Another tip for those having connection issues after reinstalling the core.. On these older misused cards I have found that applying a 30 gram copper weight on top of the core with thermal paste while installing it will help with the balls not attaching this is usually do to the board or core not being perfectly square or flat.. I use 40g with 3080 or larger cores.. The balls always look perfect under the microscope after using this procedure. Keep up the good work..
If it is a Gigabyte 2080 Super, I don't think you solved the issue (1 other comment who got the same card re balled from you say the same). The problem is random and the card can works perfect for 2 hours or then crash after 1mn. It is I believe because of Sic788A DrMos (if this one has it, I would be interested to know), they have a kind of defect when they go bad, send higher voltage spikes instead of being nicely shorted, very specific to this model. You can try to check that with an oscilloscope under load, I have also proof on video if you want so we can compare. The solution for this one is to change all DrMos, it is mandatory - all of them to be 100% sure the problem is gone, and test it for multiple hours at least.
Hi, I just wanted to thank you for your comment. I have 2 Gigabyte 2080 Supers with this issue, and I have scoured the internet looking for the cause. My cards both have the SiC788A, and your description lines up perfectly with the symptoms of my cards. I will note that the PCB around the DrMos es is noticeably darker. I may be completely wrong about this, but I seem to remember that the card is forced to run all phases all the time, which produces a lot of excess heat. I think this may be a contributing factor, but I will have to see. Anyway, I will post an update once I do the repair. Also, @northwestrepair, thank you for showing an uncut and honest video. I really enjoy them and I've learned so much from you!
You seem to be the best repair person I have witnessed on the internet. Of course Tony, Alex at Northridgefix makes all his boards "Better than factory", so how can one do better. Your fix seems better than Krisfix Churmany as you don't use lead free solder like he does. Watching you repair pads under memory or primary Graphics chips spells out to me no one is as good.
That’s the same heatsink I use when I start my bench builds. I hot glued a fan on top. Has an awesome copper heatsink. Works beautifully on every cpu, even a 13900k.
That recovery function is VERY usefull if you just know what you are doing. You can fix almost anything from the command line and you can even go around Windows passwords etc.
I’m not easily impressed, but you’ve got my trust. If my card (2019 EVGA 2080Ti) goes out, you’re getting it. Thank you for this excellent video. Solid idea making your videos your resume!!!
Most people pack things rather inadequately... taking the GPU out of the original box, using only a few sheets of bubble wrap as "padding," sometimes the component is bouncing around inside the box they ship it in which really grinds my fears when I see it. Or they put it in a dang bubble envelope.... The best thing to do for a GPU is to double box it. So in the original packaging put that inside a larger box with 1 inch shock absorbent padding covering all sides. That's all there really is to it. - 13 years eBay experience and I have packaged more heavy and fragile glass, porcelain, and electronics than I could ever remember.
In a static bag wrapped in bubble wrap and packed tightly with newspaper where you can barely tape the box shut will survive the ramp apes. It absolutely can't shift.
What do you preheat the board to and what temp do you set the hot air gun to and what temp do you expect the chip to release at? Thanks for the video. The reball is not boring!!!!!
We have been discussing this exact problem on discord in the nvidia general channel, it is a known probkem with this board, the drmoses fail in a strange way. And I haven't found a good way to find tha failing Drmos, but at least proved, that it has nothingt do with the core, the core BGA or the vrams, I even transfered a core from one of these to an other (better) card. You will see that the card behaves strangely, hard coil whine and you can measure the coil inputs on the back side on the RC snobber networks capacitor leg and 1 or 2 phase will look strange and those phases will pull really hard on the psu too, but the current steering fets and the coil for it drops some voltage (from 12V to 8V) and reduces the current a bit.
Also these 2080s cards use Samsung chips because it has HC16 chips, not HC14 ones. I haven't seen super cards with bad Micron anyways, it was discovered before the super cards really began being produced.
