As I said at the beginning of this video, I'm no graphics card repair guru. I have no experience trying to fix them so this was just an experiment. If you're looking for good advice on how to repair graphics cards be sure to read the comments on this video as there is great advice from smart people.
I had that exact same 1080 ftw. 2 of em actually. both had similar problems because evga cheaped out and didn't add thermal pads to the vrms from the factory.
1. Never, ever allow GPUs to boot past BIOS without a cooler mounted. 2. That 3060 with the green artifacts is usually an indicator that a RAM-chip is either bad or has lost connection with the PCB. You could try a reflow of the RAM chips.
2080 Tis come with only 11 of the 12 memory channels populated from the factory, also, missing phases in the various VRMs is a normal strategy for the OEM to save money while still using the reference board design.
thats what I was thinking. Isn't each vram chip usually 1GB, so maybe that was a 11GB model but the board could also be used for a 12GB model and thats why 1 of the vram chips were missing? I also was thinking it came that way from the factory because the board looked way too clean, it didn't look like anyone went at it with flux and a soldering iron or heat gun since usually there would be some flux residue on the card afterwards
Yep, exactly what I was about to say. That EVGA 2080 ti xc ultra isn't missing any components. Also, one possible reason for the stacked thermal pads would be that the card was on a waterblock, got defective, and the just swapped the stock cooler back on to sell without caring about it being done right.
Yeah I'm bothered by him not researching the pcb or breakdowns out there. Love his videos but with today's resources this is something he could have researched on the fly.
Yeah, same with some other nVidia top-end GPUs below the Titans, like the GTX 780, same GPU (GK110) but partially deactivated and half of the VRAM, just 3 GB and these are only on the GPU side, the Titan has another 3 GB on the other PCB side; they simply share the reference PCB so the same watercooler fits both cards. And also the 980 Ti and Titan X (both GM200 but Titan got 12 instead of 6 gigs VRAM), 1080 Ti and Titan X (Pascal)/Titan Xp (all GP102 but the 1080 Ti got 11 and the titans got 12 gigs).
Came here to leave the same comment. OEMs frequently make the same base pcb for a range of cards and populate based on the version to save money on the manufacturing and engineering side of things. Other times they over engineer in design and then find they don’t need all the extra phases or memory modules to hit their prospective benchmark. “Missing” components is a common sight.
1) Don't ever just install and attempt to run a GPU that you know has an issue or is likely to have one. 2) Check for shorts on the power input rails aswell as on the voltage rails the card itself creates (5V, 1.8V, PEX, vCore, vMEM, VDDCI (on AMD GPUs), MSVDD (on all GA102 GPUs except for the 3090Ti)). Each power input is it's own 12V rail, so just checking one will not suffice. vCore, vMEM, VDDCI, MSVDD and PEX have rather low resistances, so check them using resistance mode. DO NOT use continuity mode, or you might just lead yourself down the wrong path. (Getting a close to 0 Ohm reading on vCore on newer GPUs is expected since the GPU silicon has a very low internal resistance.) 3) If you have a short, don't ever just inject voltage, especially if a power input is shorted. First, check if any power can make it's way from that power input directly to the GPU or memory using your multimeter. If there is a direct connection and you just go to inject voltage, you might just be sending whatever voltage you're injecting directly into your GPU or memory, potentially frying them. 4) If you have a short, inspect the componets that are on that shorted voltage rail closely. Don't waste your time looking at componets that aren't on that rail (they can't cause the short). Look for things such as blown capacitors, mosfets, powerstages or a bubbled-up top PCB layer. Look at the solder on mosfets and powerstages, see if there's some small solder blobs or balls around or if the joints look bad and not properly flowed. Those are indications that something got quite warm. If you have a short on PEX, your core is most likely dead. AMD Vega cards a notorious for having defective cores. 5) If you don't have a short anywhere, run the card. If you have artifacting, it's very likely one or more defective memory banks, mayhaps even across multiple memory channels. For nVIDIA GPUs from Fermi to Turing, use nVIDIA MODS MATS. Run it, and it'll tell you in a plain textfile which memory bank is spitting out errors. You can probably do the same on AMD GPUs and nVIDIA Ampere cards, but I haven't looked into that yet. (As far as I'm aware, the latest version of MODS MATS that is available only supports nVIDIA GPUs from Kepler to Turing). Artifacting can also be caused by the core/memory controller. If the memory controller is dead, you'll likely find discolored underfill around the core. If you see that, the core is dead, or pretty much dead. 6) If you simply have no display, it's very likely just a missing voltage rail, probably caused by a defective logic gate somewhere on the board and/or physical damage (missing capacitor or resistor, GPUs w/o backplates are obviously quite prone to physical damage, so it might just be worth to take a closer look at the back of the PCB). That said, defective memory can also cause no display. In that instance, do step 5. There are many more things that can cause a GPU to fail POST or not to work correctly, but a short, physical damage or a missing voltage rail are the ones that I myself have seen and fixed the most (by far). Now, some things in general: 1) Any and all of the discolored PCB and componentry you saw in this video is not caused by excessive heat or overclocking. That's just silicone oil from the thermal interface materials reflecting the light in a funny way. Solicone oil is not conductive or capacitive. It does not damage your graphics card in any way, shape or form. 2) I'm not aware of any GPU that doesn't have some empty pads (and, by extension, missing comnponets) by default. So if you see an empty pad, that doesn't mean someone has used a GPU as a donor for parts. 3) Some GPU models have intentionally missing memory modules to get to the desired VRAM capacity. 4) Pointing a thermal camera at blank silicon doesn't do you anything since the silicon will reflect some IR that's already in your environment and, by extension, throw off your reading. 5) Please don't buy dead GPUs off of random people on ebay. It's a bad idea 'cause most of them are way beyond repair. Only buy dead GPUs on ebay if you can get at least a rough idea of what the fault is and, most importantly, if it's not been messed with already.
I have been fixing gpus for a living for more than a year. Nothing tells me these cards were unfixable. -The 1080 FTW just had a dead filtering capacitor. This model doesn't have fuses but its probably something minor. There are more voltages than 1V8, 5V and 3V3, there's also FBVDD, PEX and NVVDD. -The 3060(?) you ran at 140C, if it wasn't toast it probably is since you ran it at 140C... -No missing ram chip on the 2080ti. Each chip is 1GB, card is 11GB. Missing chip is because they bin the chips that have defective memory controllers and disable that one to have a better yield.
@@chilly1661 Many of his videos he's changing components for the wrong ones (Alpha And Omega DrMOS to NCP DrMOS and vice versa, etc.), making the card, well not work then he cuts the video. I don't know if he's just ignorant or doing this to then buy the "broken" cards from his clients. He also doesn't seem able to diagnose memory issues, bios issues, logic signalling issues. Basically, only thing he seems to reliably be able to fix is a broken fan wire/connector.
18:49 the missing ram chip is normal, they come like that, even the 1080 Ti's are the same 3rd chip missing. Also to note, when you get no picture and the machine appears to boot, this is an indication that the GPU is fine, but 1 ram chip is needing replacing, there is some software you can get to test your GPU memory and it will tell you witch memory chip needs replacing. Of course this runs in dos from a flash drive etc. Also if your system boots but has no screen, try using a 2nd card and it should boot into windows on the 2nd GPU and you can see in device manager that the broken GPU 1 will show up as code 21 error, this means the card is working but the memory on the card, one of them chips has died and will give a page in non page area fault, basically use the NVIDIA software tool and find out what chip it reports dead, swap it out, card fixed. Also if you boot using a 2nd GPU and install team viewer or something like that, log me in, etc etc you can shut it down, take out the 2nd GPU and then boot it up with the broken card, then jump on another PC and login remotely and you will get a very small screen but you can navigate around the system and see that in device manager it has error code 21 device cannot start. Hope with helps, Oh p.s. all the other noted missing parts, are normal, its a re-used board from reference to reference I believe. Good luck and thank you for sharing.
6:30 The fans not spinning aren't really an indication. Those newer mid-range to high-end GPUs have what's called a "Zero-Fan mode" wherein the fans will spin when it's first turned on but then go off again until the GPU reaches a certain temperature or is hit with a higher load. On a usual startup the fans only spinning at the start is not an issue.
@@benmasten5924 That 1080 has dual bios if I'm not mistaken, and the switch is right next to the 8-pin power adapter. He should've tried swapping bioses
That 2080 Ti's board isn't missing parts, that's just cost savings done by the 3rd party board partners. Perfectly normal to be missing the last memory module. Check out some of the tear-downs of that GPU.
It’s actually supposed to be missing that 1 memory module. It’s an 11GB card, and each of those chips are 1GB chips. There’s 12 open spots but only 11 will be occupied on a 2080 Ti.
Well it's clear that this is new territory for you, with the comments on the lack of components on the 2080 Ti (which is definitely a normal thing for GPUs as they re-use board designs for a number of different chips), and running a card without a heatsink on it at all. That made me cringe because seeing 130°+ on the package means the junction temp was probably a lot hotter... For testing a GPU's board for temps a good way to get a view is to either rotate the heatsink 90° and mount it that way if possible, or use a CPU cooler to give yourself a visual of the rest of the board while it's powered on. This isn't a good thermal solution for the power management and memory though, so I wouldn't recommend doing any stress tests with this type of setup, but it's good for checking the rest of the board when voltage is applied. With most GPUs now, they turn on the fans for a brief moment after boot to give a kind of "sign of life", then turn them off as the board boots up and starts controlling the fan speed based on temperature. At low enough temps, the fans will be off, hence why you see the fans turn on and then shortly back off when turning the pc on.
