Silverstone Strider ST75F cutting out

[bluemarvin]

I've got this PSU which worked fine until my new graphics card came along - a Palit GTX275. Installed OK and works fine with ordinary desktop apps - but anything requiring graphics grunt (games or video stability test) causes immediate power trip, blackscreen & re-boot. Given the PSU has quad 12v rails rated at 54 amp total, it should be able to handle the GPU.

The PSU has 3 pots accessible from the side, one of which adjusts the voltage (I tried maxing it today - to 12.6v according to the BIOS - but no effect).

It has 2 other pots which I think control protection sensitivity - which looks like it is the problem (?) - can anyone confirm this or offer any other advice?

Silverstone manuals are really just spec sheets and don't mention the pots so I'm floundering here - no idea how they should be adjusted, & trial & error has failed so far.

 

[JN]

Although its not very common, some PSU's have compatibility issues with certain types of graphics cards. May sound a bit crazy considering its all 'volts' but it does happen. I'd try the graphics card in another machine if possible to see what happens.

 

[bluemarvin]

Jim - thanks very much for that. I did resort to sending the card back to the supplier who said it worked fine in their test system. The trouble is I don't really want to swap out what should be a perfectly good PSU. I understand what you say about compatibility - could be down to something like transients, hence my focus on the PSU's protection circuitry - just annoying I can't find out how to adjust it!

 

[JN]

Jim - thanks very much for that. I did resort to sending the card back to the supplier who said it worked fine in their test system. The trouble is I don't really want to swap out what should be a perfectly good PSU. I understand what you say about compatibility - could be down to something like transients, hence my focus on the PSU's protection circuitry - just annoying I can't find out how to adjust it!

One thing you could try is powering the graphics card from molex connectors using some cheap adapters off ebay. Its not ideal but it may prove/disprove the compatibility theory.

 

[Datamonkey]

Hey up,

Just wanted to throw something else into the fire here for you to try,

I don’t know the psu in question myself, but had a quick scan of the review on this very site

"2x PCI-E cables, each with 2 PCI-E connectors attached to give the unit a total of 4x PCI-E connectors for Quad-SLI compatibility"

So there are 2 modular cables with 2 pcie connectors per cable, again I’m not familiar with the Palit GTX275 but I’m gonna have a wild stab that it also has 2 power connectors.

My question is are you using the one cable? Or 2 separate cables? The reason I ask, is its possible the GTX275 could be pulling more juice the psu could supply down the one cable?

I know with my Bequiet 1200w psu it recommends using connectors from seperate rails to the graphics card. So in your case plugging in both cables and using only 1 connector from each cable.

Edited to tidy up lol im tired.

 

[bluemarvin]

thanks for that - already been there I'm afraid. Initially tried using one split cable but then tried using 2, one from each output on the PSU - same problem. As I said above, I also tried tweaking voltage to no avail - hence I'm now suspecting protection sensitivity.

I've just followed Jim's advice and ordered a couple of molex/PCI adaptor cables off e-bay - either to use with this PSU or, for testing purposes, with an old PSU I have lying around

 

[bluemarvin]

Strider cut-out

One thing you could try is powering the graphics card from molex connectors using some cheap adapters off ebay. Its not ideal but it may prove/disprove the compatibility theory.

Jim,

just thought I'd let you know - powering from Molex adaptors worked in that it powered the card - but in terms of the problem all it did was enable the PC to run the video test for 3 sec rather than 1 before blacking out!

I'm onto both Silverstone & Palit - but having spent 3 whole days trying to get this card working, I'm thoroughly

 

[bluemarvin]

Datamonkey -ta - Silverstone actually came up with the same idea - but still stuffed I'm afraid

 

[JN]

Jim,

just thought I'd let you know - powering from Molex adaptors worked in that it powered the card - but in terms of the problem all it did was enable the PC to run the video test for 3 sec rather than 1 before blacking out!

I'm onto both Silverstone & Palit - but having spent 3 whole days trying to get this card working, I'm thoroughly

Yeah I'm not surprised ur peed mate. Out of interest do you have a multimeter? One last test (if you can face it) would be to hook up the multimeter to the back of one of the PCI-E connectors thats plugged into your GPU and then see if the voltages dip significantly when entering 3D mode/games.

Let us know how you get on contacting manufacturers tho.

 

[Rastalovich]

Current would be the issue rather than voltage, imo.

 

[bluemarvin]

Yeah I'm not surprised ur peed mate. Out of interest do you have a multimeter? One last test (if you can face it) would be to hook up the multimeter to the back of one of the PCI-E connectors thats plugged into your GPU and then see if the voltages dip significantly when entering 3D mode/games.

Let us know how you get on contacting manufacturers tho.

OK - sad update time.

Tried multimeter - voltage flickered from 12.06 to 12.00 before crashing - I'm assuming that would be normal for the GPU kicking in.

I've tried all combinations of PCI-E & Molex converter cables.

Silverstone now want pictures of all my connections and still won't tell me what the other 2 potentiometers are for.

Palit keep on plugging away, even analysing my event viewer log for me - but are now resorting to advising a new clean Vista 64 install :banghead:

 

[JN]

OK - sad update time.

Tried multimeter - voltage flickered from 12.06 to 12.00 before crashing - I'm assuming that would be normal for the GPU kicking in.

Yep perfectly fine. Obviously there could be other issues with your PSU that aren't quite so easy to detect, but to be honest I'm leaning towards it being an issue with the card now.

I've tried all combinations of PCI-E & Molex converter cables.

Silverstone now want pictures of all my connections and still won't tell me what the other 2 potentiometers are for.

I'm not so sure about your model, but normally there are one or two pots for adjusting the minimum fan voltage and what temperature it 'ramps up' at. Half the time Silverstone probably won't even know what they are for (as the units are built by an OEM) so that could well be why they're not saying.

Palit keep on plugging away, even analysing my event viewer log for me - but are now resorting to advising a new clean Vista 64 install :banghead:

Ugh, best thing would be to tell Palit you've tried the card in another machine and it has exactly the same problem. Get it RMA's then see how you get on with the new card.

 

[bluemarvin]

Took my card to be tested - ran fine on a 500W PSU in another system!

My next steps look like twiddling the pots on my PSU or trying an old PSU - but the only one I have to hand has a 4-pin 12v ATX lead rather than the 8-pin lead on my new one - not sure it can power the mobo.

(I have tried a completely clean install of the 186.18 nvidia drivers, but no joy)

Bingo! - bodged up a 300W no-name PSU I rescued from the skip - and the video test worked! I've no idea why the Strider should trip but it's clearly to blame.

Update 2 - RMA obtained from Scan & PSU sent off for exchange:beerchug:

Final update - replacement PSU arrived yesterday - worked fine with GTX275 (and even seems a bit quieter than original)