Jump to content

Possibly destroyed VRM?

Flojer0

To begin with I am considering my PGU dead. As background This GPU is a AMD 7870 XT that has lived a hard life. Along with gaming on it I mined on it from early December to late March. Along with that I did have an AIO cooler cable tied to my GPU die. I do have heatsinks on the ram but I have been depending on incidental airflow from the case for cooling the board, couple with the coin mining this probably has a very large share in my symptoms.

 

The story leading to today's symptoms started a week ago. I swapped power supplies in my machine, in the process I needed to remove my GPU and the attached radiator. This caused the pump/cold-plate to move and may have disturbed the thermal paste. After the PSU swap was done the GPU couldn't keep cool. Simple enough, just re-mount the cooler. I get some more cable ties and re-paste the thing, the whole process goes perfectly smooth, so much so I wonder what is wrong as I'm hooking the computer back up. Then my trouble begins.

 

First thing after turning my computer on was starting unengine valley to see where my temps were. The benchmark loads up and starts running, then within seconds my screens go black. This behavior continues for anything that stresses the GPU from 3D games to Luxmark, Fez was usually okay unless the framerate jumped above 60, then it artifacted everywhere. 2D games and normal usage were safe until just recently. I was surfing the web when both my screens went nuts with artifacts. 

 

Lastly I pulled the GPU assuming the worst and saw this when I pulled the board out (Mosfet heatsink removed for visibility):

EDIT: I guess I lost some text before posting, the dark area around the mosfets looks like some liquid leaked out to me.

P_20140715_224609-1.jpg

 

I am currently running on an old 4870 right now, and with my current finances I am going to be stuck with it until I can afford a new machine build anyway. My big question here today is to find out for sure if my vreg is a problem or if there may be some other issue. To me it seems a big coincidence that the card crapped out from a re-application of thermal paste (arctic silver ceramique for those wondering). But I can survive if it is dead.

 

If the card is dead take this as a PSA to not mod/mine on your GPU unless you don't mind it dying. I did get all I wanted out of this card even though I'll be sad to see it go.

My rig: 2600k(4.2 GHz) w/ Cooler Master hyper 212+, Gigabyte Z68-UD3H-B3, Powercolor 7870 xt(1100/1500) w/AIO mod,

8GB DDR3 1600, 120GB Kingston HyperX 3K SSD, 1TB Seagate, Antec earthwatts 430, NZXT H2

Verified max overclock, just for kicks: http://valid.canardpc.com/show_oc.php?id=2609399

Link to comment
Share on other sites

Link to post
Share on other sites

Off topic but how are your heatsinks attached to your ram chips?

Spoiler

Corsair 400C- Intel i7 6700- Gigabyte Gaming 6- GTX 1080 Founders Ed. - Intel 530 120GB + 2xWD 1TB + Adata 610 256GB- 16GB 2400MHz G.Skill- Evga G2 650 PSU- Corsair H110- ASUS PB278Q- Dell u2412m- Logitech G710+ - Logitech g700 - Sennheiser PC350 SE/598se


Is it just me or is Grammar slowly becoming extinct on LTT? 

 

Link to comment
Share on other sites

Link to post
Share on other sites

can't seem to see anything that might be trouble from the picture

 

if you can highlight the trouble area

Budget? Uses? Currency? Location? Operating System? Peripherals? Monitor? Use PCPartPicker wherever possible. 

Quote whom you're replying to, and set option to follow your topics. Or Else we can't see your reply.

 

Link to comment
Share on other sites

Link to post
Share on other sites

Off topic but how are your heatsinks attached to your ram chips?

Thermal tape

 

 

can't seem to see anything that might be trouble from the picture

 

if you can highlight the trouble area

 

The FET in the top right with the snapped leg maybe?

I'm only guessing too. Apart from that, I can't see anything major

Sorry, I lost some text when I accidentally pressed the back button. It looks to me like the mosfets leaked something, but I've never heard of this happening before. Though as explained I am not sure if this is entirely the demise of this card.