18:34 that's how that card is made, it's not missing any components. Also by looking at what they did with thermal pads there's no way they could remove components so cleanly.
Couple of tips for people new to GPU repair: before plugging the GPU into a working system, make sure none of the major power rails are shorted, or you might end up with more dead hardware than you started with. Also with the GPU prices recently, it's reasonable to assume that any "expensive" card sold for salvage has already been diagnosed and either deemed a no-fix or made worse (some people get desperate and put them in the oven). Get older NVidia cards like 700/900 series to get started, much higher success chances. It's usually harder to find info for AMD cards, so stick with NVidia until you get a general feel what all the parts on the card do.
The problem with dead graphics cards is that when there's a catastrophic power failure in one of the components, there's a high chance it will fry a lot of other components in the process (which may not be visually apparant). A lot of times the GPU core and/or memory modules get destroyed as well. They are extremely sensitive to voltage spikes. Even a couple of volts too high can instantly destroy them. There's also a lot of annoying issues that can happen like faulty memory modules which you have to individually test for and then replace the module. In general graphics card repair is a huge pain and only recommended if you really specialize in it and know all the little details of what's happening with each generational architecture / board design. I know a lot of repair shops avoid graphics cards, also because there's so many different models that it's hard to get replacement parts. According to NorthRidgeFix over 50% of graphics cards they work on end up being a "no-fix" either because the problem cannot be found or the GPU core is dead. Big time waste.
It's normal for graphics cards to get hot in random places especially on memory chips so heatspots don't mean the end of graphics cards. The processor on the 3060 that is discoloured is most likely bad so it is most likely unfixable. Also you should test the 12 volt 5 volt 1.8 vmam vcore for continuenty. If it shorted to ground you can apply voltage and you will see with the thermal cam . If the memory is shorted it's a lost cause. The second doesn't seem too bad but not unfixable since all the voltages are present is most like a 0 ohm resistor. You could have checked with the thermal cam for a heatspot while being connected( if there is no short the thermal cam will show nothing ). In graphics cards While using the thermal cam you should know that the heat is very very minimal you will hardly be able to see it(random spots might get hot but when something is faulty on GPUs the heat is low , unless its the processor.).
Current gpu s have really low resistance. Multimeters would identify it as a short by mistake. Just have to check if it’s above 0.7ohms. Entirely possible it’s even lower than that
What always amazes me about these is learning about the insane number of components besides just the ICs. For these to work properly all these capacitors, resistors, mosfets, etc. must work within tolerances and not fail. Even with a tiny chance of failure per year the probability of a failure somewhere becomes very high. So, the chance of failure for these individual components must be near non-existent. That's amazing manufacturing quality in fundamentally inexpensive components. That means the entire supply chain making these components must be extraordinarily tight on accuracy and lack failure points. It quickly becomes bewildering.
The "missing" components on the 2080 ti often means that it is using the same pcb as a higher tier of gpus/a better version of that card that needs those extra components.
As some comments pointed out, the missing RAM module in RTX 2080Ti is normal, that is the reason why that card has "only" 11GiB VRAM and 352bit memory bus (instead of the 24GiB and 384 bits memory bus of the Titan RTX card which shares the same chip), and also the missing components of the power phases are intended by design to save costs. Other thing I think might come in handy: when testing for RAM problems like in the second to last GPU (if i remember right it was a 3060) just use MODS/MATS (a bootable tool for internal testing at nvidia which got leaked about a year ago, beware of scams and viruses tho), that usually identifies the broken chip. The memory chip itself is dirt cheap (5-ish USD, more or less depending on model), and for someone with your skills pretty easy to repair (you will need to preheat the board quite a lot, but apart from that, I was able to perform this repair as a complete novice on a dead GTX770)
Depending the card, the fans won't spin up until needed. Also, I wouldn't run them without at least a fan on them. They heat up quick (within seconds) and can cause permanent damage.
I've been working with JLCPCB for a couple years now. They've been an economical way to test my projects. I've made and sold a couple of devices through them. I've only had one issue with them, some of my red PCBs had cosmetic issues. Nothing too serious though. Great service.
@TronicsFix Im sure someone has already mentioned this but just in case, you shouldn't be running a gpu or cpu with absolutely no heatsink... All these modern processors has a thermal limit of around 100°c. That GPU was already overheated before it even booted into windows. As for most of the cards that looks like they experienced excessive heat around the GPU itself is pretty normal for a perfectly working card. As I mentioned above these cards run around 80°, 90° normally under heavy loads especially in a case with bad air flow. And they will go up to around 100° maximum worst case scenario. So with the pcbs or gpu die's show signs of excessive Heat is not necessarily an issue...
A lot of GPU's Have places for extra memory chips and other parts but they don't use them. They're only used on higher end models like going from a 3070 to a 3080. This is due to the PCB being the same for each model.
For the 2080 ti this is normal if there are some "missing" parts, if there is no flux around that means it was like that from the factory. Removing the extra thermal pads might just "fix" it.
It's really cool to see you use your skills to try to repair something you're still new to. Really interesting content. It also helps newbies like me to see how you work through something you're new to.
With the 3060 perhaps consider reducing the Memory clocks with MSI afterburner. On the 2080ti with "missing parts"- they're probably not missing, the board design is probably just reused between different models. You could try a heat gun on the non working cards, but GPU's are difficult to fix even for people who know what they're doing.
@@ksevio They have standard board designs- he didn't test the 2080ti at all just presumed that it wouldn't work. The 3060 probably had dead gddr6 chips as some of the early RTX cards
While this was disappointing in a specific way, it was really educational and I am sure you saved a lot of people from by lots or used graphic cards like this. Thank you for helping us viewers in this way.
It's always wise to test the unknown vga cards for short circuits before installing in the MB (first 3 pci pins and the power conector pins), a shorted vga card can break the MB or psu. Really like your videos! 👍
I'd not apply thermal paste before ensuring that it at least boots properly with no artifacts! Also having an iGPU in your processor or just a second graphics card would be nice when there's no signal out to see if the card is detected. A frequent failure is no VRAM detected, so GPU Z can tell you about this. I think that's what happened to the RTX 3060 Every missing component in the 2080 Ti looks factory, so why don't you test the card? a GPU failure happens but it's rare!
when testing gpus you can insert 2 gpus in the pc. one you want to test and one you know work. when the display is connected to the known working gpu you ca go to device manager and see if the one you want to test shows up. also never stress test agpu without cooler installed, even booting it without is a risk. install the driver for eatch gpu. use ddu to delete the drivers from previous gpus.
You should watch Tech Cemetery on GPU repairs, with your testing every GPU in this video is still potentially fixable. Well you may have killed the 3060 running it at 140° C, never run a GPU without cooler, put a CPU cooler on it for testing but don't stress without cooling VRMs and memory. Also GPUs often turn off fans after initially powering on since the VBios takes control of cooling then, most GPUs have a silent fan curve where the fans only kick in when a certain temperature is exceeded (mostly around 50° C)
If your not learning from Steve then you will from the comment section. Some great advice and tips on how to test and fix these and other GPU's. Thanks for sharing Steve and people in comment section!
the EVGA RTX 2080 Ti looks perfectly fine on the PCB, no missing parts, only the amount of thermalpads is the issue here, they are too thick and no cooler contact on the gpu chip. Try to apply normal thermal pads (Thermalgrizzly are the best) and the perfect amount of thermal paste, i bet they work.
EVGA also offers a second/third owner warranty. So you could have simply tested the serial number on their RMA page and had the cards exchanged free of charge... After you've tried repairing them, of course a RMA is no longer possible. It's a shame that so few of the graphics cards can be fixed. At least they now have the perfect amount of thermal paste.
If i remember correctly evga doesn’t have any problem about warranty if you open the card, the sticker is present but it’s not for the warranty. This is because I see a lot of video of jay two cent where suggest on evga card to replace thermal compound without any problem with the warranty
@@ChristianBelotti If that's the case there should be no faulty EVGA cards for sale as they would have been replaced with new, I think you will find opening the card for any reason voids the warranty, that's why they have special seals on them.
I can watch now and see the mistakes and misconceptions. I was doing worse than that but in time I was learning even from this kind of terrifying videos. We can learn from everything if we want to learn, especially from mistakes and sometimes we can learn from someone else's mistakes. This video is funny for professionals but for beginners is just right. A beginner should always buy the cheapest of things to practice on because a beginner will almost always ruin them all. No beginner should spend thousands of dollars in hope it will magically repair the things. It doesn't work like this. It's learning first and a bit of luck, a trained eye and more or less steady hands but knowing what you're doing is a must. We all were here. We all thought that are missing parts on these boards even if they actually weren't, let's be honest about that.
It's always a real crapshoot when you buy electronics, especially GPU's, second hand. Too many bitcoin miners out there burning out GPU's and then returning them to stores or offloading them on ebay. I got lucky and picked up my MSI AMD Radeon RX 580 8GB via Ebay back in 2019 used for $180.00. I was expecting a burnt out paperweight to come out of the box when it arrived, but I was surprised to find not only did the card look almost new, after I gave it a good cleaning, it worked fine and is still going strong today.
When I first used jlcpcb service I got excellent customer service as I forgot to send them all of the nessasry Gerber files. They simply asked politely to double check if I had everything and I of course did not. The keyboard PCB turned out with extreme quality and also good shipping. would buy from them again.
Suggestion for 9:10: I have had a bunch of fully working gpus which I had trouble showing any signal on multiple monitors and cables and computer systems. An indication of a good card is if you see the fan start up AT ALL and the motherboard does not flash a red LED near the gpu. (Not nessesarily certain all motherboards will have these leds, but they generally all have some way of telling you if there is an error with the hardware) Its common for fans to spin for a brief moment while starting the card, then turned off again once the data from the temperature sensor is available. If we assume you have the same issue, what I have found which might help is the following. 1. Connect the cable to the monitor and choose the corresponding source. 2. Power on the computer normally. 3. Then press a reset button, not power on/off. You have the primary reset button (connected to the switch on the tower) and some motherboards even have a secondary one directly on the motherboard. Either of those should do the trick if we had the same problem. Though I really don't understand why I have been experiencing this.
Giving up on that 1080 was not a good thing to do. Modern gpu cores very rarely get hurt from any kind of over clocking. They would've had to flash the bios of the card or they would've had to do a shunt mod in order to cause damage to the core from over clocking. It's worth it to try to fix it
Failed bios flash on the GPU could be the issue if the card lacks dual bios. Fwik it can be fixed with another gpu. Quite common to flash other bioses onto gpus to unlock higher power limit
I love how he does some research before making the videos, also asks for people in the comments for help as he stated that he's not done this before, and is always pushing his own boundaries to improve and get better and always takes us foe the ride! Keep up the great content it's nice to see.
Never run the GPU naked, put at least a small alu sink on it even for short power tests. THe 3060 also could have flashed a mining BIOS on it, you should try to flash the manufacturer BIOS again. Also some dis-colourisation around the GPU mostly happens in around the clock running systems and is normal ageing.
Lots of good comments here, when you bought these cards $3K was probably about right, but in today's market with GPU prices dropping like they are, $3K is a big OOF. It's a shame you didn't get around to them several months ago when you bought them, but hey, we get it, you're a busy guy you can't just drop everything to try and fix these. Definitely an expensive learning experience, and perhaps if you read some of the comments, a part 2/follow up might yield some different results.
You'd be surprised how many dead GPUs I've come across that overheated from misuse (folks leaving their PC case open and not cleaning it regularly/using a case that lacks proper ventilation).
There's also the GPU software that has the fans set to Auto , and in my opinion doesn't keep the card cool enough . I always set my fan speeds before gaming. The extra fan noise is worth it.
2080Ti is not missing a chip. That's how high-end GPUs SOC is configured to work/use the amount of VRAM with the bus. Make a research first. Same with the missing VRM parts. Same apply for 1080 Ti and 3080++
EVGA 10 series cards of that era, especially the early ones did have a tendency to blow components. If you were lucky it wouldn't take out the core, but often, it did. One thing to be wary of with crds that seem to work on the short term, sometimes cards take a good while to actually crash. I had a 3060Ti that would run for a long time and then randomly go to a black screen with the fans at 100%. Only long-term tests are going to catch those, so be careful if you're going to try to sell the ones you think are good. You might be sitting on timebombs and not know it.
Yep, in Nov 2016 EVGA offered free kits for users to install and reduce those failures. I still have my 1080 that I rebuilt. It was basically just a bunch of new thermal pads.
My 1070 burned to a crisp too... I'll avoid EVGA whenever I can... I had a PNY GTX960 since 2015 and never had any problems even with many lightning it faced throughout the years...
I inherited a 1070 FTW which was in fantastic condition, but which also didn't produce any display. I couldn't even use MATS to determine if one of the VRAM chips went bad, since it absolutely refused to display anything. I would've loved to upgrade to that card from my current one, but since I didn't have the equipment or knowledge to fix it, I instead sold it for parts/repair. The PCB itself looked great as well, with nothing out of the ordinary.
RTX 2080 Ti not missing parts, if you check out a board view it's normal, you can see this kind of thing with other GPU-s or with low end laptop boards where the whole GPU is missing
Yeah I don't think that GPU had missing bits in it. That's just how they make them, they have a similar board for a lot of the GPUs, and add in memory chips and such as needed for whatever card it's meant to be.
Wow you weren't so lucky with this job lot 😢 Having a way to test the GPU's is enough to put me off I think 🤣 I hear they are one of the biggest things in repair demand at the moment though because of the shortage. Great video nonetheless bud 😊
A few years ago when money was tight and needed a graphics card for my PC , I would look in the junk pile at the used PC repair shop I worked at . He would let me pick some out and I took them home and bake them and re apply the right amount of thermo paste and hope for the best . 70% of them would work and stayed working for a good while. Now I just buy new ones . lol
Hi Man! that was nice to see. I have a couple points that could help for a further revision of this cards: 1. Usually the mining cards shows kind of a glossy spots on the PCB under the memory chips, main chip and the VRM/mosfets. So it looks like a couple of this ones were used to mining, in wich there are additional spots to check like bios, drivers and memory damage. 2. Try not to test the cards without a brick over the main chip because that temp rise over 120C could kill them even in testing 3. Usually the testing for voltage goes: 12v ext input, 12v PCI, 3.3 PCI, 5v output, 1,8v logic, 1v mem controller, 1.5v memory and 0,9v to the GPU core. usually when all the voltages are ok but the GPU, and the mosfets are ok (no short between the phase inductors and ground), could be the phase regulator or further components there. Sometimes there are issues with the regulators for the memopry controller or the memory chips and those could be quite tricky to check because the fall damage just by touching the wrong pin on testing, but replacements are cheap and could worth the repair. 4. This mining cards like the 3060 could show damage on the memory chips in form of artifacts or green spots on games/benchmarks. There's a tool for nvidia GPU to test the memory damage and usually shows the chip that has to be replaced. The tool also exists for AMD but it's not as fancy I use to get damaged cards and it use to happens that people just sells them when someone already tried to repair it and goes FUBAR for the cards. So i learned a lot from what not to do with this...
Going into gpu repair without any experience is like going into a sand trap with no one to pull you out. The people who make these things make them so finicky and most need a decent motherboard plus software to remove old gpu software before installing new ones etc. nvm the dozens of other things to think about. Basically if you do one of these videos again stick to trying to repair maybe just one or two decent cards like a 2080ti or 1080 and focus on just those. Sorry that someone convinced you to waste your money on all those. Assuming u already knew all of that well, there’s not much you can do if someone harvested pieces of the circuit board, yikes. Hope one day you have a way to get those pieces and possibly repair a board that was scrapped if that’s even possible. Would be interesting to see. Thx for the upload and GL!
I think the issue with the 3060 might be memory, try using MSI Afterburner to downclock the memory, and try running the benchmark again. For the 2080ti, I think those components are meant to be missing, as there is a more powerful card, the Titan RTX, which would use that extra VRAM chip and likely needs a bit more power from those components that are missing.
That 2080Ti has all the necessary parts on it. The 2080Ti has 11 1GB memory chips but there is space for 12 since the design is shared with the RTX TITAN which has all memory spaces populated with 2GB modules.
Generally speaking running GPUs with the potentially several hundreds of watts that run though the GPU dies without a cooler can be a pretty sure fire way to overheat and kill them it is HIGHLY recommended to not run a GPU on without a cooler attached as it can very very easily damage the GPU die itself as it gets caught with a thermal runaway before it can shut itself down
Also like many others have mentioned the 2080 Ti board was normal that had nothing missing companies use a set PCB design and add parts as they need depending on what SKU it is to save on manufacturing costs since the 2080 Ti is an 11GB card it only needs 11 ram modules
The biggest thing too is making sure to wipe and reinstall drivers between each GPU. Sometimes the Nvidia drivers don’t properly detect new devices until a restart after windows detects them.
Hi, Great content. Repairs of Nintendo, Xbox, & PlayStation are interesting but I especially like when you purchase other items for repair like these graphics cards. The video on the Go Pro's was also really interesting.
I not sure if it's important. The drivers can be wrong. When you go to change a graphics card you suppose to with old card in then go into program and features and delete the driver. Then u install new driver.
I suggest some things to do if the GPU is not sending signal to the monitor: Connect the GPU to a system with another GPU to see if it is at least recognized in Windows in "Device Manager," if it is detected I suggest you try flashing a VBIOS that you can get from the internet, there are many tutorials on how to do it. If on the other hand the GPU starts but is very unstable during testing, even there you can try flashing a VBIOS downloaded from the internet to see if the situation improves, otherwise you can try underclocking to find a value where it remains stable
Love the comments on these videos, everyone has been repairing GPUs there whole life and know exactly what's wrong with every single one in the video lol. Great video Steve keep em coming.
the trtx 3060 prolyl is fixable. that series suffered from overheating memory. which also looks possible due to teh artifacts and errors your getting - prolly one of the first or second row of mem ICs has a fault ( could also be just popped solder balls due to thermal stress ). id recommend do a memtest on it and either reball the memory ic /s or get a replacement ( their about 7 bucks rightnow )
Finally someone who knows what is saying...artifacts comes definitely from a bad memory chip, also the discoloration around GPU is probably caused by the neighbor chips gone really hot
The second-hand Ebay prices are average prices the card sells for if its working. Just to let us know what he could sell it for if it works or gets fixed. Not if is not working
@@ikwilgewoonfilmpje is not even worth that new. 1080ti, which is the better variant, is going for $400, in working condition. A plain old 1080 is not worth $500.
that one that wasnt putting out a display might have a mining bios on it if the card is being detected but not displaying anything. edit: if you have an integrated graphics on your cpu or have a cheap graphics card and plug them both in, if you plug in the one that works and go to hardware information it should tell you if the one thats not displaying is plugged in or not.
