Update: Check part 2 at th-cam.com/video/HtIr7bxwQes/w-d-xo.html Viewers are suggesting that i inject voltage on mem and see what happens. I will do a part 2 today.
From the two cards I have done that to, usually all but one VRAM chips are shorted. Usually the 12v gets through to the VRAM chips and kills them. Replacing all the VRAM chips can fix it but the GPU can get damaged sometimes as well.
@@skullteria you took the words right out my mouth....this is exactly how i feel. I also bought equipment little by little. @northridgefix gives us the confidence to try fixing it ourselves.
You always ask what we think. I think that I have never seen a more honest man. I love that even when you may fail on a specific device for whatever the reason. You always show the video. That is great to see. Many youtubers only show their success stories. You sir are an inspiration! Also my goto happy place, because of your humor! Thank you!
GPU repairs seem to be a nightmare in general. I watch other channels that only do GPU repair and often it is very complicated and you have to have detailed knowledge of all the little components and even then it's not guaranteed fixable. Then there's cards that are almost always dead / unfixable like those older AMD cards with HBM memory. Or you get half-working cards that have artifacting and/or stability issues so you have to do extensive testing in software and so on to find out which memory module is faulty or whatever. Very annoying and time consuming.
They are ment to be unfixable as possible. No company making GPUs like Asus, MSI, Gigabyte, etc give any type of schematics or service manual. This limits repair to themselves or the few people that have built up experience with loads of dead GPUs which not many have that opportunity
No, that’s incorrect. Im a PC modder / watercooler. I remove heatsinks to place waterblocks on them all the time, and they do not have any direct attachments to any of the electronics of the card, meaning it can not signal that there is a waterblock on the card. It starts up regardless.
@@AJayMustang they still have to power on for the over heat protection to kick in. He said they dont power on, yes they do, they jjst over heat immediately after powering in
I wish I could send you a video card to fix. I have one video card and an old stereo receiver I need fixed,.,.. and maybe a motherboard. I see you have become so successful that you can't handle anymore. I'm very happy you are thriving. At least somebody is back home in California.
I think You can also track that line (inyect voltage) and see if there is another thing causing a short. I know that you guide yourself by your experience but sometimes things go diferent.
The lack of blown fuses is a great sign. It couldn’t have taken that much current without blowing the fuses. @actually hardcore overclocking is a great resource for gpu and mobo power delivery. And how stuff fails. Edit: also how would the mem short not show up on any of the 12v rails? Just curious but i thought it was powered of 12v.
When I was working with electronics development, if we had to rework something we would often use kapton tape to shield off components that we didn't want to heat up too much. Maybe you could use that for shielding stuff like your electrolytic capacitor there.
@@ДимитърАндонов-ъ7е Heatsinks cool by conducting heat into the air. A heatsink cannot make something cooler than the air around the heatsink. In this situation, hot air is being blown around the heatsink which will reduce its ability to cool. A heat shield will prevent the component from directly contacting the hot air. Being an insulating material is a good thing in this case.
Looks like GPU is shutting down to prevent from overheating. Those cards (3080 and over) reach 100+ celsius in a couple of seconds without any heat dissipation. 0 ohms might be caused by heat protection. In this kind of cards, always look for any way to avoid thermal shutting down, or in a couple of seconds any measure will have no sense ;)
0 Ohms because of heat protection makes no sense. Yes the card will heat and may shut down but still that 0 ohm should not be there that's not how the protection works secondly , there is nothing that will prevent it from Posting , that's completing the built in self test load the firmware etc, if it did that it wont beep. Card is probably dead better to get these things RMA , evenever there is short its highly likely it causing secondary failures such as the chip itself
I'd love to see a how-to on replacing dead memory chips on a gpu I've got a three cards with the same dead chips but I'm not yet comfortable enough with the process to try making 2 good cards out of 3 bad ones
Ever try heat stop putty ? I used it on hvac sweats close to metering device and wet bulb . 2,400 ° oxygen acetylene torch and heat never traveled past the putty . Just a thought when nothing else works . Maybe not compatible chemically . I do not know The Zen Master is you brother ,not AMD
It's any point on the shorted rail. Because from that point, it'll all go through the shorted component to ground. Then you use a thermal camera/atomiser/Isopropyl to find what heats up. So any point that's shorted (and being careful you don't inject it into ground by mistake, you can use capacitors to know what's positive and negative) will work.