My rig: 2600k(4.2 GHz) w/ Cooler Master hyper 212+, Gigabyte Z68-UD3H-B3, Powercolor 7870 xt(1100/1500) w/AIO mod,

8GB DDR3 1600, 120GB Kingston HyperX 3K SSD, 1TB Seagate, Antec earthwatts 430, NZXT H2

Verified max overclock, just for kicks: http://valid.canardpc.com/show_oc.php?id=2609399

Link to comment
Share on other sites

Link to post
Share on other sites

Thermal tape

 

 

 

Sorry, I lost some text when I accidentally pressed the back button. It looks to me like the mosfets leaked something, but I've never heard of this happening before. Though as explained I am not sure if this is entirely the demise of this card.

yea other than the mosfet with one missing lead

 

i think your GPU is toast

Budget? Uses? Currency? Location? Operating System? Peripherals? Monitor? Use PCPartPicker wherever possible. 

Quote whom you're replying to, and set option to follow your topics. Or Else we can't see your reply.

 

Link to comment
Share on other sites

Link to post
Share on other sites

Thermal tape

 

 

 

Sorry, I lost some text when I accidentally pressed the back button. It looks to me like the mosfets leaked something, but I've never heard of this happening before. Though as explained I am not sure if this is entirely the demise of this card.

 

I can't see any leakage, and especially not from the fets.

Capacitors often leak if they go wrong. What components do you think are fets?

Can you highlight them on the photo?

 

EDIT: I see the leakage you're talking about now.

Those look like surface mount capacitors or ICs rather than fets.

Could well be the reason your card is misbehaving, but I can't say for certain

Link to comment
Share on other sites

Link to post
Share on other sites

The FET in the top right with the snapped leg maybe?

 

yea other than the mosfet with one missing lead

 

i think your GPU is toast

 

I can't see any leakage, and especially not from the fets.

Capacitors often leak if they go wrong. What components do you think are fets?

Can you highlight them on the photo?

 

EDIT: I see the leakage you're talking about now.

Those look like surface mount capacitors or ICs rather than fets.

Could well be the reason your card is misbehaving, but I can't say for certain

Photo is highlighted now. I was somehow led to believe that fets were under the vrm heatsink, probably wrong though. Caps make a lot more sense. I was actually looking at the larger caps in the area to see if any of them had leaked but didn't find anything.

 

If it is indeed leaky caps and a snapped mosfet leg then I my guess is the leg got snapped somehow and the electrical issues caused when the card went under load wreaked havoc on those surface mount caps. Meaning that if the card is salvageable then it would need a new mosfet and caps, assuming that is all the damage.  If I could of caught the fet earlier I could have soldered a jumper wire in there. Oh well.

 

Edit: Just looked up mosfets to figure more out and came across this little article:

http://www.techpowerup.com/articles/overclocking/voltmods/21

If we are talking about the same thing then it looks fine to me. *shrug*

My rig: 2600k(4.2 GHz) w/ Cooler Master hyper 212+, Gigabyte Z68-UD3H-B3, Powercolor 7870 xt(1100/1500) w/AIO mod,

8GB DDR3 1600, 120GB Kingston HyperX 3K SSD, 1TB Seagate, Antec earthwatts 430, NZXT H2

Verified max overclock, just for kicks: http://valid.canardpc.com/show_oc.php?id=2609399

Link to comment
Share on other sites

Link to post
Share on other sites

Photo is highlighted now. I was somehow led to believe that fets were under the vrm heatsink, probably wrong though. Caps make a lot more sense. I was actually looking at the larger caps in the area to see if any of them had leaked but didn't find anything.

 

If it is indeed leaky caps and a snapped mosfet leg then I my guess is the leg got snapped somehow and the electrical issues caused when the card went under load wreaked havoc on those surface mount caps. Meaning that if the card is salvageable then it would need a new mosfet and caps, assuming that is all the damage.  If I could of caught the fet earlier I could have soldered a jumper wire in there. Oh well.

ah now i see them 

 

yikes that leakage is really bad for the surface mount caps

 

and yea the mosfet with the missing leg

 

 

if you can ID the caps and the Mosfet used on the card

 

you might be able to use DIY soldering 

 

it may be possible to bring it back to life

Budget? Uses? Currency? Location? Operating System? Peripherals? Monitor? Use PCPartPicker wherever possible. 

Quote whom you're replying to, and set option to follow your topics. Or Else we can't see your reply.