Common issue on the gpu ,on of them are vrms mosfets, vrm controller like 5v 3.3v 1.2-1.3v and 900mili volts to 1 volt, and the most common video ram colder joints and failure u have to replace it with the donor board
Many GPUs (especially higher end 70 and 80 series models) often come with unpopulated VRAM packages and power supply phases. That is completely normal and is done due to the fact that: 1. They are using a reference design from Nvidia and didn't bother removing unneeded components -> cheaper to design and less work. 2. The actual GPU chip itself has been binned in the factory. Not every GPU chip is perfect. Some might for example come with a few non functional memory channels and sections inside of the GPU that don't work. These large chips are quite prone to manufacturing defects due to their large size, which results in parts of the GPU being broken and useless. Instead of just throwing these otherwise perfectly fine chips away, Nvidia go ahead and overspec their reference design, so that the manufacturer (ASUS, Gigabyte,...) can just leave the required external circuitry for the broken section out. If for example GPU sections A,B,C,D work fine but section E is defective then it makes no sense so e.g. solder a VRAM chip to section E, since section E is dead. Same goes for the power delivery of that section. So the reference PCB design comes with more external circuitry than what's actually used. It's also possible that the same bare PCB is used in multiple GPU versions and they just didn't populate all the chips due to cost saving measures. There's usually low and high end versions of the same GPU. (E.g ASUS TUF vs Rog Strix, Strix being the premium models). If a GPU chip has too many broken sections to still be considered for example an 80 series card, they disable the extra stuff and rank it as a 70 series card instead. This is binning. It's also how Intel i7,i5 and i3 CPUs work. Nvidia doesn't make a new chip for every GPU model out there. They only have 3 or 4 "base chips" per generation where they just go ahead and disable features based on the amount of manufacturing defects. Long story short: That GPU with the missing VRAM chip and the doubled thermal pads probably still works just fine. What I think happened there is that someone installed new thermal pads but didn't pay attention to their thickness, so the GPU chip didn't make any contact with the heatsink and therfore instantly overheated and throttled itself, making them think they broke it.
Another thing I didn't see you do or mention, wiping the drivers. Having multiple gpus and multiple drivers can be conflicting. I would try some gpus again and DDU the drivers.
'Missing' components on the board is a normal thing in the GPU world. It's just a way to reuse the board design for different models for graphics cards. Also, bold of you to turn your system on with a faulty GPU. You can fry the rest of your system that way.
18:55 Look up what the PCB is supposed to look like before you assume someone took parts off. It has 11 memory chips for 11 gigabytes of memory. Also EVGA has several versions of most cards and I don't know if some have more diodes. I think the problem with this card may have just been someone putting so many thermal pads on it that the heatsink didn't touch the chip or paste.
That 2080 Ti wasn't missing anything. That model of it just doesn't come with those components. The same PCB is often used across multiple models. If you do this again, you should look for teardown images of the GPU in question first, which are usually easy to find in reviews.
Small tip, with gpus you need to cover the entire Die with thermal paste :) also the EVGA 1080s a lot of then had fail issues. I love your videos. Such a fascinating watch every time ❤️❤️
You just need to put enough on the die that it will spread across the whole thing when pressed against the heatsink. Same with any chip without an IHS over the die, like the consoles he often fixes, or most laptops.
The EVGA 1080 FTW he showed had a flaw that the V-RAM could get to hot. EVGA had a fix by useing a set of thermal pads, which you could get from them. Till today my EVGA 1080 FTW is still working 😁 and i was lucky enough to read about the flaw, otherwise it was also RIP
you could check the serialnumbers on the EVGA Cards if theyre still under warranty. The 3060 could be, open an RMA and send it in and get a new one. You also could check to try to reflash the VGA Bios on the cards, since they could got corrupted and flashing could fix it
@@AnotherDud3 wrong as long the card is not physically damaged. It’s still under warranty. Cards are still under warranty if you change the thermal paste or the cooler.
1 thing to point ☝️ is you should check the pcie version in bios and match that for your gpu! I know it won't always cause no signal but it might just do that some times
Great video as always! GPUs can be very tricky, but it's interesting watching something different. I hope you can sell the parts and get your money back!
Fans usually only turn on when the card reaches a temp threshold. Otherwise it will run without them spinning. I almost choked when I saw you boot up without any cooling at all. I then died when you ran a benchmark lol
try installing the drivers for the cards. That most likely will fix the issue. If not you can try updating the BIOS on the MOBO to see if that will fix the issue
I have a similar EVGA 3060. I have 3 years warranty. I'm 100% sure that yours should be still under warranty. Before you open, you need to check the serial number on their website. But since you rip the warranty sticker, it's no use then.
I would Keep all the cards you have not fixed, So that once you are familiar with Mosfets and Vram and how to test them when shorted or Open you can replace the ones that have failed, with both components they should diode test much as each other so when one reads totally different, that's the one that needs replacing along with perhaps one or two supporting components that caused either the Vram or Mosfet to fail..
@UCK0_WC3SrgMhmMCfOQ_Wp0Q 🤦♂Oh REALLY???..🤣 Yeah I don't see this as legit because for 64 plus years I have never EVER won anything. 🧐 I think someone is about to go down for going against what is obviously the universes intent that I never EVER be a winner without entering as a contender..
Hey. What happend to my comment. You missed a few things. Glitchy screen is 87% memory issue. Run mats/mods to find out. Discoloration is okay. The pads will leak oil and other cemicals and it will discolor the pcb even at 60-70°C. Depends on brands. About the missing components. Nothing is missing. Is is just a reference board. In the same generarion, those spaces will be populated on higher end gpu's. Check google pics. About the one that does not turn on: Check the voltage power up line. 12v, 5v, 3.3v, 1.8v pex, vmem, vcore. They have a specific order. The last thing is you can get mining vbios that disables the output to gain some more hash. Flash them back.
Hey Steve! Nice vids btw, Ive seen in the recent videos you've been playing with the aound modifier and its cool keep doin it, but an idea for you... In a dramatic scene like in this video show a video of yourself doin a sad face as if it didnt work, but dont say it... Think it would be rly funny!
Many GPU's come with thermal pads. Obviously that card with the stacked pads was modified, But just because the pads weren't cut to size for the individual components doesn't mean its been modified - its a labour cost cutting excercise. Its quicker to slap on 1 thermal pad for 4 chips than to cut them to size for each individual chip.
I had to look up 0 ohm resistor, because there is no such thing. 0 ohms means 0 resistance, so it can't be a resistor. Apparently they are just wires packaged in a resistor-sized package, to make design easier (it's a way to connect components without using exposed wires. Anyway...I wonder if the bad/glitching GPU's were ruined by mining. That explains the 1030 working, that one would be useless for mining. For the 3060, my guess is some of the RAM is bad. Probably died from overclocking (again due to mining).
As I said at the beginning of this video, I'm no graphics card repair guru. I have no experience trying to fix them so this was just an experiment. If you're looking for good advice on how to repair graphics cards be sure to read the comments on this video as there is great advice from smart people.
do more of these pc parts videos
I had that exact same 1080 ftw. 2 of em actually. both had similar problems because evga cheaped out and didn't add thermal pads to the vrms from the factory.
@@iamdmc there is no need to be rude, this is a learning opportunity for him, even though better research could've of been done.
@@iamdmc Why does it matter to you just click off the video hes trying something new bozo
@@MoMo-qw9ro your are absolutely right, a simple google research would show up the board for the GPU he is working on
1. Never, ever allow GPUs to boot past BIOS without a cooler mounted.
2. That 3060 with the green artifacts is usually an indicator that a RAM-chip is either bad or has lost connection with the PCB. You could try a reflow of the RAM chips.
2080 Tis come with only 11 of the 12 memory channels populated from the factory, also, missing phases in the various VRMs is a normal strategy for the OEM to save money while still using the reference board design.
thats what I was thinking. Isn't each vram chip usually 1GB, so maybe that was a 11GB model but the board could also be used for a 12GB model and thats why 1 of the vram chips were missing? I also was thinking it came that way from the factory because the board looked way too clean, it didn't look like anyone went at it with flux and a soldering iron or heat gun since usually there would be some flux residue on the card afterwards
Yep, exactly what I was about to say. That EVGA 2080 ti xc ultra isn't missing any components. Also, one possible reason for the stacked thermal pads would be that the card was on a waterblock, got defective, and the just swapped the stock cooler back on to sell without caring about it being done right.
Yeah I'm bothered by him not researching the pcb or breakdowns out there. Love his videos but with today's resources this is something he could have researched on the fly.
Yeah, same with some other nVidia top-end GPUs below the Titans, like the GTX 780, same GPU (GK110) but partially deactivated and half of the VRAM, just 3 GB and these are only on the GPU side, the Titan has another 3 GB on the other PCB side; they simply share the reference PCB so the same watercooler fits both cards.
And also the 980 Ti and Titan X (both GM200 but Titan got 12 instead of 6 gigs VRAM), 1080 Ti and Titan X (Pascal)/Titan Xp (all GP102 but the 1080 Ti got 11 and the titans got 12 gigs).
Came here to leave the same comment. OEMs frequently make the same base pcb for a range of cards and populate based on the version to save money on the manufacturing and engineering side of things. Other times they over engineer in design and then find they don’t need all the extra phases or memory modules to hit their prospective benchmark. “Missing” components is a common sight.
