Topic: [WTS] 1x Sapphire 5970 (reference cooler) - 45BTC (Read 2107 times)
When I'm using an air rework station to remove a component, I usually put a square of kapton tape around it. It will help insulate a little and will keep any components that might get loose firmly attached.
Never tried the oven trick. With the heatsink removed aren't there still pieces of plastic that could suffer damage at these temperatures?
Maybe 9 minutes isn't enough to cook them and make them brittle. I'll have to try it next time...
Edit: I was thinking of material used for the DVI/HDMI connectors, not sure what they are and how they react to these temperatures.
Maybe 9 minutes isn't enough to cook them and make them brittle. I'll have to try it next time...
Edit: I was thinking of material used for the DVI/HDMI connectors, not sure what they are and how they react to these temperatures.
yea, the inside of those connectors are usually some sort of a high temp or abs plastic. The metal shielding around them helps a bit in keeping the heat from getting concentrated on any one spot as well. They will of course melt, I'm just not sure at what point.
@OP I read your latest report back. With what "MrTeal" has said about the other components I am not so sure i would risk any longer either. We could build an asbestos or ceramic mold to hold the underside stuff in place and maybe get a bit hotter. But that is probably not feasible for such a project.
Never tried the oven trick. With the heatsink removed aren't there still pieces of plastic that could suffer damage at these temperatures?
Maybe 9 minutes isn't enough to cook them and make them brittle. I'll have to try it next time...
Edit: I was thinking of material used for the DVI/HDMI connectors, not sure what they are and how they react to these temperatures.
Maybe 9 minutes isn't enough to cook them and make them brittle. I'll have to try it next time...
Edit: I was thinking of material used for the DVI/HDMI connectors, not sure what they are and how they react to these temperatures.
Anything rubber will melt, but anything that's plastic can handle the heat because it's only in there for a few short minutes. I was able to plug my DVI dummy plug into both ports with no issues.
Never tried the oven trick. With the heatsink removed aren't there still pieces of plastic that could suffer damage at these temperatures?
Maybe 9 minutes isn't enough to cook them and make them brittle. I'll have to try it next time...
Edit: I was thinking of material used for the DVI/HDMI connectors, not sure what they are and how they react to these temperatures.
Maybe 9 minutes isn't enough to cook them and make them brittle. I'll have to try it next time...
Edit: I was thinking of material used for the DVI/HDMI connectors, not sure what they are and how they react to these temperatures.
aye, maybe a dry solder joint or bad vrm. You could mail it to me for diagnosis and we could split some part of the sale of it. if it can be fixed that is.
So I took the Diamond apart and looked at it. I didn't see anything noticeable, so I took your advice and thought maybe it was a solder problem. I threw it in the oven at 385 F for 9 minutes, and it better! Not great, but better.
I was able to enable the 2nd core and mine with it! Before it would insta-crash. I had them both at 625/300/0.95 (I always undervolt), and it was hashing at just under 600MH/s. It's still a bit flaky, and as soon as it crashes once, you can't mine again until you reboot the pc. This is only after playing with it for about 20 min, so I'm still not really sure. If you think you can work with it, PM me an offer, and we'll see what we can do.
I woulnd't mind playing with it but honestly, as far as fixing it goes, it sounds like it is in better hands with you. As I would not be confident enough to attempt a home oven reflow. =)
You wouldn't want to do a full reflow on a double sided PCB, so be careful about doing it in the oven. If you actually melt the solder the surface tension is usually enough to keep most bottom side components attached, but you could lose the larger ones like a RAM chip if you bake it too long.
No the oven was preheated to 385, and then I put it in there for 9 minutes, and immediately took it out again. I've heard of people saying to cook it at 200C for 10 min, but I was feeling a little cautious, because if it gets too hot, other components can start to get loose like sadpanda was saying. I think 8-9 minutes is long enough to soften the solder inside the chip, but not long enough to really melt anything.
Also, I tried keeping the stock voltages, while still downclocking to 625,300, and it still crashed. GPU-Z shows both VRMs as outputting a consistent amp while mining, with no fluctuations, and decent temp, too. Thus, I"m pretty sure the VRMs are good.
aye, maybe a dry solder joint or bad vrm. You could mail it to me for diagnosis and we could split some part of the sale of it. if it can be fixed that is.
So I took the Diamond apart and looked at it. I didn't see anything noticeable, so I took your advice and thought maybe it was a solder problem. I threw it in the oven at 385 F for 9 minutes, and it better! Not great, but better.
I was able to enable the 2nd core and mine with it! Before it would insta-crash. I had them both at 625/300/0.95 (I always undervolt), and it was hashing at just under 600MH/s. It's still a bit flaky, and as soon as it crashes once, you can't mine again until you reboot the pc. This is only after playing with it for about 20 min, so I'm still not really sure. If you think you can work with it, PM me an offer, and we'll see what we can do.
I woulnd't mind playing with it but honestly, as far as fixing it goes, it sounds like it is in better hands with you. As I would not be confident enough to attempt a home oven reflow. =)
You wouldn't want to do a full reflow on a double sided PCB, so be careful about doing it in the oven. If you actually melt the solder the surface tension is usually enough to keep most bottom side components attached, but you could lose the larger ones like a RAM chip if you bake it too long.
aye, maybe a dry solder joint or bad vrm. You could mail it to me for diagnosis and we could split some part of the sale of it. if it can be fixed that is.