 

Link to comment
Share on other sites

Link to post
Share on other sites

if you can ID the caps and the Mosfet used on the card

I managed to get all the numbers off.

 

The small caps have:

6811

88LW

1236

 

The big caps:

6894

AM82

1232

 

And the mosfet says:

GS7805D

F2WBH

 

Not sure if or when I'll attempt a repair. But it would be an excuse to play with a soldering iron. 

My rig: 2600k(4.2 GHz) w/ Cooler Master hyper 212+, Gigabyte Z68-UD3H-B3, Powercolor 7870 xt(1100/1500) w/AIO mod,

8GB DDR3 1600, 120GB Kingston HyperX 3K SSD, 1TB Seagate, Antec earthwatts 430, NZXT H2

Verified max overclock, just for kicks: http://valid.canardpc.com/show_oc.php?id=2609399

Link to comment
Share on other sites

Link to post
Share on other sites

I managed to get all the numbers off.

 

The small caps have:

6811

88LW

1236

 

The big caps:

6894

AM82

1232

 

And the mosfet says:

GS7805D

F2WBH

 

Not sure if or when I'll attempt a repair. But it would be an excuse to play with a soldering iron. 

keep us updated if you are able to do it

 

head by your electronics store and check if they can find the caps and mosfet with the same spec or just bring the card to let the tech guy ID them for you

 

I will take my hat off to you if you can pull it off

 

cheers

Budget? Uses? Currency? Location? Operating System? Peripherals? Monitor? Use PCPartPicker wherever possible. 

Quote whom you're replying to, and set option to follow your topics. Or Else we can't see your reply.

 

Link to comment
Share on other sites

Link to post
Share on other sites

Thermal tape

 

 

 

Sorry, I lost some text when I accidentally pressed the back button. It looks to me like the mosfets leaked something, but I've never heard of this happening before. Though as explained I am not sure if this is entirely the demise of this card.

 

There's nothing liquid in a MOSFET, and the heat required to melt the materials used would easily melt the PCB away long before the MOSFET

CPU: AMD Ryzen 7 3700X - CPU Cooler: Deepcool Castle 240EX - Motherboard: MSI B450 GAMING PRO CARBON AC

RAM: 2 x 8GB Corsair Vengeance Pro RBG 3200MHz - GPU: MSI RTX 3080 GAMING X TRIO

 

Link to comment
Share on other sites

Link to post
Share on other sites

There's nothing liquid in a MOSFET, and the heat required to melt the materials used would easily melt the PCB away long before the MOSFET

Alright, though apparently I was mistaken and it was capacitors that were leaking. 

My rig: 2600k(4.2 GHz) w/ Cooler Master hyper 212+, Gigabyte Z68-UD3H-B3, Powercolor 7870 xt(1100/1500) w/AIO mod,

8GB DDR3 1600, 120GB Kingston HyperX 3K SSD, 1TB Seagate, Antec earthwatts 430, NZXT H2

Verified max overclock, just for kicks: http://valid.canardpc.com/show_oc.php?id=2609399

Link to comment
Share on other sites

Link to post
Share on other sites

What appears to be leaked liquid around the mosfets is actually just a greasy substance that comes from the thermal pads.

      

Link to comment
Share on other sites

Link to post
Share on other sites

sorry, double post

      

Link to comment
Share on other sites

Link to post
Share on other sites

What appears to be leaked liquid around the mosfets is actually just a greasy substance that comes from the thermal pads.

That was my second thought but this use case was at high risk of overheating. If it is not the vrm then what? Maybe I'll try mounting the cooler again just to see if that was the problem.

 

I also did notice since yesterday that there is some new scarring on the copper plate of the cooler. I don't see any visible damage on the die, is it possible I've damaged the core beyond use without visible damage?

My rig: 2600k(4.2 GHz) w/ Cooler Master hyper 212+, Gigabyte Z68-UD3H-B3, Powercolor 7870 xt(1100/1500) w/AIO mod,

8GB DDR3 1600, 120GB Kingston HyperX 3K SSD, 1TB Seagate, Antec earthwatts 430, NZXT H2

Verified max overclock, just for kicks: http://valid.canardpc.com/show_oc.php?id=2609399

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×