1) Don't ever just install and attempt to run a GPU that you know has an issue or is likely to have one.
2) Check for shorts on the power input rails aswell as on the voltage rails the card itself creates (5V, 1.8V, PEX, vCore, vMEM, VDDCI (on AMD GPUs), MSVDD (on all GA102 GPUs except for the 3090Ti)). Each power input is it's own 12V rail, so just checking one will not suffice. vCore, vMEM, VDDCI, MSVDD and PEX have rather low resistances, so check them using resistance mode. DO NOT use continuity mode, or you might just lead yourself down the wrong path. (Getting a close to 0 Ohm reading on vCore on newer GPUs is expected since the GPU silicon has a very low internal resistance.)
3) If you have a short, don't ever just inject voltage, especially if a power input is shorted. First, check if any power can make it's way from that power input directly to the GPU or memory using your multimeter. If there is a direct connection and you just go to inject voltage, you might just be sending whatever voltage you're injecting directly into your GPU or memory, potentially frying them.
4) If you have a short, inspect the componets that are on that shorted voltage rail closely. Don't waste your time looking at componets that aren't on that rail (they can't cause the short). Look for things such as blown capacitors, mosfets, powerstages or a bubbled-up top PCB layer. Look at the solder on mosfets and powerstages, see if there's some small solder blobs or balls around or if the joints look bad and not properly flowed. Those are indications that something got quite warm.
If you have a short on PEX, your core is most likely dead. AMD Vega cards a notorious for having defective cores.
5) If you don't have a short anywhere, run the card. If you have artifacting, it's very likely one or more defective memory banks, mayhaps even across multiple memory channels. For nVIDIA GPUs from Fermi to Turing, use nVIDIA MODS MATS. Run it, and it'll tell you in a plain textfile which memory bank is spitting out errors. You can probably do the same on AMD GPUs and nVIDIA Ampere cards, but I haven't looked into that yet. (As far as I'm aware, the latest version of MODS MATS that is available only supports nVIDIA GPUs from Kepler to Turing). Artifacting can also be caused by the core/memory controller. If the memory controller is dead, you'll likely find discolored underfill around the core. If you see that, the core is dead, or pretty much dead.
6) If you simply have no display, it's very likely just a missing voltage rail, probably caused by a defective logic gate somewhere on the board and/or physical damage (missing capacitor or resistor, GPUs w/o backplates are obviously quite prone to physical damage, so it might just be worth to take a closer look at the back of the PCB). That said, defective memory can also cause no display. In that instance, do step 5.
There are many more things that can cause a GPU to fail POST or not to work correctly, but a short, physical damage or a missing voltage rail are the ones that I myself have seen and fixed the most (by far).
Now, some things in general:
1) Any and all of the discolored PCB and componentry you saw in this video is not caused by excessive heat or overclocking. That's just silicone oil from the thermal interface materials reflecting the light in a funny way. Solicone oil is not conductive or capacitive. It does not damage your graphics card in any way, shape or form.
2) I'm not aware of any GPU that doesn't have some empty pads (and, by extension, missing comnponets) by default. So if you see an empty pad, that doesn't mean someone has used a GPU as a donor for parts.
3) Some GPU models have intentionally missing memory modules to get to the desired VRAM capacity.
4) Pointing a thermal camera at blank silicon doesn't do you anything since the silicon will reflect some IR that's already in your environment and, by extension, throw off your reading.
5) Please don't buy dead GPUs off of random people on ebay. It's a bad idea 'cause most of them are way beyond repair. Only buy dead GPUs on ebay if you can get at least a rough idea of what the fault is and, most importantly, if it's not been messed with already.
Something i'd like to add is that some GPU's perform just fine on benchmarks but fail on actual gamming, I have one here with this problem.
Very meticulous comment, I think he should do a colab with northbridge fix and they can hopefully share some knowledge with each other.
Graphic cards are so confusing. I'll stick to me ps5 and buy a new one if it dies! 🤣🤣🤣
Who are you the GPU police
This -essay- comment is a gold mine, thanks for the tips!
I have been fixing gpus for a living for more than a year. Nothing tells me these cards were unfixable.
-The 1080 FTW just had a dead filtering capacitor. This model doesn't have fuses but its probably something minor. There are more voltages than 1V8, 5V and 3V3, there's also FBVDD, PEX and NVVDD.
-The 3060(?) you ran at 140C, if it wasn't toast it probably is since you ran it at 140C...
-No missing ram chip on the 2080ti. Each chip is 1GB, card is 11GB. Missing chip is because they bin the chips that have defective memory controllers and disable that one to have a better yield.
Do you have any advice for someone looking to get into fixing GPUs or electronics in general? How did you learn? Thank you =)
@@phyro4143 hi, why do you consider northridgefix a scammer? Just wanted to know.
@@chilly1661 Many of his videos he's changing components for the wrong ones (Alpha And Omega DrMOS to NCP DrMOS and vice versa, etc.), making the card, well not work then he cuts the video. I don't know if he's just ignorant or doing this to then buy the "broken" cards from his clients. He also doesn't seem able to diagnose memory issues, bios issues, logic signalling issues. Basically, only thing he seems to reliably be able to fix is a broken fan wire/connector.
I guess the 3060 had a bad RAM chip.
@@phyro4143 yeah the boss fix the easy one. For complex repairs with multiples ic failure leave to the most junior employee 😬
18:49 the missing ram chip is normal, they come like that, even the 1080 Ti's are the same 3rd chip missing. Also to note, when you get no picture and the machine appears to boot, this is an indication that the GPU is fine, but 1 ram chip is needing replacing, there is some software you can get to test your GPU memory and it will tell you witch memory chip needs replacing. Of course this runs in dos from a flash drive etc.
Also if your system boots but has no screen, try using a 2nd card and it should boot into windows on the 2nd GPU and you can see in device manager that the broken GPU 1 will show up as code 21 error, this means the card is working but the memory on the card, one of them chips has died and will give a page in non page area fault, basically use the NVIDIA software tool and find out what chip it reports dead, swap it out, card fixed.
Also if you boot using a 2nd GPU and install team viewer or something like that, log me in, etc etc you can shut it down, take out the 2nd GPU and then boot it up with the broken card, then jump on another PC and login remotely and you will get a very small screen but you can navigate around the system and see that in device manager it has error code 21 device cannot start.
Hope with helps,
Oh p.s. all the other noted missing parts, are normal, its a re-used board from reference to reference I believe.
Good luck and thank you for sharing.
@Mr Prongles Agreed
6:30 The fans not spinning aren't really an indication. Those newer mid-range to high-end GPUs have what's called a "Zero-Fan mode" wherein the fans will spin when it's first turned on but then go off again until the GPU reaches a certain temperature or is hit with a higher load.
On a usual startup the fans only spinning at the start is not an issue.
Maybe that card has corrupted bios?. Im just guessing honestly
@@leexy3395 I was thinking that or mining BIOS
@@benmasten5924 That 1080 has dual bios if I'm not mistaken, and the switch is right next to the 8-pin power adapter. He should've tried swapping bioses
@@razmann4k for sure. He's learning though.
yup. I think they turn on at 60c if I'm not mistaken.
That 2080 Ti's board isn't missing parts, that's just cost savings done by the 3rd party board partners. Perfectly normal to be missing the last memory module. Check out some of the tear-downs of that GPU.
It’s actually supposed to be missing that 1 memory module. It’s an 11GB card, and each of those chips are 1GB chips. There’s 12 open spots but only 11 will be occupied on a 2080 Ti.
That's correct, but I think the top left memory chip is broken. Still fixable probably tho
yup some GPU's share the same board as the higher tiers. Like the 2080 ti shares the same board as the RTX Titan.
Well it's clear that this is new territory for you, with the comments on the lack of components on the 2080 Ti (which is definitely a normal thing for GPUs as they re-use board designs for a number of different chips), and running a card without a heatsink on it at all. That made me cringe because seeing 130°+ on the package means the junction temp was probably a lot hotter... For testing a GPU's board for temps a good way to get a view is to either rotate the heatsink 90° and mount it that way if possible, or use a CPU cooler to give yourself a visual of the rest of the board while it's powered on. This isn't a good thermal solution for the power management and memory though, so I wouldn't recommend doing any stress tests with this type of setup, but it's good for checking the rest of the board when voltage is applied.
With most GPUs now, they turn on the fans for a brief moment after boot to give a kind of "sign of life", then turn them off as the board boots up and starts controlling the fan speed based on temperature. At low enough temps, the fans will be off, hence why you see the fans turn on and then shortly back off when turning the pc on.
only do this if you are going to turn it off right away after seeing that it posts. usually 6 seconds then shut off.
18:34 that's how that card is made, it's not missing any components. Also by looking at what they did with thermal pads there's no way they could remove components so cleanly.
Powering a GPU without a heat sink...
Couple of tips for people new to GPU repair: before plugging the GPU into a working system, make sure none of the major power rails are shorted, or you might end up with more dead hardware than you started with.
Also with the GPU prices recently, it's reasonable to assume that any "expensive" card sold for salvage has already been diagnosed and either deemed a no-fix or made worse (some people get desperate and put them in the oven). Get older NVidia cards like 700/900 series to get started, much higher success chances. It's usually harder to find info for AMD cards, so stick with NVidia until you get a general feel what all the parts on the card do.