So I took the Diamond apart and looked at it. I didn't see anything noticeable, so I took your advice and thought maybe it was a solder problem. I threw it in the oven at 385 F for 9 minutes, and it better! Not great, but better.
I was able to enable the 2nd core and mine with it! Before it would insta-crash. I had them both at 625/300/0.95 (I always undervolt), and it was hashing at just under 600MH/s. It's still a bit flaky, and as soon as it crashes once, you can't mine again until you reboot the pc. This is only after playing with it for about 20 min, so I'm still not really sure. If you think you can work with it, PM me an offer, and we'll see what we can do.
I woulnd't mind playing with it but honestly, as far as fixing it goes, it sounds like it is in better hands with you. As I would not be confident enough to attempt a home oven reflow. =)
Have you tried reseating that Diamond? You know pulling the heatsink and reapplying the thermal compound? Did you happen to check the idle temps on it?
It crashes my driver as soon as I even try to mine on core #2. When I got the one core mining, fan was set at 50%, and the good core was ~55C, and the bad core was ~45C (this was according to CGMiner). I tried enabling the bad core at 600,300,0.950 and it still produced the same results - it crashed as soon as I enabled it.
Edit: the symptoms do not point at an overheating or temp issue.
aye, maybe a dry solder joint or bad vrm. You could mail it to me for diagnosis and we could split some part of the sale of it. if it can be fixed that is.
So I took the Diamond apart and looked at it. I didn't see anything noticeable, so I took your advice and thought maybe it was a solder problem. I threw it in the oven at 385 F for 9 minutes, and it better! Not great, but better.
I was able to enable the 2nd core and mine with it! Before it would insta-crash. I had them both at 625/300/0.95 (I always undervolt), and it was hashing at just under 600MH/s. It's still a bit flaky, and as soon as it crashes once, you can't mine again until you reboot the pc. This is only after playing with it for about 20 min, so I'm still not really sure. If you think you can work with it, PM me an offer, and we'll see what we can do.
Have you tried reseating that Diamond? You know pulling the heatsink and reapplying the thermal compound? Did you happen to check the idle temps on it?
It crashes my driver as soon as I even try to mine on core #2. When I got the one core mining, fan was set at 50%, and the good core was ~55C, and the bad core was ~45C (this was according to CGMiner). I tried enabling the bad core at 600,300,0.950 and it still produced the same results - it crashed as soon as I enabled it.
Edit: the symptoms do not point at an overheating or temp issue.
aye, maybe a dry solder joint or bad vrm. You could mail it to me for diagnosis and we could pslit some part of the sale of it. if it can be fixed that is.
I'm thinking I'll take the cooler off and take a look at it tonight when I get home from work.
Have you tried reseating that Diamond? You know pulling the heatsink and reapplying the thermal compound? Did you happen to check the idle temps on it?
It crashes my driver as soon as I even try to mine on core #2. When I got the one core mining, fan was set at 50%, and the good core was ~55C, and the bad core was ~45C (this was according to CGMiner). I tried enabling the bad core at 600,300,0.950 and it still produced the same results - it crashed as soon as I enabled it.
Edit: the symptoms do not point at an overheating or temp issue.
aye, maybe a dry solder joint or bad vrm. You could mail it to me for diagnosis and we could split some part of the sale of it. if it can be fixed that is.
Have you tried reseating that Diamond? You know pulling the heatsink and reapplying the thermal compound? Did you happen to check the idle temps on it?
It crashes my driver as soon as I even try to mine on core #2. When I got the one core mining, fan was set at 50%, and the good core was ~55C, and the bad core was ~45C (this was according to CGMiner). I tried enabling the bad core at 600,300,0.950 and it still produced the same results - it crashed as soon as I enabled it.
Edit: the symptoms do not point at an overheating or temp issue.
Have you tried reseating that Diamond? You know pulling the heatsink and reapplying the thermal compound? Did you happen to check the idle temps on it?
Update: The Sapphire is still for sale, but I've also got a Diamond 5970 that is being sold for parts/repair. It shows up, and I've gotten it to mine on with core 1, but it was pretty flaky. It's a reference style blower cooler, with a backplate. Fan works great. PM me with offers.
I'm surprised this hasn't sold yet. He's only asking $30 more than I got for a for parts/not working 5970 on ebay a couple months ago.
I'd like to purchase it, but anything coming out of European Union is subject to a 30% tax for anything above ~$60. Customs can even overrule the estimated value if they think it is too low.It's OK if I purchase coffee or maple syrup (the risk isn't big) but not good for graphic cards (which are meant to mine enough to pay for electricity, sometimes fan replacement and resale at lower price point).
I'm shipping from the US
I'm surprised this hasn't sold yet. He's only asking $30 more than I got for a for parts/not working 5970 on ebay a couple months ago.
I'd like to purchase it, but anything coming out of European Union is subject to a 30% tax for anything above ~$60. Customs can even overrule the estimated value if they think it is too low.It's OK if I purchase coffee or maple syrup (the risk isn't big) but not good for graphic cards (which are meant to mine enough to pay for electricity, sometimes fan replacement and resale at lower price point).
I'm surprised this hasn't sold yet. He's only asking $30 more than I got for a for parts/not working 5970 on ebay a couple months ago.
Bump.
Bump.
Bump: It runs at 600+MH/s undervolted, or 750MH/s at stock Volts.
CGMiner 2.4.1 couldn't OC w/o a dummy plug (Win7x64, 12.4, SDK 2.5), so I will include the dummy plug if you want/need it.
Jump to: