[SOLVED] Sapphire R9 270x Dual-X Black Screen? Here is how i fixed mine.

[ronelz1101]

Hi Guys,

So as the title says, i managed to fix my Sapphire R9 270x Dual-X graphic card.

I've done so much research on the problem that i had finally decided to take matters into my own hands and try to fix it by all means.
And wanted to poste my results so it might help others even though thats and outdated graphic card, i know there are still people holding on to one of these.

Things i tried before getting to this point:
Test older drivers - NOK
Flash GPU Bios - NOK
Read all existing threads on the internet about the problem - NOK but learned a few things.

So i started to disassemble the card to check the PCB and place it on alluminium foil paper.

I started by cleaning the board, all the components and their connectors with, believe me or not, acetone ( nail polish remover ) as i would normally had use Isopropyl Alcohol, but there no such thing in the country i live. (well i looked everywhere but couldn't find any)
But sill Acetone did the trick.
I applied it on a smooth tooth brush and even dropped some on the board and brushed it smoothly but thoroughly until the connecters/solders were all clean and shiny.

after that (believe it or not too) i used a hairdryer to reflow the PCB. (my heatgun was borrowed indefinitely) but that still did the trick. So hairdryer on max temp with a little bit of flux on the board to remake the solder. (keep in mind that a heatgun would do a better job for that and will take far less time.

Keep the board fixed while doing this on one side.
let it cool down and do the same on the other side of the PCB.
10/15 minutes reflow on each side with 5 minutes cool down in between.

Afterwards, re-apply thermal paste to the GPU and re-assemble card!
TEST.

Side note:

My card died right after test the latest Adrenaline 2020 driver 19.12.3 so i DDU'ed the drivers and tested the Adrenaline 2019 19.12.1 which works fine with the card (without any tweaks or overclock) haven't test any overclock as it's working fine since i revived it.

So let me know how it goes on your side if it's the last resort and you wan't to try the fix.

Cheers.

 

[delaro]

A hairdryer doesn't get hot enough to reflow solder. Typical hair dryers get between 40-50 °C, the melting point of "Soft" Solder starts around 180 to 190 °C. At best you loosened up the protecting coating and when it hardened something decided to make contact, I can't imagine that it will last long.

Most people would have probably tossed the card since it is a little slower than a 1050 GTX but uses much more power and outputs more heat.

 

[Gam3r01]

Agreed, that would not have created conditions to reflow the card, and even if it did "baking" of graphics card is not a fix, but rather a temporary solution to an underlying problem that will resurface later.

 

[ronelz1101]

Totally agree on that! i almost fried the dryer keeping it on max but board heated enough for the soldering points to shine again.
But i did keep the dryer steady on one side of the board so as to accumulate more heat than the dryer should be able to. does that make sense?

What made me do this experiment is the old YLOD of the ps3 which the fix was to heat the GPU and CPU enough to realign them but still the permanent fix was reballing them. So i heated the GPU as well!! but on the PS3 i used to heat 'em up from 250 to 400 degrees.

So far the graphic card is still working after a month but i can tell that the latest amd driver killed it.
I'm still hesitating to trying the adrenaline 2020 edition as it was the one that killed it.
I will OC stress it when i get my heatgun back to reflow properly so to make sure it doesn't die again.

But fact is, this could be pointing out to the main cause of the drivers killing the card.
a possible component overheat and i'm dying to find the culprit of this massive graphic card murder! if it's not the GPU itself.

 

[delaro]

An OEM driver isn't going to fry a GPU, messing with 3rd party BIOS files will. If your running a stock BIOS file then your limited to a capacity of voltage that is well within the limits of the card. I'm not sure what you're suggesting there but no conspiracy exists to fry out outdated GPU's.

 

[RememberThe5th]

An OEM driver isn't going to fry a GPU, messing with 3rd party BIOS files will. If your running a stock BIOS file then your limited to a capacity of voltage that is well within the limits of the card. I'm not sure what you're suggesting there but no conspiracy exists to fry out outdated GPU's.

But I've fixed with different bios version an GTX 960?950? and it still works till this day. (Ofc techpowerup GPU bios collection)

I've fixed my current motherboard with dish washing detergent and drying out, re flowed bent pins and it still works and im using it right now.

Fixed also GT 9800 GX2 (with industrial heat gun and flux) but I've killed it after I tried to see how much does my PSU drop voltage in Furmark.