Hello, can i have link to the discord server ? I have 1 galax gtx 1070ti have blown vrm phase. After remove the blown phase it's still short on 12v rail
We need to push you forward so more people can see you work and you well deserve it. I try to like and comment on each of your videos. We desperately need someone with your level of skills and price range for repairs here in EU. Btw, Linus would only watch you if you had more than 1m subs. He's kind of an ass to his followers. Especially in person.
Amazing video, I love these longer vids, stupid question, when you are reflowing the core what is the purpose of the wire you place on the card when it's in the vice? Sorry if asked before I'm just curious 😅
I have this model and a weird problem. Under light load you can hear inductors oscillating and a second later card crashes. The interesting part is, rebooting resolves the problem until I reboot again. I dont think it is temperature related. Is that something common, never saw a fault like that.
Your cat is looking for the mouse, that's why it's attacking usb 😂 my cats exactly the same, meows early morning, especially during night. I have to kick outta my room, hahah 😅
Experience. As he said in the beginning of the video, he runs into this problem often and most of the time, the solution is the same. When troublshooting things like this, it's always best to try the quickest/cheapest solutions first. If he had assumed the core was bad and replaced it right away, that could've been very time consuming and a waste of money when something quicker and cheaper could've worked. If reballing the core hadn't worked, then he might've moved on to the more time consuming task of testing the core, the components, and so on.
got a question for Northwest or someone who knows about electronics: I took apart my gpu to change the pads and I noticed that some of the inductors were slightly cracked on top. I read that they should be molded inductors and its the coil inside that is important. gpu was running fine. (6900xt)
My GPU doing well during FurMark stress test, however crash on any benchmark preset and on games. What can be issue? tried change other GPU same, tried different slots same, changed thermal paste for gpu and cpu same. The only thing what comes in mind maybe PSU?
where can i check correct thermal pad specification for my gpu model wise and your discord link is saying invalid? do i need to use email to contact you?
My issue is pc crash while gaming after 5 or 10 minutes.I tested it on furmark even the temperature goes to 81 degrees.eventhen pc not crash.What is the problem.
Did you hear anything about pumpout effect? Or I imagine it's the customer problem who face it after a few months of usage(then the hotspot temperature 25-30° warmer than core). But who cares(not me, I'm just curious) cuz I'm far away from you Btw: do you using US cleaner? If not, youre experienced in diy stuff, maybe that's exiting project. Imagine: ultrasonic cleaner wit rgb, blackjack and b****
@@LiveType In my county I have access to shinetsu, denko and dow corning thermalpastes. All of these are way cheaper than anything I can find in usual computer stores, also all of them better in all terms. About 17usd for 30 grams. tpcm materials little bit pricier but not much. I prefer to use best materials I can find
I actually sent my card, Gigabyte 2080 Super, to NWR and reballing didn't help. The issue is believed to be the temp sensor within the chip which was deemed unfixable.
Again a epic and entertaining repair and that you show us even when you encounter some errors you don’t edit that out! Thank you Tony. 👍 😊 🍻
My pleasure!
great content@@northwestrepair
Another tip for those having connection issues after reinstalling the core.. On these older misused cards I have found that applying a 30 gram copper weight on top of the core with thermal paste while installing it will help with the balls not attaching this is usually do to the board or core not being perfectly square or flat.. I use 40g with 3080 or larger cores.. The balls always look perfect under the microscope after using this procedure. Keep up the good work..
This is roughly the weight of 5-6 quarters..
30 g or 3, cause 30 is insane
30, 3 gram won't do shit@@trouble_maker023
Always a fun riddle or puzzle. These repairs are very interesting to come along with. Thanks for bringing us!
man i wish i could work with you and learn these stuff... you sound like a chill teacher
3am gang, can i send my bed for repair, the pillow needs a reball and the butt has loose solder joints ;(
Bro 😂😂😂
Set your multi meter to ZZZZZZZZZZZ
You might get away with underclocking. Put the clock under the bed so you don't know what time it is:)
That's what I told my wife 😅😅😅😅
If it is a Gigabyte 2080 Super, I don't think you solved the issue (1 other comment who got the same card re balled from you say the same).
