PDA

View Full Version : BGF 6800GT OC PCI-E complaining about "Not enough Power"

mr.curmudgeon
08-23-05, 01:06 PM
"To protect your hardware from potential damage or causing a potential system lockup, the graphics processor has lowered its performance to a level that allows continued safe operations."

This is the warning I get, which randomly appears whenever I play games that put a serious load on the card. I've got the 6-pin PCI-E power cable connected directly to the card, and I'm running a 550W ANTEC EPS12V power supply. The system idles at 45-50C ambient / 50-60C load. The GPU idles at 45-50C ambient / 68-72C load. I've got the NV5 Silencer hooked up, but I had this problem before when the stock heatsink/fan was connected.

My questions are:

1) Is it a hardware or software/driver issue?
2) Do I call tech support for the PSU, Mainboard or Graphics card?

My motherboard is a Dual-Xeon serverboard which requires (3) power connections from the PSU; a 24-pin connector, +12v 8-pin connector, and a additional +12V 4-pin connector. I've got all three connected...although the 4-pin connector comes off of the 8-pin connector. (That's the way the PSU is built). I had to buy an 4pin-to-4pin 12V extender because once the 8-pin cable is plugged in the 4-pin will not reach the plug on the motherboard. I mention this because I'm wondering if this is the culprit.

I've played HL2 for a whole day without any trouble, but CS:S crashes within 5 minutes?!

I'm looking for the pros here to help me troubleshoot this. I hate calling Tech Support since the communication gap is usually about as wide as the Grand Canyon!!


My system specs:

MB: SUPERMICRO X6DAE-G2 (http://www.supermicro.com/products/motherboard/Xeon800/E7525/X6DAE-G2.cfm) Dual Socket 603/604
CPU: Intel Xeon 3.2 Nocona (http://www.newegg.com/Product/Product.asp?Item=N82E16819117039) Processors x (2)
GPU: BFGTech GeForce (http://www.newegg.com/product/product.asp?item=N82E16814143025) 6800GT OC
DRV: NVIDIAŽ ForceWare 77.77
PSU: ANTEC TruePower (http://www.antec.com/us/productDetails.php?ProdID=22551) 2.0 TP2-550 EPS12V ATX12V 550W
RAM: KINGSTON 2GB (http://www.newegg.com/product/product.asp?item=N82E16820144167) (2 x 1GB) DDR2 SDRAM DDR2 400 (PC2 3200) ECC
HDD: Western Digital Caviar SE 80GB (http://www.newegg.com/product/product.asp?item=N82E16822144122) IDE 7200RPM
HDD: Western Digital Caviar RE 250GB (http://www.newegg.com/product/product.asp?item=N82E16822144179) SATA 7200RPM
HDD: Western Digital Caviar RE 250GB (http://www.newegg.com/product/product.asp?item=N82E16822144179) SATA
hardestofcore
08-23-05, 08:43 PM
I doubt its driver related, Not positive but I'd have to go with that its your wiring.

Sure calling for assistance is always good, could possibly have a bad connection from the power connector to the board itself, never know.
mr.curmudgeon
08-26-05, 10:06 AM
Could it be that my PSU isn't supplying enough amperage on the +12V line?
How much do the 6800 GT OC's need? I think my PSU has 19A on +12V...

mrC
toxikneedle
08-26-05, 01:21 PM
This msg only comes up if your PCI-E power cable isn't plugged into the vid card. It won't give you that msg no matter what the ampage is on the 12v rail. So this means your PCI-E plug isn't giving the card power or you just simply forgot to plug it into the PSU or something.
J-Mag
08-26-05, 02:15 PM
Get the Turbo Cool 510w ASL then forget about your problems.
mr.curmudgeon
08-29-05, 11:05 AM
Ok, there is no way this is a power supply issue. I've tried three different PSU's and I'm still getting the warning. The PCI-E 6-pin cable is directly connected to the graphics card, so there should be NO WAY that it's not getting enough power.