He injects voltage on the shorted line. If the line is shorted and you inject voltage it will heat up what component is bad or it will heat up what ever component that has the lowest resistance to ground.
Look for the power cord :D I guess you need to figure out the short first, that means you need to figure out where the buck converters are first, if he had to explain everything in every video it will be boring for educated audience.
Hi, thanks for all your videos. What I'm not understanding is why the fuse doesn't blow if there is a short on the board? Also, how come not every competent on the shorted line (e.g. The 5v line) gets hot? You were measuring shorts on many components but only one component got hot, why? Thanks again
Fuses don't blow with shorts, they blow when there's too much current to prevent damage, which often occurs due to a short but a short does not always mean high enough current to break fuses in microelectronics. The short breaks the circuit flow and since there's no longer any resistance the failed component starts emitting current as heat from where the circuit fails, at higher voltages probably heat would spread to other components too but he injects very low voltages to contain heat emission to the failed component/that's how he pinpoints which component failed
I'm not very educated but my guess is fuse doesn't blow because the PSU is protecting itself, i.e. it's a 20A fuse, that's a lot and the PSU might be shutting down way earlier. Your 2nd question, is just ohm law, traces doesn't get hot as their resistant is very low, other components in series like fuses, coils are also very low resistance in dc, then non faulty components in parallel like caps and good ICs are high impedance, so much power gets to the shorted component only and heat is generated there first. Try a 1 ohm resistor with some wires (0 ohm) connected to a dc power supply, it will burn the resistor, not the wire
Hello sorry for my question which is not related to the subject, I have a big problem with the search for flash bios graphics card Are there any sites specializing in downloading flash bios? Thank you,,
IDK why most of the 3090 cards have this problem (Suddenly won't turn on) , as a 3090 owner i'm honestly scared , thank god the card has a 5 years warranty
You need to change thermal pads regularly with very good ones. And also good thermal paste. I recommend thermal grizzly minus pad 8 and hydronaut or kryonaut paste also arctic mx5 is also good
As per buildzoids analysis from actually hardcore overclocking channel, the vrms and controllers are really lacking for the rtx 3090, where he tweeted out the only reason you'd want to buy the 3090ti over a 3090 is to not have recycled vrms and controllers from 6th or 7th Gen Intel motherboards
voltage controller and protections are garbage, unless the retailer is changing something, most of them have it so the over current protection will never be triggered because it is set almost twice as much as the card is able to even pull
Hey Alex, you inspired me to buy a thermal camera, unfortunately I can`t buy from you since I get charged 58% from taxes in my country so, I had to find another way to gete it. it wiil be a uni-t.
Good job. Can i ask a question? I have a lil more simple card, it doesnt have another power cable, only slot feeds it. On the card all 3.3v power rails are short (this must be clue tho). Used power injection 3.3 v 2amp, it pulls 1.4 amp stable and nothing else getting hot but gpu, it is getting slightly hot thats all. There is no damage on the board, clean and doesnt turn on. 12 volt power rails are ok. Just confused about that all 3.3v power rails are not connected right? i decided that gpu is gone, am i wrong?
Of course, we do haha Here are a few: Use a separate 1000 Watt PSU with watt-meter check vbios for factroy defaults use heatsink and fan have the manufacturer drivers preinstalled lots things Do you know what avoinics mean? haha
Not like 10series nvidia and old amd..20 and 30 series card is very hot in split second.. u need to put temporary heatsink before turn it on..even with short on memory (0 0hm).. the voltage on vcore and memory should still present
Why does everyone think 30 series cards need a fan plugged in to turn on? I've water blocked tons and never have fan connectors on the card plugged into anything and they run fine ;). He's correct that the card itself is toast, it goes through it's power cycle all the way until it powers the memory & chip. Then the chip get's super hot, super fast, the card goes into protect mode and throws the beep code that the card isn't functional. At least we got the "and the short is gone" before he diagnosed it as a brick. lol
@@NorthridgeFix without it in front of me I can't really assume anything. But a 0ohm read on either vmem or vcore is probably either dead GPU or dead ram chip
@@Coniass " Silver thermal compounds may have a conductivity of 3 to 8 W/(m·K) or more, and consist of micronized silver particles suspended in a silicone/ceramic medium. However, metal-based thermal paste can be electrically conductive and capacitive; if some flows onto the circuits, it can lead to malfunction and damage. " en.wikipedia.org/wiki/Thermal_paste
I have a card that does not show any output, but I am access and flash the BIOS - does anyone know if the GPU might be dead or shorted? I thought if there is a short, I won't be able to access the vBIOS either.