Acetone is good thing tho as isopropyl alternative.

The main issues that happens to those GPU's is using them, then turning off PC, moisture gets under "Somewhere" and "Bumps" the components off the BGA array grid.

 

[delaro]

But I've fixed with different bios version an GTX 960?950? and it still works till this day. (Ofc techpowerup GPU bios collection)

I've fixed my current motherboard with dish washing detergent and drying out, re flowed bent pins and it still works and im using it right now.

Fixed also GT 9800 GX2 (with industrial heat gun and flux) but I've killed it after I tried to see how much does my PSU drop voltage in Furmark.

Acetone is good thing tho as isopropyl alternative.

The main issues that happens to those GPU's is using them, then turning off PC, moisture gets under "Somewhere" and "Bumps" the components off the BGA array grid.

*Shudders"Your methods....errrrr..

1. Tap Water is bad to clean electronics with, it leaves deposits.. lots of it in places you can't easily get to regardless of the soap you use. If your using water it should be distilled or de-ionized.

2. VLSI components are safe to immerse in clean water, but SMT electronics are not and guess what? You have both on motherboards and GPU's.

3. Even with a seal coat, a PCB is permeable and dunking it in water gets it in places you don't have access to like under the trace leads and the chipsets themselves, so unless your heating it out or shaking it with an ultrasonic cleaning station you're leaving yourself open to problems.

4. Acetone is a horrible product for cleaning with, it can dissolve the seal coating along with Flux which leaves you prone to possible corrosion or in the worst cases an Arc. Acetone also can dissolve the minuscule layers of many of the chipsets themselves because they are refined using plastic. 😔

But I've fixed with different bios version an GTX 960?950? and it still works till this day. (Ofc techpowerup GPU bios collection)

You reflashed a corrupt BIOS... that was pretty common on those cards and it has nothing to do with the drivers.[/QUOTE]

 

[RememberThe5th]

1. Why is still working after 1 year?
2. My gpu was soaked in water with Detergent and still working.
3. Im not Soaking components for 3 hours, im washing it.
4. I used acetone cleaning my CPU pads of 775, ram pads etc.

About corrupt bios, someone on tech said that wrong bios might damage the GPU, I guess they are also wrong.

My tap water is soft and drinkable, idk about the rest of the world, maybe because I've used hot water it didnt damage my components.

 

[delaro]

About corrupt bios, someone on tech said that wrong bios might damage the GPU, I guess they are also wrong.

Yes a wrong BIOS can mess up your GPU and nowhere in this did anyone say otherwise.

1. Why is still working after 1 year?
2. My gpu was soaked in water with Detergent and still working.
3. Im not Soaking components for 3 hours, im washing it.
4. I used acetone cleaning my CPU pads of 775, ram pads etc.

Well If you cleaned it with Urine once and it worked would you still clean it with Urine and expect the same result? If you strapped 1000 bottle rockets to your feet and managed to fly up without burning your legs off does that mean the next time you would get so lucky? If you drove down a busy highway once blinded folded and didn't get into an accident does that mean you could do it again with the same result?

Do you get my point?

Just because you can or did something doesn't mean it's the right way or going to work every time. When your maintaining or repairing things you take steps to minimize the risks and in a world where all the information is at your fingertips why are you even arguing about this?

My tap water is soft and drinkable, idk about the rest of the world, maybe because I've used hot water it didnt damage my components.

Do you understand the difference between Tap water and distilled water? I'm thinking you have no clue.

If you doubt what I am saying then do this.

Take a tub of tap water, drop a 9V battery in and test the Conducitity.
Do the same with Distilled water.

Do you see a difference? Yes, you do that is what a 3 grade science project?

Take a pot and boil tap water completely out.. what is left in the pot?
Take a pot of distilled tap water and completely boil it out.. what is left in the pot?

That white stuff that is left... it's conductive. The more you have the greater likelihood you might have issues cleaning sensitive electronics. Remember what I said before about minimizing the risks? Why take the risks in the first place when it's pretty easy to just use the right products and be safe all the time.

 

[ronelz1101]

delaro
Thanks for the informations provided on what you should and should not do.
I've already learned a lot from your posts.