The problem with dead graphics cards is that when there's a catastrophic power failure in one of the components, there's a high chance it will fry a lot of other components in the process (which may not be visually apparant). A lot of times the GPU core and/or memory modules get destroyed as well. They are extremely sensitive to voltage spikes. Even a couple of volts too high can instantly destroy them. There's also a lot of annoying issues that can happen like faulty memory modules which you have to individually test for and then replace the module. In general graphics card repair is a huge pain and only recommended if you really specialize in it and know all the little details of what's happening with each generational architecture / board design. I know a lot of repair shops avoid graphics cards, also because there's so many different models that it's hard to get replacement parts. According to NorthRidgeFix over 50% of graphics cards they work on end up being a "no-fix" either because the problem cannot be found or the GPU core is dead. Big time waste.
It's normal for graphics cards to get hot in random places especially on memory chips so heatspots don't mean the end of graphics cards. The processor on the 3060 that is discoloured is most likely bad so it is most likely unfixable. Also you should test the 12 volt 5 volt 1.8 vmam vcore for continuenty. If it shorted to ground you can apply voltage and you will see with the thermal cam . If the memory is shorted it's a lost cause. The second doesn't seem too bad but not unfixable since all the voltages are present is most like a 0 ohm resistor. You could have checked with the thermal cam for a heatspot while being connected( if there is no short the thermal cam will show nothing ). In graphics cards While using the thermal cam you should know that the heat is very very minimal you will hardly be able to see it(random spots might get hot but when something is faulty on GPUs the heat is low , unless its the processor.).
GPU was fine, before he cooked it. Probably an issue with one of these RAM chips.
Current gpu s have really low resistance. Multimeters would identify it as a short by mistake. Just have to check if it’s above 0.7ohms. Entirely possible it’s even lower than that
I wouldn't trust a card that's only stability test was heaven.
It doesn’t matter what Steve is fixing I’m gonna watch it and it will make my day better.
What always amazes me about these is learning about the insane number of components besides just the ICs. For these to work properly all these capacitors, resistors, mosfets, etc. must work within tolerances and not fail. Even with a tiny chance of failure per year the probability of a failure somewhere becomes very high. So, the chance of failure for these individual components must be near non-existent. That's amazing manufacturing quality in fundamentally inexpensive components. That means the entire supply chain making these components must be extraordinarily tight on accuracy and lack failure points. It quickly becomes bewildering.
kudos to tronics for getting out of comfort zone and trying new things
The "missing" components on the 2080 ti often means that it is using the same pcb as a higher tier of gpus/a better version of that card that needs those extra components.
My soul died when you ran it under load without any heatsink 😮 Melty in seconds.
As some comments pointed out, the missing RAM module in RTX 2080Ti is normal, that is the reason why that card has "only" 11GiB VRAM and 352bit memory bus (instead of the 24GiB and 384 bits memory bus of the Titan RTX card which shares the same chip), and also the missing components of the power phases are intended by design to save costs.
Other thing I think might come in handy: when testing for RAM problems like in the second to last GPU (if i remember right it was a 3060) just use MODS/MATS (a bootable tool for internal testing at nvidia which got leaked about a year ago, beware of scams and viruses tho), that usually identifies the broken chip. The memory chip itself is dirt cheap (5-ish USD, more or less depending on model), and for someone with your skills pretty easy to repair (you will need to preheat the board quite a lot, but apart from that, I was able to perform this repair as a complete novice on a dead GTX770)
Depending the card, the fans won't spin up until needed. Also, I wouldn't run them without at least a fan on them. They heat up quick (within seconds) and can cause permanent damage.
I've been working with JLCPCB for a couple years now. They've been an economical way to test my projects. I've made and sold a couple of devices through them. I've only had one issue with them, some of my red PCBs had cosmetic issues. Nothing too serious though. Great service.
@TronicsFix
Im sure someone has already mentioned this but just in case, you shouldn't be running a gpu or cpu with absolutely no heatsink... All these modern processors has a thermal limit of around 100°c. That GPU was already overheated before it even booted into windows.
As for most of the cards that looks like they experienced excessive heat around the GPU itself is pretty normal for a perfectly working card. As I mentioned above these cards run around 80°, 90° normally under heavy loads especially in a case with bad air flow. And they will go up to around 100° maximum worst case scenario. So with the pcbs or gpu die's show signs of excessive Heat is not necessarily an issue...
You know the Tronicsfix video is gonna be good when he goes out of his comfort zone!
A lot of GPU's Have places for extra memory chips and other parts but they don't use them. They're only used on higher end models like going from a 3070 to a 3080. This is due to the PCB being the same for each model.
For the 2080 ti this is normal if there are some "missing" parts, if there is no flux around that means it was like that from the factory. Removing the extra thermal pads might just "fix" it.
It's really cool to see you use your skills to try to repair something you're still new to. Really interesting content. It also helps newbies like me to see how you work through something you're new to.
With the 3060 perhaps consider reducing the Memory clocks with MSI afterburner. On the 2080ti with "missing parts"- they're probably not missing, the board design is probably just reused between different models. You could try a heat gun on the non working cards, but GPU's are difficult to fix even for people who know what they're doing.
Considering it has a space for a memory chip and it had a memory error, I'd suspect it should have been there
@@ksevio They have standard board designs- he didn't test the 2080ti at all just presumed that it wouldn't work. The 3060 probably had dead gddr6 chips as some of the early RTX cards
The board is used with the RTX Titan and the 2080ti. Like how the 3080 ti is still the same chip as the 3090.
Considering that mans tried to run Heaven bench with no heatsink installed the thing is fried lmao
@@ksevio 2080ti has 11gb of vram and all the boards have one empty space(2080ti’s).
While this was disappointing in a specific way, it was really educational and I am sure you saved a lot of people from by lots or used graphic cards like this. Thank you for helping us viewers in this way.
At least he is not speaking very fast or with strange accent
It's always wise to test the unknown vga cards for short circuits before installing in the MB (first 3 pci pins and the power conector pins), a shorted vga card can break the MB or psu. Really like your videos! 👍
I'd not apply thermal paste before ensuring that it at least boots properly with no artifacts!
Also having an iGPU in your processor or just a second graphics card would be nice when there's no signal out to see if the card is detected.
A frequent failure is no VRAM detected, so GPU Z can tell you about this. I think that's what happened to the RTX 3060
Every missing component in the 2080 Ti looks factory, so why don't you test the card?
a GPU failure happens but it's rare!
when testing gpus you can insert 2 gpus in the pc. one you want to test and one you know work. when the display is connected to the known working gpu you ca go to device manager and see if the one you want to test
shows up. also never stress test agpu without cooler installed, even booting it without is a risk. install the driver for eatch gpu. use ddu to delete the drivers from previous gpus.
that is if he has a motherboard with 2 PCI-E slots.
Or just use a processor with an iGPU
@@H3LLGHA5T yes, that too :)
That 3060 is still covered under EVGA’s 3 year warranty. Not sure why the previous owner didn’t Get it repaired/replaced under warranty
You should watch Tech Cemetery on GPU repairs, with your testing every GPU in this video is still potentially fixable. Well you may have killed the 3060 running it at 140° C, never run a GPU without cooler, put a CPU cooler on it for testing but don't stress without cooling VRMs and memory. Also GPUs often turn off fans after initially powering on since the VBios takes control of cooling then, most GPUs have a silent fan curve where the fans only kick in when a certain temperature is exceeded (mostly around 50° C)
I've used JLPCB for years and I won't change as the service is great and the product stunning !...cheers.
the 3060 it looked like you just hadn't installed the driver, hence the low resolution on the TV and big icons etc
If your not learning from Steve then you will from the comment section. Some great advice and tips on how to test and fix these and other GPU's. Thanks for sharing Steve and people in comment section!
the EVGA RTX 2080 Ti looks perfectly fine on the PCB, no missing parts, only the amount of thermalpads is the issue here, they are too thick and no cooler contact on the gpu chip. Try to apply normal thermal pads (Thermalgrizzly are the best) and the perfect amount of thermal paste, i bet they work.
EVGA also offers a second/third owner warranty. So you could have simply tested the serial number on their RMA page and had the cards exchanged free of charge... After you've tried repairing them, of course a RMA is no longer possible. It's a shame that so few of the graphics cards can be fixed.
At least they now have the perfect amount of thermal paste.
The cards had already been taken apart before he bought them, if warranty was valid you can be sure the previous owner would have sent it back.
If i remember correctly evga doesn’t have any problem about warranty if you open the card, the sticker is present but it’s not for the warranty. This is because I see a lot of video of jay two cent where suggest on evga card to replace thermal compound without any problem with the warranty
@@ChristianBelotti If that's the case there should be no faulty EVGA cards for sale as they would have been replaced with new, I think you will find opening the card for any reason voids the warranty, that's why they have special seals on them.
that 3060 was probably fixable until you ran it with no heatsink , if it wasn't fried before it definitely is now.
I can watch now and see the mistakes and misconceptions. I was doing worse than that but in time I was learning even from this kind of terrifying videos. We can learn from everything if we want to learn, especially from mistakes and sometimes we can learn from someone else's mistakes.
This video is funny for professionals but for beginners is just right.
A beginner should always buy the cheapest of things to practice on because a beginner will almost always ruin them all.
No beginner should spend thousands of dollars in hope it will magically repair the things. It doesn't work like this.
It's learning first and a bit of luck, a trained eye and more or less steady hands but knowing what you're doing is a must.
We all were here. We all thought that are missing parts on these boards even if they actually weren't, let's be honest about that.