When you measure the Ohms to see if there is a short, what are the OHMs telling you. I think you said 440 Ohms was good. Is it that more resistance is good? I'm just trying to follow your logic. A lower number would mean an open short to ground? It that what you are checking for? I' just want to understand the relationship of the ohms reading to the short. Thanks.
When reading ohms one probe is always on ground and measuring a positive voltage side should always be at least over 100 ohms and anything close to 0-1 ohm is basically a direct link to ground which is never good for a voltage side line.
It appears, to me, that the owner of this card's a PC enthusiast. Nothing wrong with being a PC enthusiast, IMO. So, I wonder if over-clocking killed/hastened its demise? I have a 3070 Ti, by MSI, so I hate to see a card, in the shop, manufactured by them. Anyway, I do hope that this card can be saved, but I think it might be a lost cause. However, when it comes to electronics, I surely wouldn't bank on anything I say, lol.
What about BIOS or overcooking, customers are in most cases lie . Maby he was tweaking a card and fried memory, once had that kinda customer with 2080ti was overcoocking GPU and flashed with some moded BiOs and bricked it stateing that was working fine before and sudenly next day it stopped working . No mentioned about mods he put on. .
@@jambikaus311 just reassemble the shroud, heatsink, and fans before u mail it back? i mean if RMA cant fix it i would expect a new card. even those void warranty screw stickers arent legal (unverified) if thats a worry.
@@ShadowSingularity88 put it back together and none the wiser? also these AIBs fucking know ppl are gonna water block their cards, especially a 3090. u cant void a card if u just open it for cleaning. The FTC also stated that warranty void stickers are illegal. i mean srslyif northridge cant fix it this guy actually NEEDS to RMA asap or nice paper weight
@@Yoblokit those stickers are illegal and meant as a scare tactic by shitty companies and we dont know for sure if that card had warranty void stickers to begin with. not all of the cards have them. RMA is not possible until rma people tell you you voided warranty with xyz but until then its still technically under warranty.
too bad you cant stock up on memory chips and gpu's for the best video cards that are worth repairing but i know that would cost more than a new video card so would not be worth your time
I watercool mine to lower the working temperature of the GPU and reduce internal system temperature. My system is whisper quiet and it doesn’t burn my leg when I’m gaming.
Update: Check part 2 at th-cam.com/video/HtIr7bxwQes/w-d-xo.html
Viewers are suggesting that i inject voltage on mem and see what happens. I will do a part 2 today.
Thanks i appreciate that.
From the two cards I have done that to, usually all but one VRAM chips are shorted. Usually the 12v gets through to the VRAM chips and kills them. Replacing all the VRAM chips can fix it but the GPU can get damaged sometimes as well.
@@skullteria you took the words right out my mouth....this is exactly how i feel. I also bought equipment little by little. @northridgefix gives us the confidence to try fixing it ourselves.
You always ask what we think. I think that I have never seen a more honest man. I love that even when you may fail on a specific device for whatever the reason. You always show the video. That is great to see. Many youtubers only show their success stories. You sir are an inspiration! Also my goto happy place, because of your humor! Thank you!
.. you don't know what he doesn't show...
Did you buy all the stuff he was selling?
GPU repairs seem to be a nightmare in general. I watch other channels that only do GPU repair and often it is very complicated and you have to have detailed knowledge of all the little components and even then it's not guaranteed fixable. Then there's cards that are almost always dead / unfixable like those older AMD cards with HBM memory. Or you get half-working cards that have artifacting and/or stability issues so you have to do extensive testing in software and so on to find out which memory module is faulty or whatever. Very annoying and time consuming.
Which channels are those?
@@kkgt6591 TechCemetery is my favorite.
@@ApocDevTeam
Techcemetry is an arrogant prick but he knows his stuff
They are ment to be unfixable as possible. No company making GPUs like Asus, MSI, Gigabyte, etc give any type of schematics or service manual. This limits repair to themselves or the few people that have built up experience with loads of dead GPUs which not many have that opportunity
@@zacharytaylor8523 Why do not they want to fixed voltages? To sell new ones?🤔👀
I think these newer cards only power on when the fan and the heat sink are attached. I had a similar issue. Great repair 👍👍
It won't power on if there is a dead short on the memory side. That pretty much killed it.
what do you know about these newer cards boy
No, that’s incorrect. Im a PC modder / watercooler. I remove heatsinks to place waterblocks on them all the time, and they do not have any direct attachments to any of the electronics of the card, meaning it can not signal that there is a waterblock on the card. It starts up regardless.