The latest PSU I've tried is a OCZ 600W PowerStream, which I've read will run a SLI 6800 GT system, so again, there's no reason it shouldn't be able to power my single 6800GT.

I've also seen people run the 6800GT on my motherboard (X6DAE-G2), so compatibility shouldn't be an issue.

I'm at a loss?! :confused:

Could it be a hardware defect in the GPU or the Motherboard? Or should I try an earlier version of ForceWare drivers, or something? (I've tried 77.72 & 77.77 versions)
J-Mag
08-29-05, 12:01 PM
The latest PSU I've tried is a OCZ 600W PowerStream, which I've read will run a SLI 6800 GT system, so again, there's no reason it shouldn't be able to power my single 6800GT.

Could it be a hardware defect in the GPU or the Motherboard? Or should I try an earlier version of ForceWare drivers, or something? (I've tried 77.72 & 77.77 versions)

Your board could be extremely picky as SuperMicro says they require 450w minimum and then adding a high end GPU to your machine further compounds the problem. Personally I would avoid dual 12v rali PSU's for your specific situation unless you calculate the load whcih will be applied ot each connector.

However, It could very well be the power regulators on the Video Card itself. This doesn't sound like a driver problem to me, but you never know...
mr.curmudgeon
08-29-05, 12:27 PM
Your board could be extremely picky as SuperMicro says they require 450w minimum and then adding a high end GPU to your machine further compounds the problem.

The strange thing is the motherboard is SLI-ready, and I've read reviews of people using it in a SLI-system w/ two 6800GT's, so it should be able to handle one card, especially on a 600W (+12V@34A) supply. I'm going to doublecheck the load on the 12V line supplying the GPU, but I'm pretty sure I've balanced the load well enough already.

After that, I guess I'll try to take the card into a service shop and have them run it in a different system, just to see if the video card itself has issues.
rohit
08-29-05, 04:44 PM
i think its the vid card. RMA it.
Slammin
08-29-05, 04:59 PM
I've seen that come up when playing with different bioses with my BFG.
J-Mag
08-29-05, 05:05 PM
especially on a 600W (+12V@34A) supply.

According to my understanding the 600w OCZ has a 20amp and 18amp 12v rail, so technically you have 38amps which should be more than enough as long as one rail isn't getting overloaded.

Personally I would just find another mahcine with a 16x PCI-E slot and test it out in that one...
mr.curmudgeon
08-29-05, 07:33 PM
According to my understanding the 600w OCZ has a 20amp and 18amp 12v rail, so technically you have 38amps which should be more than enough as long as one rail isn't getting overloaded.

I think you just answered another question I had...but just to clarify, you're saying the OCZ will automatically "balance" the amperage between to two +12V rails? If so, that'll surely help.

Personally I would just find another mahcine with a 16x PCI-E slot and test it out in that one...

This is a challenge...but it's what I'm hoping to do next. After that it's RMA, as suggested below.


mrC
ricercar
08-30-05, 03:52 AM
Aim a good fan directly at the card and see if the warnings stop.

My 5900 Ultra and NV35GL (workstation 5900U) used to complain about "low power" when they got too hot. Since about 53.03, the drivers pop that dialogue up about power both for a true power issue and when the GPU overheats.
mr.curmudgeon
08-30-05, 09:22 AM
Aim a good fan directly at the card and see if the warnings stop.

My 5900 Ultra and NV35GL (workstation 5900U) used to complain about "low power" when they got too hot. Since about 53.03, the drivers pop that dialogue up about power both for a true power issue and when the GPU overheats.

I have already installed an NV Artic Silencer v5 on the card, and I'm getting 45-50C/idle...I've monitored the card with RivaTuner while under heavy load (CS:Source) and it seems pop up the warning around 70C, which is not really that hot, as far as I understand. I got this same error with the stock fan, too.

At this point it seems like the card is trying to spin the fan speed up a bit more at 70C, thereby requiring more voltage/amps...and something is going wrong at that point.