The problem is random and the card can works perfect for 2 hours or then crash after 1mn.
It is I believe because of Sic788A DrMos (if this one has it, I would be interested to know), they have a kind of defect when they go bad, send higher voltage spikes instead of being nicely shorted, very specific to this model.
You can try to check that with an oscilloscope under load, I have also proof on video if you want so we can compare.
The solution for this one is to change all DrMos, it is mandatory - all of them to be 100% sure the problem is gone, and test it for multiple hours at least.
Noted for sure
Hi, I just wanted to thank you for your comment. I have 2 Gigabyte 2080 Supers with this issue, and I have scoured the internet looking for the cause. My cards both have the SiC788A, and your description lines up perfectly with the symptoms of my cards. I will note that the PCB around the DrMos es is noticeably darker. I may be completely wrong about this, but I seem to remember that the card is forced to run all phases all the time, which produces a lot of excess heat. I think this may be a contributing factor, but I will have to see. Anyway, I will post an update once I do the repair. Also, @northwestrepair, thank you for showing an uncut and honest video. I really enjoy them and I've learned so much from you!
@@mronion6417 Any update? I have the same card, the same DrMos and the same problem :(
This man's humor hits me every time
perfect
good work
i love watch your videos to learn
Good work as always. Much love from Turkiye.
🦃 I love turkey 😂
@@northwestrepairyummmmm
You seem to be the best repair person I have witnessed on the internet. Of course Tony, Alex at Northridgefix makes all his boards "Better than factory", so how can one do better. Your fix seems better than Krisfix Churmany as you don't use lead free solder like he does. Watching you repair pads under memory or primary Graphics chips spells out to me no one is as good.
Love your repair videos! keep up the good work!
LOVE the reball music!
That’s the same heatsink I use when I start my bench builds. I hot glued a fan on top. Has an awesome copper heatsink. Works beautifully on every cpu, even a 13900k.
You, sir, are inspiring me a lot.
That recovery function is VERY usefull if you just know what you are doing. You can fix almost anything from the command line and you can even go around Windows passwords etc.
I’m not easily impressed, but you’ve got my trust. If my card (2019 EVGA 2080Ti) goes out, you’re getting it. Thank you for this excellent video. Solid idea making your videos your resume!!!
I really appreciate the work you do to film the re-ball it seems very awkward.
You must be expert in packing and shipping GPUs. It'd be interesting to see how you box and protect them for shipping after so much experience!
Most people pack things rather inadequately... taking the GPU out of the original box, using only a few sheets of bubble wrap as "padding," sometimes the component is bouncing around inside the box they ship it in which really grinds my fears when I see it. Or they put it in a dang bubble envelope....
The best thing to do for a GPU is to double box it. So in the original packaging put that inside a larger box with 1 inch shock absorbent padding covering all sides. That's all there really is to it.
- 13 years eBay experience and I have packaged more heavy and fragile glass, porcelain, and electronics than I could ever remember.
In a static bag wrapped in bubble wrap and packed tightly with newspaper where you can barely tape the box shut will survive the ramp apes. It absolutely can't shift.
What do you preheat the board to and what temp do you set the hot air gun to and what temp do you expect the chip to release at? Thanks for the video. The reball is not boring!!!!!
We have been discussing this exact problem on discord in the nvidia general channel, it is a known probkem with this board, the drmoses fail in a strange way. And I haven't found a good way to find tha failing Drmos, but at least proved, that it has nothingt do with the core, the core BGA or the vrams, I even transfered a core from one of these to an other (better) card. You will see that the card behaves strangely, hard coil whine and you can measure the coil inputs on the back side on the RC snobber networks capacitor leg and 1 or 2 phase will look strange and those phases will pull really hard on the psu too, but the current steering fets and the coil for it drops some voltage (from 12V to 8V) and reduces the current a bit.