Yes 30 series overheat protection hits instantly without a heatsink on the card.
@@AJayMustang they still have to power on for the over heat protection to kick in. He said they dont power on, yes they do, they jjst over heat immediately after powering in
I wish I could send you a video card to fix. I have one video card and an old stereo receiver I need fixed,.,.. and maybe a motherboard. I see you have become so successful that you can't handle anymore. I'm very happy you are thriving. At least somebody is back home in California.
"and the short is gone" is the best part of the video, always :D .
I dont know how you make fixing electronics suspenseful, but you do. Good work!
Your the man. I used to be able to do this type of work. Now my eyes and hands are not friendly to me these days.
I think You can also track that line (inyect voltage) and see if there is another thing causing a short. I know that you guide yourself by your experience but sometimes things go diferent.
The lack of blown fuses is a great sign. It couldn’t have taken that much current without blowing the fuses. @actually hardcore overclocking is a great resource for gpu and mobo power delivery. And how stuff fails.
Edit: also how would the mem short not show up on any of the 12v rails? Just curious but i thought it was powered of 12v.
Unfortunately, it is very important to try, I wish you good luck ❤
This guy and big boss would be as asset to any doomsday bunker. I’ll bring the milling machine and the lathe.
When I was working with electronics development, if we had to rework something we would often use kapton tape to shield off components that we didn't want to heat up too much. Maybe you could use that for shielding stuff like your electrolytic capacitor there.
He used to use kapton but he hates it, he finds aluminium tape a notch above kapton and ofc heatsinks are even better.
A heatsink may conduct the heat in the hot air into the capacitor if it's not so large. Heat shielding is better
Not his first rodeo... but I like your attitude JOHN WICK. You are a good person.
I use copper tape.
@@ДимитърАндонов-ъ7е Heatsinks cool by conducting heat into the air. A heatsink cannot make something cooler than the air around the heatsink. In this situation, hot air is being blown around the heatsink which will reduce its ability to cool.
A heat shield will prevent the component from directly contacting the hot air. Being an insulating material is a good thing in this case.
They really gonna have to improve the buck converter Circuit to protect the gpu die from receiving a direct 12v , when a mosfet fail
@@laboratorioassembler BINGO! Planned obsolescence. ;)
@@laboratorioassembler you dont understand, its not about economy or sales,
Looks like GPU is shutting down to prevent from overheating. Those cards (3080 and over) reach 100+ celsius in a couple of seconds without any heat dissipation. 0 ohms might be caused by heat protection. In this kind of cards, always look for any way to avoid thermal shutting down, or in a couple of seconds any measure will have no sense ;)
0 Ohms because of heat protection makes no sense. Yes the card will heat and may shut down but still that 0 ohm should not be there that's not how the protection works secondly , there is nothing that will prevent it from Posting , that's completing the built in self test load the firmware etc, if it did that it wont beep. Card is probably dead better to get these things RMA , evenever there is short its highly likely it causing secondary failures such as the chip itself
Yeah is exactly what I was going to suggest
how the hell would a 0hms dead short be cause by heat protection? It wouldn't.
14:00 probably without heatsink the gpu core is frying
I'd love to see a how-to on replacing dead memory chips on a gpu
I've got a three cards with the same dead chips but I'm not yet comfortable enough with the process to try making 2 good cards out of 3 bad ones
Thank you bro , but where we can obtain such mosfets or collections of most used mosfets , appreciate your help
Put a cooler on the GPU and retest. Probably going into a thermal shutdown. My 3090 gets so insanely hot, its idles at over 100 watt power draw.
No he explains it at the end with the memory shorted (0 ohms).
A short on Vmem can hardly be caused by a warm GPU.
As Paris would say.. "That's Hot". Cheers TD
Obviously, one should try with Heatsink on it.
What's your queue time nowadays? (ie the time it arrives at your door to the time it leaves your door?)