Not talking about conspiracy here but just that fact that there's "something" in the latest drivers that triggered "something" on the card that killed it! Could be a compatibility issue with older amd graphic cards (software/hardware not thoroughly tested to ensure maximum stability with all older models) but i see too many similar cases online for it to be a coincidence. specially for the R9 200 series/300 series and even some RX 400 series. So in my opinion, Adrenaline pushes those cards to some kind of limit they are not supposed to have.

As the issue that still remains to this day even with the latest drivers, the famous screen flickering which the card started having more frequently before it died.

I've used amd products long enough to know the weakness of their products.

As for the bios i only flashed stocks ones! no mod....YET.. i'm not into OC'ing until i'm sure it can maintain stability on stock.

Another thing i noticed is that, one fan working and the other one on standby..
That ZeroCore Power technology that comes certainly with the latest amd softwares.
That card didn't have that feature when it used to be on the legacy drivers (to my knowledge). I wonder if there's a way to disable this feature and see how it performs.

 

[RememberThe5th]

Yes a wrong BIOS can mess up your GPU and nowhere in this did anyone say otherwise.



Well If you cleaned it with Urine once and it worked would you still clean it with Urine and expect the same result? If you strapped 1000 bottle rockets to your feet and managed to fly up without burning your legs off does that mean the next time you would get so lucky? If you drove down a busy highway once blinded folded and didn't get into an accident does that mean you could do it again with the same result?

Do you get my point?

Just because you can or did something doesn't mean it's the right way or going to work every time. When your maintaining or repairing things you take steps to minimize the risks and in a world where all the information is at your fingertips why are you even arguing about this?



Do you understand the difference between Tap water and distilled water? I'm thinking you have no clue.

If you doubt what I am saying then do this.

Take a tub of tap water, drop a 9V battery in and test the Conducitity.
Do the same with Distilled water.

Do you see a difference? Yes, you do that is what a 3 grade science project?

Take a pot and boil tap water completely out.. what is left in the pot?
Take a pot of distilled tap water and completely boil it out.. what is left in the pot?

That white stuff that is left... it's conductive. The more you have the greater likelihood you might have issues cleaning sensitive electronics. Remember what I said before about minimizing the risks? Why take the risks in the first place when it's pretty easy to just use the right products and be safe all the time.

Motherboard which you wash does not have active voltage nor current if its propetly discharged, you are talking same as using live voltage line/cablr and dropping it into water vs dropping a piece of cable which isnt connected to anything.

Then prove me, why isnt my motherboard dead yet? Why isnt my 5 year old gpu dead yet? Hmm?

If i did these things for years and still do not have any issue nor corrosion (Cauee im not peeling off the coating, the metals ARE COATED!) The acetone does not remove right away all 10 layers of insulation between copper layers, im not pouring 10 liters of it and leave it onto it to dissolve it, damn there is time vs effectiveness for all materials.

Same as in my degree the Polyurethane foam wont dry in next few days but it takes weeks.

Oh and last thing, why is der8auer washing mothrrboard in dishwasher to wash off the vaseline? Oh no corrosion will happen.

 

[Deleted member 14196]

lol, this thread is pretty funny....

 

[delaro]

Motherboard which you wash does not have active voltage nor current if its propetly discharged, you are talking same as using live voltage line/cablr and dropping it into water vs dropping a piece of cable which isnt connected to anything.

Then prove me, why isnt my motherboard dead yet? Why isnt my 5 year old gpu dead yet? Hmm?

If i did these things for years and still do not have any issue nor corrosion (Cauee im not peeling off the coating, the metals ARE COATED!) The acetone does not remove right away all 10 layers of insulation between copper layers, im not pouring 10 liters of it and leave it onto it to dissolve it, damn there is time vs effectiveness for all materials.

Same as in my degree the Polyurethane foam wont dry in next few days but it takes weeks.

Oh and last thing, why is der8auer washing mothrrboard in dishwasher to wash off the vaseline? Oh no corrosion will happen.

I don't think you read a thing I typed. 🤣I think even if I did go into detail on the how and the what you would still argue with me. Feel free to do as you please just don't toss up guides promoting products and procedures that an entire industry avoids.

There is a reason why any place that does restoration, refab and repair uses only three primary products.

Distilled Water
Soap " Designed for this application"
Isopropyl Alcohol in the pre-clean process.

This is what is used on a professional level, yes it's quite similar to a dishwasher but it uses precious products to clean with.

How to Use an Ultrasonic Cleaner to Restore Electronics and PCBs

There are a lot of things you seem to not understand about PCB's and how corrosion affects them.