Also these 2080s cards use Samsung chips because it has HC16 chips, not HC14 ones. I haven't seen super cards with bad Micron anyways, it was discovered before the super cards really began being produced.
Hello, can i have link to the discord server ? I have 1 galax gtx 1070ti have blown vrm phase. After remove the blown phase it's still short on 12v rail
You are wounded, such a warrior.
Very impressed!
We need to push you forward so more people can see you work and you well deserve it. I try to like and comment on each of your videos. We desperately need someone with your level of skills and price range for repairs here in EU. Btw, Linus would only watch you if you had more than 1m subs. He's kind of an ass to his followers. Especially in person.
Amazing video, I love these longer vids, stupid question, when you are reflowing the core what is the purpose of the wire you place on the card when it's in the vice? Sorry if asked before I'm just curious 😅
temperature sensor...
Very informative great vid
Awesome video
wOWOWOWO
Good job, old man!
Great video! What is the IR camera you use?
in my point of view you are not an idiot old man , you are a genius :) love from sri lanka
Love the honky tonk piano
Nice job❤
which gpu manufacturer you think uses best quality PCB and components ?
thanks
What? No Alex face? I feel so deprived (or is that depraved?) I can't tell anymore.
Great.
I have this model and a weird problem. Under light load you can hear inductors oscillating and a second later card crashes. The interesting part is, rebooting resolves the problem until I reboot again. I dont think it is temperature related.
Is that something common, never saw a fault like that.
It looks like you need a better reflow equipment, with temp profiles and hooke.. fans for cooling.
Very interesting job
You need a designated droid that does nothing but reballs.
Your cat is looking for the mouse, that's why it's attacking usb 😂 my cats exactly the same, meows early morning, especially during night. I have to kick outta my room, hahah 😅
You were doing probes when it was in the heater clamps, so the screw down bits could have caused shorts.
this was hilarious hhh the poor cat inhaled so much led it has a personality disorder lol
amazing fix.
I just got a gigabyte 2080 super off Facebook last week and I am having this same issue! How much would a repair like this run me?!
Have you had any experience in removing and replacing a socket on a mothrboard?
AS you were going along, how did you know it was not the core that went bad or was bad as you kept going? Thanks for the video.
Experience. As he said in the beginning of the video, he runs into this problem often and most of the time, the solution is the same. When troublshooting things like this, it's always best to try the quickest/cheapest solutions first. If he had assumed the core was bad and replaced it right away, that could've been very time consuming and a waste of money when something quicker and cheaper could've worked. If reballing the core hadn't worked, then he might've moved on to the more time consuming task of testing the core, the components, and so on.
Ask yourself, what would kill a core?
got a question for Northwest or someone who knows about electronics:
I took apart my gpu to change the pads and I noticed that some of the inductors were slightly cracked on top. I read that they should be molded inductors and its the coil inside that is important. gpu was running fine. (6900xt)
What model and did they have thermal pads connecting them to the cooler?
@@The_Man_In_Red 6900xt liquid devil and no thermal pads came on them, but I did put some on when I put the block back on.
pc crash and no signal with buzz sound when playing game but it pass all benchmark. Any idea what the problem
good work
Maybe you can reball your cat and retest to check if the cats mosfets and memory is reading as cat.
My GPU doing well during FurMark stress test, however crash on any benchmark preset and on games. What can be issue? tried change other GPU same, tried different slots same, changed thermal paste for gpu and cpu same. The only thing what comes in mind maybe PSU?
hi what tools you using do you got a video of the setup?
where can i check correct thermal pad specification for my gpu model wise and your discord link is saying invalid? do i need to use email to contact you?
A penguin! LMAO. Definitely get em snipped 😂
What is that stuff you are applying to the slot connection?
Lift the core 🤣 your so pro !!
i can watch this for hours
when you heat up the board with the preheater I see, that you put a thermocouple on top of the PCB. What temperature do you target ?
Can you turn sound lil' bit up? 1.5-2 times :)
Ahhhhh gigabyte, there’s your problem right there Lolol
Yeah I agree with the windows start message like that. It’s absolutely useless imo
rep Solid caps may can make it better
My issue is pc crash while gaming after 5 or 10 minutes.I tested it on furmark even the temperature goes to 81 degrees.eventhen pc not crash.What is the problem.
Perfect diagnostic.
You need a reball
What do you think might have gone wrong during the first reball attempt?
The sensor was in the wrong place showing higher temperatures.
What's the little wire for when you are re-balling a core??
I believe it is a temperature sensor indicating the board temperature. He has a complete video on his modifications to the board preheater.
Is your cat's name Big Boss 🤷♂️🤷♂️🤣🤣
LTT background is sales, yours is repair and you know, knowing what you're talking about :P
lol im i the only one who like the old intro more..this one is cool but the old one is special
Did you hear anything about pumpout effect? Or I imagine it's the customer problem who face it after a few months of usage(then the hotspot temperature 25-30° warmer than core). But who cares(not me, I'm just curious) cuz I'm far away from you
Btw: do you using US cleaner? If not, youre experienced in diy stuff, maybe that's exiting project. Imagine: ultrasonic cleaner wit rgb, blackjack and b****
Mx5 is one of them I imagine
@@LiveType have you tried Thermalright TFX? Any experience with that? It's decently thick from what I can tell.
@@LiveType In my county I have access to shinetsu, denko and dow corning thermalpastes. All of these are way cheaper than anything I can find in usual computer stores, also all of them better in all terms. About 17usd for 30 grams. tpcm materials little bit pricier but not much. I prefer to use best materials I can find
@@LiveType Running honeywell 7950 on my CPU (AM5 7700x) at the moment. Curious to try Roman's new Kryosheet. He talks it up.
I have the same issue on a 2080 super I acquired. What's the cost to reball the cpu?
I actually sent my card, Gigabyte 2080 Super, to NWR and reballing didn't help. The issue is believed to be the temp sensor within the chip which was deemed unfixable.
My cat plays fetch like a dog.
Not bad for a passive cooler
Got cats that come to a whistle and play fetch
IT'S REBALL TIME!
Disable Windows AutoBootRepair:
bcdedit /set {current} recoveryenabled no
lucky that my 2080super has samsung memory too, but most of my cards have had samsung
not sure how lucky that is
most cards have micron. At least for nvidia
I think the reason why Nvidia released the Super was to distinct it from the defective Micron chips.
i wish i can send my GPU to you for repair :-(
I have a hard drive out of a modded Xbox with a lot of emulators on it does anybody know how or if I can hook it up to my PS4 pro
Micron? Please remind us.
Love the LTT comment. That dude fell far. Love your channel brother.
He fell pretty hard, but I think they're actually taking steps to remedy the issues they had. Time will tell of course.
@@ArtisChronicles Got so big he lost control. Full corp model will change everything. Interesting drama to follow.
Guys, what for is the round wire that touches the core at 9:25 ?
It is a thermocouple that is used to indicate the board temperature. He has a complete video on his modifications to the preheating plate.
@@theoldbigmoose thanks man. watching more videos I learned that too. Tony mentioned that in one of them .
Yes reball!! I like this process
Edit: way too fast .. 😂😢
I saw that flux at the beginning of the video - and u want to not seen it? hmmm
this is exactly what my 3090 does
You can disable that try fix thing. hehe
Its 6pm
Is your finger ok?👉
ltt dizz ? :( :)
Video is too long 15 min max for me
go away, meow my a55 :)))
7:05 🔥 🔥 🔥 🔥 😂 Outdated like LTT! 😂😂😂😂
.
Every time you turned it on with the scissors/tweezers underneath i was cringing.
comment
Linus…………🤮. The worlds worst computer tech.