There's also most nightmare GPU high end owner : Burnt VRAM GDDR6X cause too overheat. Which can't be RMA cause counted as "human error"
It's just thermally cycling. Put a small heat sink on it, and plug in some sort of fan into the GPU fan header just in case.
so soon tho as in a split second?
You are simply the best!! I really enjoy watching your educational program. I commend you for sharing your knowledge.
new 30 series cards does not start without fans you should try that
So you are saying that there is none 30 series with waterblock?🫠
So you are saying that there is none 30 series with waterblock?🫠
@@tomtebloss89 water block still have fan.
i could watch these videow whole day.
Can you check with a termo camera when device is plug in , from were the shortcut of memory is coming from ?
Ever try heat stop putty ? I used it on hvac sweats close to metering device and wet bulb . 2,400 ° oxygen acetylene torch and heat never traveled past the putty . Just a thought when nothing else works . Maybe not compatible chemically . I do not know
The Zen Master is you brother ,not AMD
Customer should send in complete card... Heatsink, fan etc.
Reject partial cards.
nice ! i was looking for a gpu repair
Sir, you always do a great job.
How do you know where to inject voltage? On any electrical board not only gpus....can you please make a video about that?
It's any point on the shorted rail. Because from that point, it'll all go through the shorted component to ground. Then you use a thermal camera/atomiser/Isopropyl to find what heats up. So any point that's shorted (and being careful you don't inject it into ground by mistake, you can use capacitors to know what's positive and negative) will work.
He injects voltage on the shorted line. If the line is shorted and you inject voltage it will heat up what component is bad or it will heat up what ever component that has the lowest resistance to ground.
Look for the power cord :D I guess you need to figure out the short first, that means you need to figure out where the buck converters are first, if he had to explain everything in every video it will be boring for educated audience.
I would have check if the bios was corrupted due to the power malfunction, and re-wrote the bios. You would need an i-GPU enabled motherboard.
Good idea chris, I was thinkin' the same
IS NOT A BIOS PROBLEM. It is shorted to ground in the memorys phase, missing voltage then graphic is not power on
Hi, thanks for all your videos. What I'm not understanding is why the fuse doesn't blow if there is a short on the board?
Also, how come not every competent on the shorted line (e.g. The 5v line) gets hot? You were measuring shorts on many components but only one component got hot, why?
Thanks again
Fuses don't blow with shorts, they blow when there's too much current to prevent damage, which often occurs due to a short but a short does not always mean high enough current to break fuses in microelectronics.
The short breaks the circuit flow and since there's no longer any resistance the failed component starts emitting current as heat from where the circuit fails, at higher voltages probably heat would spread to other components too but he injects very low voltages to contain heat emission to the failed component/that's how he pinpoints which component failed
I'm not very educated but my guess is fuse doesn't blow because the PSU is protecting itself, i.e. it's a 20A fuse, that's a lot and the PSU might be shutting down way earlier. Your 2nd question, is just ohm law, traces doesn't get hot as their resistant is very low, other components in series like fuses, coils are also very low resistance in dc, then non faulty components in parallel like caps and good ICs are high impedance, so much power gets to the shorted component only and heat is generated there first. Try a 1 ohm resistor with some wires (0 ohm) connected to a dc power supply, it will burn the resistor, not the wire
@@glmnet thanks for info, i liked this.
Shame about the card there very expensive aswell. Great video as always though
Very good video. And your explainings are perfect, like a very good teacher. Thank you for that video. Maybe you have more luck next time. Greets
Try to add cooling an retry. New cards have a power safe shutdown if no cooling is detected
Hello sorry for my question which is not related to the subject, I have a big problem with the search for flash bios graphics card Are there any sites specializing in downloading flash bios? Thank you,,
IDK why most of the 3090 cards have this problem (Suddenly won't turn on) , as a 3090 owner i'm honestly scared , thank god the card has a 5 years warranty
You need to change thermal pads regularly with very good ones. And also good thermal paste. I recommend thermal grizzly minus pad 8 and hydronaut or kryonaut paste also arctic mx5 is also good
As per buildzoids analysis from actually hardcore overclocking channel, the vrms and controllers are really lacking for the rtx 3090, where he tweeted out the only reason you'd want to buy the 3090ti over a 3090 is to not have recycled vrms and controllers from 6th or 7th Gen Intel motherboards
A few of the board partners arent actually honoring their warrenty hence why Alex has so many Gpus in for repair especially these 3090 cards.
@@heavygunt10 That's because MSI is trash, don't know why people keep buying them.
voltage controller and protections are garbage, unless the retailer is changing something, most of them have it so the over current protection will never be triggered because it is set almost twice as much as the card is able to even pull
Hey Alex, you inspired me to buy a thermal camera, unfortunately I can`t buy from you since I get charged 58% from taxes in my country so, I had to find another way to gete it. it wiil be a uni-t.
great job !!!
Good job. Can i ask a question? I have a lil more simple card, it doesnt have another power cable, only slot feeds it. On the card all 3.3v power rails are short (this must be clue tho). Used power injection 3.3 v 2amp, it pulls 1.4 amp stable and nothing else getting hot but gpu, it is getting slightly hot thats all. There is no damage on the board, clean and doesnt turn on. 12 volt power rails are ok. Just confused about that all 3.3v power rails are not connected right? i decided that gpu is gone, am i wrong?
Hi Alex, try with the heatsink on.
Do you think we'll ever get socketed GPUs? These cards are essentially motherboards, and those don't come with soldered on CPUs...
I think bios chips and firmware should be socketed.
13:59 🤣🤣🤣🤣🤣🤣🤣🤣..
That's me
I’m sure all the scientists will have many suggestions for you Alex…
Of course, we do haha
Here are a few:
Use a separate 1000 Watt PSU with watt-meter
check vbios for factroy defaults
use heatsink and fan
have the manufacturer drivers preinstalled
lots things
Do you know what avoinics mean? haha
@@igrewold Oh well, at least we know you haven't got a clue
@@trollmeister69
Man you know where to get a cheap RX 580 from?
Wanna try it on linux
They got some on aliexpress but maybe fakes
Put a heat sink on the gpu it’s shutting off power as a safety mechanism as without the heat sink it’s instantly overheating
Some old boards BIOS doesn't contain microcode to detect new cards say for 8GB and bigger so try a newer board
I don't understand 15:15 why low or no resistance in memory means dead GPU ? did the die itself went kaput ?
heartbreaking after all that work on such an expensive card
could that short cause 12v to pass to the gpu or vram?
Not like 10series nvidia and old amd..20 and 30 series card is very hot in split second.. u need to put temporary heatsink before turn it on..even with short on memory (0 0hm).. the voltage on vcore and memory should still present
Why does everyone think 30 series cards need a fan plugged in to turn on? I've water blocked tons and never have fan connectors on the card plugged into anything and they run fine ;). He's correct that the card itself is toast, it goes through it's power cycle all the way until it powers the memory & chip. Then the chip get's super hot, super fast, the card goes into protect mode and throws the beep code that the card isn't functional. At least we got the "and the short is gone" before he diagnosed it as a brick. lol
How do you inject volts into the card?
you did a great job but the catd decided to sleep....love brom Bangladesh 🇧🇩
Try to connect the fan, some boards will not turning on if the fan is not connected.
Which Fluke multimeter model are you using?
think he uses fluke 87v
Would do a test with a fan plugged in. May have a failsafe? Same with it getting hot then turning off, add a heatsink and try again
Just use common sense. A 0 ohm reading on mem while power is off has nothing to do with whether the fan is plugged in or not.
Are you sure that 0ohm coil you measured is Vmem? Probably it is a V core.
Check the schematic
@@NorthridgeFix without it in front of me I can't really assume anything. But a 0ohm read on either vmem or vcore is probably either dead GPU or dead ram chip
@@NorthridgeFix That's a douchey comment for someone who wasn't being rude, I think having 257k subscribers is getting to your head.
The customer should have sent the whole card, this may need the fans connected or CPU is overheating and shutting down.
There is a dead short on the memory, the heat is coming from the current rushing in, no gpu will get to 100 C in 0.5s at start.
The customer's cooler setup is a complicated water radiator; traditional fan & heatsink omitted.
good job mate.
good luck with the card
Keep up the good work
4:51
Is thermal paste conductive if hot/warm or cool ?
No
@@Coniass
" Silver thermal compounds may have a conductivity of 3 to 8 W/(m·K) or more, and consist of micronized silver particles suspended in a silicone/ceramic medium. However, metal-based thermal paste can be electrically conductive and capacitive; if some flows onto the circuits, it can lead to malfunction and damage. "
en.wikipedia.org/wiki/Thermal_paste
I have a card that does not show any output, but I am access and flash the BIOS - does anyone know if the GPU might be dead or shorted?