It's always a real crapshoot when you buy electronics, especially GPU's, second hand. Too many bitcoin miners out there burning out GPU's and then returning them to stores or offloading them on ebay. I got lucky and picked up my MSI AMD Radeon RX 580 8GB via Ebay back in 2019 used for $180.00. I was expecting a burnt out paperweight to come out of the box when it arrived, but I was surprised to find not only did the card look almost new, after I gave it a good cleaning, it worked fine and is still going strong today.
Can't mine Bitcoin on graphics cards.
When I first used jlcpcb service I got excellent customer service as I forgot to send them all of the nessasry Gerber files. They simply asked politely to double check if I had everything and I of course did not. The keyboard PCB turned out with extreme quality and also good shipping. would buy from them again.
Suggestion for 9:10:
I have had a bunch of fully working gpus which I had trouble showing any signal on multiple monitors and cables and computer systems.
An indication of a good card is if you see the fan start up AT ALL and the motherboard does not flash a red LED near the gpu.
(Not nessesarily certain all motherboards will have these leds, but they generally all have some way of telling you if there is an error with the hardware)
Its common for fans to spin for a brief moment while starting the card, then turned off again once the data from the temperature sensor is available.
If we assume you have the same issue, what I have found which might help is the following.
1. Connect the cable to the monitor and choose the corresponding source.
2. Power on the computer normally.
3. Then press a reset button, not power on/off.
You have the primary reset button (connected to the switch on the tower) and some motherboards even have a secondary one directly on the motherboard.
Either of those should do the trick if we had the same problem. Though I really don't understand why I have been experiencing this.
Giving up on that 1080 was not a good thing to do. Modern gpu cores very rarely get hurt from any kind of over clocking. They would've had to flash the bios of the card or they would've had to do a shunt mod in order to cause damage to the core from over clocking.
It's worth it to try to fix it
especially FTW which is factory overclocked.
Failed bios flash on the GPU could be the issue if the card lacks dual bios. Fwik it can be fixed with another gpu. Quite common to flash other bioses onto gpus to unlock higher power limit
I love how he does some research before making the videos, also asks for people in the comments for help as he stated that he's not done this before, and is always pushing his own boundaries to improve and get better and always takes us foe the ride! Keep up the great content it's nice to see.
Love the PC part repair videos. Motherboards, gpus, notebooks... great content.
Never run the GPU naked, put at least a small alu sink on it even for short power tests. THe 3060 also could have flashed a mining BIOS on it, you should try to flash the manufacturer BIOS again. Also some dis-colourisation around the GPU mostly happens in around the clock running systems and is normal ageing.
Lots of good comments here, when you bought these cards $3K was probably about right, but in today's market with GPU prices dropping like they are, $3K is a big OOF. It's a shame you didn't get around to them several months ago when you bought them, but hey, we get it, you're a busy guy you can't just drop everything to try and fix these. Definitely an expensive learning experience, and perhaps if you read some of the comments, a part 2/follow up might yield some different results.
You'd be surprised how many dead GPUs I've come across that overheated from misuse (folks leaving their PC case open and not cleaning it regularly/using a case that lacks proper ventilation).
There's also the GPU software that has the fans set to Auto , and in my opinion doesn't keep the card cool enough .
I always set my fan speeds before gaming. The extra fan noise is worth it.
2080Ti is not missing a chip. That's how high-end GPUs SOC is configured to work/use the amount of VRAM with the bus. Make a research first. Same with the missing VRM parts.
Same apply for 1080 Ti and 3080++
EVGA 10 series cards of that era, especially the early ones did have a tendency to blow components. If you were lucky it wouldn't take out the core, but often, it did.
One thing to be wary of with crds that seem to work on the short term, sometimes cards take a good while to actually crash. I had a 3060Ti that would run for a long time and then randomly go to a black screen with the fans at 100%. Only long-term tests are going to catch those, so be careful if you're going to try to sell the ones you think are good. You might be sitting on timebombs and not know it.
Yep, in Nov 2016 EVGA offered free kits for users to install and reduce those failures. I still have my 1080 that I rebuilt. It was basically just a bunch of new thermal pads.
My 1070 burned to a crisp too... I'll avoid EVGA whenever I can... I had a PNY GTX960 since 2015 and never had any problems even with many lightning it faced throughout the years...
I inherited a 1070 FTW which was in fantastic condition, but which also didn't produce any display. I couldn't even use MATS to determine if one of the VRAM chips went bad, since it absolutely refused to display anything. I would've loved to upgrade to that card from my current one, but since I didn't have the equipment or knowledge to fix it, I instead sold it for parts/repair. The PCB itself looked great as well, with nothing out of the ordinary.
RTX 2080 Ti not missing parts, if you check out a board view it's normal, you can see this kind of thing with other GPU-s or with low end laptop boards where the whole GPU is missing
Yeah I don't think that GPU had missing bits in it. That's just how they make them, they have a similar board for a lot of the GPUs, and add in memory chips and such as needed for whatever card it's meant to be.
Wow you weren't so lucky with this job lot 😢 Having a way to test the GPU's is enough to put me off I think 🤣 I hear they are one of the biggest things in repair demand at the moment though because of the shortage. Great video nonetheless bud 😊
Yes, there’s a lot of people looking to have these fixed but from my limited experience there’s a lot of them that are just burned out.
A few years ago when money was tight and needed a graphics card for my PC , I would look in the junk pile at the used PC repair shop I worked at .
He would let me pick some out and I took them home and bake them and re apply the right amount of thermo paste and hope for the best .
70% of them would work and stayed working for a good while. Now I just buy new ones . lol
Hi Man! that was nice to see. I have a couple points that could help for a further revision of this cards:
1. Usually the mining cards shows kind of a glossy spots on the PCB under the memory chips, main chip and the VRM/mosfets. So it looks like a couple of this ones were used to mining, in wich there are additional spots to check like bios, drivers and memory damage.
2. Try not to test the cards without a brick over the main chip because that temp rise over 120C could kill them even in testing
3. Usually the testing for voltage goes: 12v ext input, 12v PCI, 3.3 PCI, 5v output, 1,8v logic, 1v mem controller, 1.5v memory and 0,9v to the GPU core. usually when all the voltages are ok but the GPU, and the mosfets are ok (no short between the phase inductors and ground), could be the phase regulator or further components there. Sometimes there are issues with the regulators for the memopry controller or the memory chips and those could be quite tricky to check because the fall damage just by touching the wrong pin on testing, but replacements are cheap and could worth the repair.
4. This mining cards like the 3060 could show damage on the memory chips in form of artifacts or green spots on games/benchmarks. There's a tool for nvidia GPU to test the memory damage and usually shows the chip that has to be replaced. The tool also exists for AMD but it's not as fancy
I use to get damaged cards and it use to happens that people just sells them when someone already tried to repair it and goes FUBAR for the cards. So i learned a lot from what not to do with this...
also they might have a no display bios
@@rixas20 that's a good one. I didn't even know about it!
Going into gpu repair without any experience is like going into a sand trap with no one to pull you out. The people who make these things make them so finicky and most need a decent motherboard plus software to remove old gpu software before installing new ones etc. nvm the dozens of other things to think about. Basically if you do one of these videos again stick to trying to repair maybe just one or two decent cards like a 2080ti or 1080 and focus on just those. Sorry that someone convinced you to waste your money on all those. Assuming u already knew all of that well, there’s not much you can do if someone harvested pieces of the circuit board, yikes. Hope one day you have a way to get those pieces and possibly repair a board that was scrapped if that’s even possible. Would be interesting to see. Thx for the upload and GL!
I think the issue with the 3060 might be memory, try using MSI Afterburner to downclock the memory, and try running the benchmark again.
For the 2080ti, I think those components are meant to be missing, as there is a more powerful card, the Titan RTX, which would use that extra VRAM chip and likely needs a bit more power from those components that are missing.
Thanks for keeping things fresh and entertaining.
Got to remember test before re-assembly! For all the other DIYers
That 2080Ti has all the necessary parts on it. The 2080Ti has 11 1GB memory chips but there is space for 12 since the design is shared with the RTX TITAN which has all memory spaces populated with 2GB modules.
Generally speaking running GPUs with the potentially several hundreds of watts that run though the GPU dies without a cooler can be a pretty sure fire way to overheat and kill them it is HIGHLY recommended to not run a GPU on without a cooler attached as it can very very easily damage the GPU die itself as it gets caught with a thermal runaway before it can shut itself down
Also like many others have mentioned the 2080 Ti board was normal that had nothing missing companies use a set PCB design and add parts as they need depending on what SKU it is to save on manufacturing costs since the 2080 Ti is an 11GB card it only needs 11 ram modules
The biggest thing too is making sure to wipe and reinstall drivers between each GPU. Sometimes the Nvidia drivers don’t properly detect new devices until a restart after windows detects them.
@@GhostTurtleYT or just install them full stop
Hi, Great content. Repairs of Nintendo, Xbox, & PlayStation are interesting but I especially like when you purchase other items for repair like these graphics cards. The video on the Go Pro's was also really interesting.
I not sure if it's important. The drivers can be wrong.
When you go to change a graphics card you suppose to with old card in then go into program and features and delete the driver. Then u install new driver.
try to flash the bios on the card at 9:00. might have been used for mining or something and may have a no display bios
I suggest some things to do if the GPU is not sending signal to the monitor:
Connect the GPU to a system with another GPU to see if it is at least recognized in Windows in "Device Manager," if it is detected I suggest you try flashing a VBIOS that you can get from the internet, there are many tutorials on how to do it.
