X1950pro problems continued - someone save me please

[Barrichello]

Wondering if someone can give me some advice, as I have lost patience trying to fix my current issue.

Basically it is in regard to my X1950pro card which keeps powering off – it’s a well known issue and there are threads all over the net, I even started one here below….

http://forums.hexus.net/showthread.p...290#post935290

To summarize the above, basically once the card hits around 55-60c (ie in intensive games) the screen goes black – the pc stays on and you have to reboot.

The way to fix this problem seems to involve the VRM chips (as circled in the pic in the thread above) with the default bog standard cooler not cooling them enough. So lots of people seem to have modded their card etc and fixed their issues and I thought about doing this. For a while I continued using my card with the fan at 100% and overall it was ok – only “black screening” every now and then, but it seemed to get gradually worse. The games which seem to cause this most are BF2 and Company of heroes and HL2.

At last though it appeared that my saviour had arrived – not wanting to self mod my card I was pleased to see the new Arctic cooling Accelero X2 cooler (a version of its cooler that was specifically usable for my card - includes a custom heatsink for the VRM) here it is http://www.arctic-cooling.com/vga2.p...0&data=3&disc=

So anyway yesterday I take out my X1950pro remove the bog standard HSF, and carefully follow the instructions to the word, fit the actual hsf fine – fit the VRM heatsink, all seems fine. Card back in system – all looks good (in fact the new cooler seems to keep temps down a fair bit compared to old one which makes sense) but fire up BF2 and after 10mins of gaming – you guessed it “black screen power off” again 

I am now at a loss to what to do. Despite the mod all being perfect that I have done I am now thinking maybe it is something else that is causing my problem? The official thread on my issue on the sapphire forum keeps mentioning insufficient power supplies here - http://www.sapphiretech.com/en/forum...2&page=1&pp=10 although some people specifically went and got more powerful PSU’s but still had the issue. My PSU is surely not the greatest in the world but I am loath to shell out more money to try and fix this issue. Part of me just wants to throw this card out the window and go but a new 8800GTS or something. For info my PSU is as in “my system” sig details. It’s basically a 500w PSU here’s the details http://www.tweaktown.com/reviews/829...ion/index.html I notice that when I opened my pc – the PSU has the following details on it….
+5V +12V1 +12V2 +3.3V
28A 16A 17A 30A

What does this mean and should it be able to power my system? The packaging of my card says 450PSU min with 30A+ on +12V rail.

I will add that yesterday I spend a long time trying different things, I have used a Antec spot fan blowing extra air over my graphic card, I have cleared the overclock of my core duo cpu etc to lessen power usage. I have tried countless versions of catalyst drivers but im still having the “power off black screen issue”. I’m now thinking that after having used my card for so long maybe it’s to late to have put the correct cooling hsf on my card, maybe its perm damaged from black screening from before. I have to say that apart from this issue my pc is absolutely rock solid stable – no other issues. To sum up, I think my options are the following….

1. Give up on this card and buy a new graphic card – I like the look of the 8800GTS.
2. Try a new power supply – although seeing other people tried this in the official sapphire thread and still have the issue

Sorry it’s been such a long thread – but any experience advice would be appreciated as always.

best regards
B

[Clunk]

there is someone else with a similar problem and a similar psu here

http://forums.hexus.net/showthread.php?t=104962

Why dont you just RMA the card and have done with it?

[IAmATeaf]

I'm sure I read that the problem was with a component on the other side of the card overheating and suggested that a small heatsink be placed there, post up a pic of the reverse side of the card and I'll see it it jogs my memory.

Could have been a diff card to be honest so am clutching at straws.

[Barrichello]

there is someone else with a similar problem and a similar psu here

http://forums.hexus.net/showthread.php?t=104962

Why dont you just RMA the card and have done with it?

I was going to rma when i got the card back in dec - but i thought i would try the custom cooler - which hasnt worked so invalidated my warranty now

The PSU is a cheap one so i might just get a better one anway. Any recommendations for an ASUS mobo? I was thinking of this.... http://www.scan.co.uk/Products/Produ...oductID=567861

Enermax infiniti - although as my pc case takes the psu at the top of the case would this be ok? looks like the psu fan should be pointing up for this pic

[staffsMike]

http://www.scan.co.uk/Products/Produ...oductID=446787

this ones highly regarded

[Clunk]

I think the Infiniti has the same size fan as the Galaxy, and if it is the same one, its not silent, so be aware of that. If you have any problems, just ring Enermax and usually they will just swap you one out the following day, which is the main reason I kept this Galaxy.

[Barrichello]

can these power supplies go in the top as well as bottom of a case? My pc is still the old style top loading - can the psu be turned upside down to accomodate?

[Clunk]

Yes, either way, as long as you have clearance for the loom.

[DratUK]

The Infiniti is fine in a top loader. Fan in mine is very very quiet, it only has a low hum if I stick my ear next to the mesh outlet. Being stuck in the bottom of the case its unlikely to run the fan at high speed as it it doesent take in any of the heat from the cpu etc.

Voltages are rock solid and will power any current video card. It also has the next gen 8 pin PCI-E connector and would qiet happily run an R2900XTX whenever they arrive

More details here

http://www.jonnyguru.com/forums/showthread.php?t=1833

Dont be concerned about the Tssst sound they comment on, dont exist on mine. Cables: the 8 pin cpu cable is short but on most top loading cases you wont have a problem.
The rest of the cables have plenty of length.

I would second the rma'ing of the card. If your going to lash out dosh on an expensive PSU like the Infiniti then it deserves a decent graphics card.

[Barrichello]

I would second the rma'ing of the card. If your going to lash out dosh on an expensive PSU like the Infiniti then it deserves a decent graphics card.

Yep in hindsight i should of RMA'ed it from the start, i just gave it the benefit of the doubt that it was just the card itselfs cooling that wasnt up to scratch.

The PSU that i have was only a £40 cheapy, so yeah next purchase i make will def be something like the Enermax infiniti, and then yes a DX10 graphic card i like the look of the 8800gts 320mb cards, not sure if i can stretch to the £400+ ones.

[edna]

Sorry to resurrect this old thread, but I had a slightly different problem with mine which I've now solved, so wanted to spread the word ...

I've also got one of these cards (Connect3D X1950 Pro) second hand and had an overheating issue. Every time the card got to about 73 degrees it would crash my PC.

For me, and I haven't read this anywhere else on the web, I found that when I removed the stock heatsink there were two small (about 5mm x 15mm) bits of material that looked a bit like small plasters. These covered the blobs of white heat conducting putty that went over the three voltage modulator chips.

Looks like these were intended to protect the surface of the putty during manufacture and the factory forgot to remove them. I could be wrong about that, but either way I found that once I removed them and reseated the cooler (using Arctic Silver 5 on the graphics chip itself) my card now performs perfectly.

Never gets above 63 degrees (well, not during Bioshock or Oblivion on max settings, anyway) and doesn't crash at all. Took a few days for it to work properly, which might be to do with the Arctic 5 or the putty (or both) needing to settle in properly, but straight away it would last maybe 10 minutes before crashing rather than the 1 or 2 minutes it took previously. Presumably the Arctic 5 helped reduce GPU temperatures, and that combined with removing the sticker things helped to keep the voltage modulators cooler.

Hope that helps somebody. Saved me the cost and effort of an Accelero cooler.

[IAmATeaf]

Where these plastic strips that you removed then?

[edna]

If you look at the very first link in this post, it links to a page with a picture where the volt modulator chips are (picture shows a blue board, three chips with red circles around them). The white putty stuff was to cover those chips, and the strips were over the putty.