I thought if there is a short, I won't be able to access the vBIOS either.
When you measure the Ohms to see if there is a short, what are the OHMs telling you. I think you said 440 Ohms was good. Is it that more resistance is good? I'm just trying to follow your logic. A lower number would mean an open short to ground? It that what you are checking for? I' just want to understand the relationship of the ohms reading to the short. Thanks.
When reading ohms one probe is always on ground and measuring a positive voltage side should always be at least over 100 ohms and anything close to 0-1 ohm is basically a direct link to ground which is never good for a voltage side line.
Very good job 👏 👍 👌
It appears, to me, that the owner of this card's a PC enthusiast. Nothing wrong with being a PC enthusiast, IMO. So, I wonder if over-clocking killed/hastened its demise? I have a 3070 Ti, by MSI, so I hate to see a card, in the shop, manufactured by them. Anyway, I do hope that this card can be saved, but I think it might be a lost cause. However, when it comes to electronics, I surely wouldn't bank on anything I say, lol.
What about BIOS or overcooking, customers are in most cases lie . Maby he was tweaking a card and fried memory, once had that kinda customer with 2080ti was overcoocking GPU and flashed with some moded BiOs and bricked it stateing that was working fine before and sudenly next day it stopped working . No mentioned about mods he put on. .
Is it ok to power the GPU without the heatsink on?
NO
why the 3090 owner not go thru rma if still under warranty? is it cuz northridgefix is faster, typically?
The card is factory air cooled and the person voided his warranty because it was replaced with water cooling.
@@jambikaus311 just reassemble the shroud, heatsink, and fans before u mail it back? i mean if RMA cant fix it i would expect a new card. even those void warranty screw stickers arent legal (unverified) if thats a worry.
Rma not possible. Voided the warranty sticker for the waterblock.
@@ShadowSingularity88 put it back together and none the wiser? also these AIBs fucking know ppl are gonna water block their cards, especially a 3090. u cant void a card if u just open it for cleaning. The FTC also stated that warranty void stickers are illegal. i mean srslyif northridge cant fix it this guy actually NEEDS to RMA asap or nice paper weight
@@Yoblokit those stickers are illegal and meant as a scare tactic by shitty companies and we dont know for sure if that card had warranty void stickers to begin with. not all of the cards have them. RMA is not possible until rma people tell you you voided warranty with xyz but until then its still technically under warranty.
Amazing clip 👏🏼😁
Alex, where is your safety glasses? Just kidding 🙂
connect the fan and heatsink and try again.
giant gpu and yet if any supporting transistor dies, whole gpu dies too. I am sure if they turn off couple cores on factory it would work.
When GPU is shorted, then RIP 3090 Card!
Greetings and respect to you my dear brother, I am Abbas, your follower from the country of Iraq
Surprised the guy didn’t just send it in for RMA
Sometimes, one gets away. Burnt out GPU?
nice Vedio ☮️👍
I know this old but I thought I saw a memory chip was missing when u scanned through
Check Vcc pin bug converter
Awesome ✌🏻✌🏻✌🏻
غير المكثفات الكيميائية على خط الميمورى اكيد شورت المكثفات عن تجربة
First like first comment amazing video as always keep it up brother your the best 👍👍
❤️
put on heat sync then try
too bad you cant stock up on memory chips and gpu's for the best video cards that are worth repairing but i know that would cost more than a new video card so would not be worth your time
goodness
Nice 👍🏾
GPU reballing ?
resistance is futile!
👍
You cannot save them all,
Due to thermal it will happen...... Check once with heatsink
Why the hell would he send it to you the thing should be under warranty, that is a $2k video card? It is over heating!!!
🖤
Rip memory card
I hope you finger didn't get too burned😀
yayy
The only reason people water cool their graphics card is to over clock it... this can cause components to fail no wonder it's broken
100% User got greedy, wasnt happy with a stock 3090, he probably fried it. RIP.
I watercooled my 6900xt & I don't overclock. I did it to keep the temps & noise down & because it looks cool.
Plus cooler temps = longer life
I watercool mine to lower the working temperature of the GPU and reduce internal system temperature.
My system is whisper quiet and it doesn’t burn my leg when I’m gaming.
Not at all true, it's mostly for better noise & temps.
Overclocking is almost pointless with modern silicon.
Came in for no power.