Ultimate Guide to Cleaning Electronics

Going further with this would be like trying to do shop talk with my wife, I know I'm going to get a crossed eyed stare and a clueless expression.

 

[RememberThe5th]

Im not promoting, im saying that it can be done, plus thing is im bastard at it cause it did not fail me yet.

About that "Indian guy" over techpowerup said that different bios can burn up that "output driver controler" (like duuuh).

Usually laptops are repaired since they aio. (If you spill water or whatever on closed circuit on voltage vs my talk i do without voltage.)
Like Louis Rossman group.

To say to your 3rd grade test we did not have (yeah welcome to 3rd world country).
But my water is drinkable from tap. And when my brother did boil the water it had minimal. (for the kid ofc.)

I agree and disagree, as i said ive been there and done it and its still working. When it dies and ill inspect "why". If it was due tap water i wont do it again. But as it still lives on ill keep doing it. (maybe the flux that was beneath chipset is stuck soo somewhat saves the component).

Why would you buy an equipment for it just for some sideproject? Im not doing it everyday to sell the repaired stuff, im not refurbisher, im doing it for fun, buy broken part, repair, keeping it.
Like ive done to motherboard, ssd, psu.

Btw you said to drop battety into water, not an electrical part without any voltage or current whatsoever, im not that idiot

 

[ronelz1101]

Hi Guys,

So as the title says, i managed to fix my Sapphire R9 270x Dual-X graphic card.

I've done so much research on the problem that i had finally decided to take matters into my own hands and try to fix it by all means.
And wanted to poste my results so it might help others even though thats and outdated graphic card, i know there are still people holding on to one of these.

Things i tried before getting to this point:
Test older drivers - NOK
Flash GPU Bios - NOK
Read all existing threads on the internet about the problem - NOK but learned a few things.

So i started to disassemble the card to check the PCB and place it on alluminium foil paper.

I started by cleaning the board, all the components and their connectors with, believe me or not, acetone ( nail polish remover ) as i would normally had use Isopropyl Alcohol, but there no such thing in the country i live. (well i looked everywhere but couldn't find any)
But sill Acetone did the trick.
I applied it on a smooth tooth brush and even dropped some on the board and brushed it smoothly but thoroughly until the connecters/solders were all clean and shiny.

after that (believe it or not too) i used a hairdryer to reflow the PCB. (my heatgun was borrowed indefinitely) but that still did the trick. So hairdryer on max temp with a little bit of flux on the board to remake the solder. (keep in mind that a heatgun would do a better job for that and will take far less time.

Keep the board fixed while doing this on one side.
let it cool down and do the same on the other side of the PCB.
10/15 minutes reflow on each side with 5 minutes cool down in between.

Afterwards, re-apply thermal paste to the GPU and re-assemble card!
TEST.

Side note:

My card died right after test the latest Adrenaline 2020 driver 19.12.3 so i DDU'ed the drivers and tested the Adrenaline 2019 19.12.1 which works fine with the card (without any tweaks or overclock) haven't test any overclock as it's working fine since i revived it.

So let me know how it goes on your side if it's the last resort and you wan't to try the fix.

Cheers.

UPDATE
So for people still interested in giving a second life to their somewat dead amd r9 which died after the lastest amd driver installation.

After a few months of the hairdryer and acetone fix, card is still running with some minor issues but running and gaming.

What i noticed:
Frame drops in games after 5 minutes of gameplay.
After monitoring the temp and coreclock,
Temp was rising to 80c 90c and core clock was stuck at 450mhz with 99% gc usage.
It was just thermal throttling. Adding some artic silver thermalpaste fixed that and i got to install the adrenaline 2020 4.2.2.


After fixing the thermal throttling, what i noticed is that it auto-underclocks after 5 minutes ingame but not much and games are stable.
With a clock starting at 1070 mhz and goes down to 1020mhz i don't see or feel much of a difference while gaming.
Last game tested was Shadow of War on high settings with an avg of 44 fps.

Another issue is the black screen when the screen goes on standby, the graphic card won't resync with the display until i've unplugged and replugged the hdmi cable.

So, the base starting issue could either be the gpu itself that faulted or the memory chips but still don't understand how the driver killed it.

I'll test a complete reflow of the board with professional equipments and flux tomorrow to see if i can get rid of the rest of the issues.

I'll post the results.