmmmm I’m getting a Msi 1070 with the same type of conditions.. I’ll have a look if it’s the same problem
possibly a resistor or a fuse / 0 ohm resistor issue
What happened perhaps somehow during cleaning customer fried component with electro static
Most of the time when the resistor value increase the reason may be there will be. Leaking diode near by.
My concerns would be customer took it apart to clean. A did they put new thermal compound on it and did the screw it down level. Could be a board fracture or warped. Overall I think gpu has gone. Would be nice to know what you find.
Waiting for part 2
Yea it didnt happen
these card use UP9511P phase controller for Vcore?? maybe it missing EN signal to it
Pls check EN signal which should be a bit above 2 volts, Vin should be 3.3 V , vdd 5 v, if any one or more of them is missing then check the circuitry around most probaly there is a faulty resistor. if all ok change the controller.
Is that what a GPU looks like? I wouldn't know because I can't get a hold of one.
My brother thank u this is not related to the video i was too buys repairing this hash board for that had a broken cap. job looks great i found my soldering iron, soldering iron with the sucktion adaptor, heat gun with a small tip. All cuz of your video i have learned how to replair cap and how they can blow. worked from the back of the board and the doner board one just fell out. Anyways take care my friend. I need some of that kill flux u have.
How is the GPU, could it have received a spike when the resistor blew?
We can’t repair much without the Schematic when it comes to knowing the component values
I've got exactly the same card but don't have 12v on step down converter but have 3v3 enable.
Try to flash the bios and replace the IC of the VRM.
dont you think that this resistor died because of a proper reason ?
there is some thing around it caused it to die , its not a filter or 0 ohm resistor that would normally die in case of power problem , so you should suspect a chip being shorted around it ,
did you ever fix this one?
I just bought 2 radeon r9 390x and on both gpu's the capacitors are full of shorts.
Is it real that both are fried?
I'd love to work with your show I've got really good at graphics card repair and look forward to see more videos from you
Yeah, it's time to inject voltage and look for the hot spots.
I did see where moisture corroding certain parts could be a problem and that should show up on the heat camera. Then, I see that if a graphics card is run on a game NOT using the right driver could be the problem as well
And of course the ever present over heating of a fried component but if not visually evident then go for the voltage inject and look for the hot spot.
Oh, and what about pulling the GPU and cleaning the old paste off and then resetting it with some fresh past to deter any heat problems.
One more idea is dirty board causing problems.
An alcohol dip maybe?
Might be a reason that 1k ohm resistor failed. The controller at U2 might be knacked.
Part 2 please
Cliffhanger!
Very good explain thanku sir
if you look on thermal the resistor you replaced will be heating up, the chip it feeds will likely be faulty, if you leave power on the card long enough the resistor. will fail again
Since the customer said I clean it and it worked for a day, I assume the GPU chip has disconecting. And when the card has a separator solder, it means it overheats while working. So that resistance burns out due to overheating. The best way to test it, put just a little heat on GPU, if the card works you have to re-ball it
i guess the missing vcore cause the no detect. done, no fix xD
btw: greetings from germany. thanks for sharing your experience! Keep going
You need PEX rail ... To power vcore ... Greetings from Portugal ...
When you replace a capacitor how do you know the value of that specific capacitor?
@@BilalKayaTR thank you for your reply. I’m asking in general if I found a faulty capacitor how I know which one to install.
@@BilalKayaTR if you know please let me know. Thanks
@@golamrabby5335 Either you know it from schematics, or you can take as good that, in the same productive process, capacitors with same color and same size should have same value. Taking this as a fact, you could measure (off circuit) another capacitor with same characteristics installed in the same board.
Last time I was this early I got 2 worms
Where is part 2?
Hi sir, Please provide me second part for that GPU.
But could you modd 4 more gigs of ram onto my 8-gig 1070?
Where is the part 2
I have a 3080 when I turn on my PC graphics card cpu Everythings is running but VGA light on motherboard stays on and no display i changed ram psu and everything but display
its an dead GPU for sure
you forgot to put on the eye glasses as you always do when powering a card
Part 2?
usually about now this will be a no fix as it might be the gpu, just a wild guess.
EN on PWM controller (up9511?) pin3
re connect the heat sink fan to the board
Love new videos.
what happend with this card??we waiting for part 2 :D
Thermalcam probobly will show a short in some of VCORE mosfett. Thats my guess. I hope that it doesn't show heat spot on GPU-DIE. It is possible because self degradation of thermal paste and bad heat spread. If is THAT case the GPU is completely dead 😢 Hope GPU is not fried and this graphics card can be resurrected!
We will see in the next video!
Hey guys, can anyone point me to a place where I can learn to diagnose GPUs issues ?
Like, there has to be some place that can teach me more ?
I think there is water in the gingle pin...
Awesome ✌🏻✌🏻✌🏻
My MSI GTX 1080's have been doing this as of late
Love from Pakistan 🇵🇰😍💕👍
New super hero name V-man
nice Vedio ☮️👍
When it comes to electronic repairs you can only try to fix the device
I love you boss 😍😍❤️❤️🇮🇳🇮🇳🇮🇳
part 2 plz
Your back ground it like you are in a spaceship
Turn on the thermal camera
Thermal camera is useless if you dont have a short.
it looks so easy when he is doing it xD
looks line GPU SHORTED ITS bit scary when find problem with GPU V CORE
My experience is when a resistor is blown something cause it to go
Nice