If on the other hand the GPU starts but is very unstable during testing, even there you can try flashing a VBIOS downloaded from the internet to see if the situation improves, otherwise you can try underclocking to find a value where it remains stable
Love the comments on these videos, everyone has been repairing GPUs there whole life and know exactly what's wrong with every single one in the video lol. Great video Steve keep em coming.
the trtx 3060 prolyl is fixable. that series suffered from overheating memory. which also looks possible due to teh artifacts and errors your getting - prolly one of the first or second row of mem ICs has a fault ( could also be just popped solder balls due to thermal stress ). id recommend do a memtest on it and either reball the memory ic /s or get a replacement ( their about 7 bucks rightnow )
Finally someone who knows what is saying...artifacts comes definitely from a bad memory chip, also the discoloration around GPU is probably caused by the neighbor chips gone really hot
GPU prices are falling really nicely. The 1080 is no where near being worth $500 especially broken.
The second-hand Ebay prices are average prices the card sells for if its working. Just to let us know what he could sell it for if it works or gets fixed. Not if is not working
@@ikwilgewoonfilmpje is not even worth that new. 1080ti, which is the better variant, is going for $400, in working condition. A plain old 1080 is not worth $500.
@@FranktheTank319 Got it. Then it is expensive
that one that wasnt putting out a display might have a mining bios on it if the card is being detected but not displaying anything. edit: if you have an integrated graphics on your cpu or have a cheap graphics card and plug them both in, if you plug in the one that works and go to hardware information it should tell you if the one thats not displaying is plugged in or not.
Common issue on the gpu ,on of them are vrms mosfets, vrm controller like 5v 3.3v 1.2-1.3v and 900mili volts to 1 volt, and the most common video ram colder joints and failure u have to replace it with the donor board
Many GPUs (especially higher end 70 and 80 series models) often come with unpopulated VRAM packages and power supply phases.
That is completely normal and is done due to the fact that:
1. They are using a reference design from Nvidia and didn't bother removing unneeded components -> cheaper to design and less work.
2. The actual GPU chip itself has been binned in the factory. Not every GPU chip is perfect. Some might for example come with a few non functional memory channels and sections inside of the GPU that don't work. These large chips are quite prone to manufacturing defects due to their large size, which results in parts of the GPU being broken and useless.
Instead of just throwing these otherwise perfectly fine chips away, Nvidia go ahead and overspec their reference design, so that the manufacturer (ASUS, Gigabyte,...) can just leave the required external circuitry for the broken section out. If for example GPU sections A,B,C,D work fine but section E is defective then it makes no sense so e.g. solder a VRAM chip to section E, since section E is dead. Same goes for the power delivery of that section.
So the reference PCB design comes with more external circuitry than what's actually used.
It's also possible that the same bare PCB is used in multiple GPU versions and they just didn't populate all the chips due to cost saving measures. There's usually low and high end versions of the same GPU. (E.g ASUS TUF vs Rog Strix, Strix being the premium models).
If a GPU chip has too many broken sections to still be considered for example an 80 series card, they disable the extra stuff and rank it as a 70 series card instead. This is binning. It's also how Intel i7,i5 and i3 CPUs work.
Nvidia doesn't make a new chip for every GPU model out there. They only have 3 or 4 "base chips" per generation where they just go ahead and disable features based on the amount of manufacturing defects.
Long story short: That GPU with the missing VRAM chip and the doubled thermal pads probably still works just fine. What I think happened there is that someone installed new thermal pads but didn't pay attention to their thickness, so the GPU chip didn't make any contact with the heatsink and therfore instantly overheated and throttled itself, making them think they broke it.
Another thing I didn't see you do or mention, wiping the drivers. Having multiple gpus and multiple drivers can be conflicting. I would try some gpus again and DDU the drivers.
'Missing' components on the board is a normal thing in the GPU world. It's just a way to reuse the board design for different models for graphics cards. Also, bold of you to turn your system on with a faulty GPU. You can fry the rest of your system that way.
Wow! I think this channel can turn into a multi-repairing channel not only repairing consoles but PCs too. Thanks for posting Steve!
18:55 Look up what the PCB is supposed to look like before you assume someone took parts off. It has 11 memory chips for 11 gigabytes of memory. Also EVGA has several versions of most cards and I don't know if some have more diodes. I think the problem with this card may have just been someone putting so many thermal pads on it that the heatsink didn't touch the chip or paste.
That 2080 Ti wasn't missing anything. That model of it just doesn't come with those components. The same PCB is often used across multiple models. If you do this again, you should look for teardown images of the GPU in question first, which are usually easy to find in reviews.
Please make more videos about GPUs. I learned a lot.
Small tip, with gpus you need to cover the entire Die with thermal paste :) also the EVGA 1080s a lot of then had fail issues. I love your videos. Such a fascinating watch every time ❤️❤️
You just need to put enough on the die that it will spread across the whole thing when pressed against the heatsink. Same with any chip without an IHS over the die, like the consoles he often fixes, or most laptops.
The EVGA 1080 FTW he showed had a flaw that the V-RAM could get to hot. EVGA had a fix by useing a set of thermal pads, which you could get from them. Till today my EVGA 1080 FTW is still working 😁 and i was lucky enough to read about the flaw, otherwise it was also RIP
you could check the serialnumbers on the EVGA Cards if theyre still under warranty. The 3060 could be, open an RMA and send it in and get a new one.
You also could check to try to reflash the VGA Bios on the cards, since they could got corrupted and flashing could fix it
Evga won't touch this now, warranty stickers are voided
@@AnotherDud3 wrong as long the card is not physically damaged. It’s still under warranty. Cards are still under warranty if you change the thermal paste or the cooler.
Maybe send these cards to Tech Cemetery, they are US based and specialise in graphics card repair.
I bet they can get a lot of them working.
1 thing to point ☝️ is you should check the pcie version in bios and match that for your gpu! I know it won't always cause no signal but it might just do that some times
Great videos, try toggle GPU bios switch , use onboard GPU and check device manager
Great video as always! GPUs can be very tricky, but it's interesting watching something different. I hope you can sell the parts and get your money back!
lol, running heaven without a cooler attached. what a mad lad.
On the second card you may want to try to re-float the GPU chip
Fans usually only turn on when the card reaches a temp threshold. Otherwise it will run without them spinning. I almost choked when I saw you boot up without any cooling at all. I then died when you ran a benchmark lol
Not all cards do
been waiting for this video for a long time! im about to watch it. cant wait to see if you fix em!
9:23 u need to unnistall old GPU drivers & reinstall correct one for the GPU u testing.
U can use DDU is to do it.⚠️
try installing the drivers for the cards. That most likely will fix the issue. If not you can try updating the BIOS on the MOBO to see if that will fix the issue
also could try updating the VBIOS on the cards to see if that will fix the issue.
You may try flashing a rom on that dead evga 1080, sometimes corrupt bios can make a card look dead
I have a similar EVGA 3060. I have 3 years warranty. I'm 100% sure that yours should be still under warranty. Before you open, you need to check the serial number on their website. But since you rip the warranty sticker, it's no use then.
I would Keep all the cards you have not fixed, So that once you are familiar with Mosfets and Vram and how to test them when shorted or Open you can replace the
ones that have failed, with both components they should diode test much as each other so when one reads totally different, that's the one that needs replacing along with perhaps one or two supporting components that caused either the Vram or Mosfet to fail..
@UCK0_WC3SrgMhmMCfOQ_Wp0Q
🤦♂Oh REALLY???..🤣
Yeah I don't see this as legit because for 64 plus years I have never EVER won anything.
🧐
I think someone is about to go down for going against what is obviously the universes intent that I never EVER be a winner without entering as a contender..
Hey. What happend to my comment.
You missed a few things.
Glitchy screen is 87% memory issue. Run mats/mods to find out.
Discoloration is okay. The pads will leak oil and other cemicals and it will discolor the pcb even at 60-70°C. Depends on brands.
About the missing components. Nothing is missing. Is is just a reference board. In the same generarion, those spaces will be populated on higher end gpu's. Check google pics.
About the one that does not turn on:
Check the voltage power up line. 12v, 5v, 3.3v, 1.8v pex, vmem, vcore. They have a specific order.
The last thing is you can get mining vbios that disables the output to gain some more hash. Flash them back.
Hey Steve! Nice vids btw, Ive seen in the recent videos you've been playing with the aound modifier and its cool keep doin it, but an idea for you... In a dramatic scene like in this video show a video of yourself doin a sad face as if it didnt work, but dont say it... Think it would be rly funny!
Always amazes me at the money you still need to pay for BROKEN cards or systems. Like they might NEVER work, they should be 80% off.
Many GPU's come with thermal pads. Obviously that card with the stacked pads was modified, But just because the pads weren't cut to size for the individual components doesn't mean its been modified - its a labour cost cutting excercise. Its quicker to slap on 1 thermal pad for 4 chips than to cut them to size for each individual chip.
I had to look up 0 ohm resistor, because there is no such thing. 0 ohms means 0 resistance, so it can't be a resistor. Apparently they are just wires packaged in a resistor-sized package, to make design easier (it's a way to connect components without using exposed wires. Anyway...I wonder if the bad/glitching GPU's were ruined by mining. That explains the 1030 working, that one would be useless for mining. For the 3060, my guess is some of the RAM is bad. Probably died from overclocking (again due to mining).
I wonder if the card with oil on it was highly overclocked and submerged in some type of mineral oil build.