Just ran into the weirdest thing. Insufficient power message

[Keysplayr]

For the past few weeks, I would occasionally get a "Your graphics card is receiving insufficient power. The card will run in reduced power mode to maintain stability. Check external power connectors for the card" <---- Or something like this.

It's a GTX280. Running full speed the 3D Clocks are 602/1296 1100. In the reduced power mode, It would run 400/600 650. Brutally crippling the card. Even COD2 was a challenge.

Anyway, this didn't happen all the time. It used to be, I got the message, then I shut down and removed/reconnected the external power connectors and the problem did not return.
As time went on, I received this message more often. I checked the PCI-e slot, tried it in several PC's. Same problem. I thought the card was fried. Nothing I did worked any longer. The message was coming up 100% of the time now. Card was basically useless.

Last resort and I figured what the heck. I downloaded latest BIOS from mvtech for the GTX280, made a USB boot drive and flashed the GTX280's BIOS.

I was expecting a brick when I was done, but the damn thing booted up properly. No message. Full 3D clocks were back to normal.

There is no explanation other than perhaps this cards original BIOS was a very early one, and maybe didn't like the newer Nvidia drivers.

Anyone else run into this, or know anyone who had the same prob?
If so, pass this on to them.

/share

 

[terentenet]

Keys, I remember I also had this problem. For me it was an early indicator that the PSU was going bad. Video card was OK, but PSU blew up a couple of weeks later.

 

[Keysplayr]

Keys, I remember I also had this problem. For me it was an early indicator that the PSU was going bad. Video card was OK, but PSU blew up a couple of weeks later.

Good to know. One of my first initial thoughts was the PSU in the first rig I encountered this event. I was a bit miffed as this was/is a brand new Corsair 750TX. Then I tried the GTX280 in 2 other rigs. One with an 1100W TopPower, and the other with a 1200W Silverstone. Same event on all systems.

Heh, I even took apart the GTX280 to check the solder points on the external power connectors to see if any were loose or the solder was broken. I also thought that the pins on the cards slot connector may have broken off, but no, there were all there and intact.
Those are the 3 places the card receives power. All seemed fine. All gold pins were there without any breaks or tears. The solder points on the external PCI-e power connectors were solid as a rock.

Side note: Both the 1100W TopPower and the 1200W SilverStone have run QuadSLI (2x GTX295 easily.

After much troubleshooting, I narrowed it down to something wrong with the GTX280 itself. Something internal. When I had the card apart, I inspected every sq. mm for a missing cap or resistor etc. Nothing was Physically wrong with the card. Not a thing.
My very last thought was to try to update the BIOS as it may have either become corrupt, or it was such an early BIOS (My GTX280 is a review sample from NV. I had this before the GTX280 became retail for testing purposes) that it did not handle newer Forceware drivers (somehow). The BIOS flash did the trick though. Much to my pleasant surprise. The GTX280 lives on. hehe.

 

[cusideabelincoln]

Knowing you had such an early sample and a way-more-than-sufficient PSU, a BIOS problem does sound like the most likely culprit.

 

[terentenet]

The PSU that blew up on me was a Tagan 1100W unit. It was stressed enough, for one year with constant load over 700W. Back then I used to have a Single Stage cooling unit in the belly of a LianLi V2000 case. CPU was a HOT QX6700, Intel Confidential, overclocked at 4.2-4.4; I eventually took it down to 4 for stability, then to 3.6 (i guess the silicon went bad after all, even if the CPU max temp was -20 celsius). GPU's were 2 8800GTX's in SLI. I didn't get the error alot, maybe 4 or 5 times, before the PSU blew up in smoke. Mobo died soon after, due to poor head insulation and condensation. CPU died along with the motherboard, RAM too. That was the end of my extreme cooling.

 

[dajeepster]

my gtx280 is exhibiting problems too. For the last two days when the card is under load (gpugrid), I lose video to the monitor. I have to reboot to get display back. As long as i don't put a load on the card, i'm ok. I know it's not a heat problem because i have it on water.

I'll try the bios update to see if that helps.

 

[tweakboy]

This is for sure a PSU issue.

Make sure its seated and you give power to the video card.

If your getting that message, then its your PSU. Your going to need a new PSU or RMA what you have. good luck. Lucky the video card didn't get messed up. Or jump to Frys and buy a proper PSU for your setup and video card... gl

 

[tcsenter]

I think this message is triggered more often by VGA ROM or driver glitches than actual and real power issues. There were a rash of them being reported a couple years ago and IIRC most were resolved by a revised VGA ROM or driver.

 

[Tempered81]

In my experience, flashing to the newest video card and motherboard bioses has always been in my best interest. Same goes for running newest drivers, chipset software, and operating system patches.

 

[Keysplayr]

This is for sure a PSU issue.

Make sure its seated and you give power to the video card.

If your getting that message, then its your PSU. Your going to need a new PSU or RMA what you have. good luck. Lucky the video card didn't get messed up. Or jump to Frys and buy a proper PSU for your setup and video card... gl

I got the message. Repeatedly. Dozens upon dozens of times during my troubleshooting.
And the PSU was not the culprit. I thought I had explained that though.

 

[dajeepster]

my gtx280 is exhibiting problems too. For the last two days when the card is under load (gpugrid), I lose video to the monitor. I have to reboot to get display back. As long as i don't put a load on the card, i'm ok. I know it's not a heat problem because i have it on water.

I'll try the bios update to see if that helps.

i installed windows7 business 64 bit on it and the problem seems to have corrected itself. It had vista ultimate 64bit on it before. Didn't do the bios update... i'm sure I will eventually.

 

[Seero]

Parts do age. While parts perform as specified when it is new, things may change after a while. The original BIOS may consists of a "power check" that is based on the original specification without having degrades in mind, lack of tolerance. The new BIOS may have that particular "power check" specification eased.

In other words, it was a false alarm, and the backup plan (downclock) kicked in too soon and the new BIOS fixed it.

Not that I know this before, just trying to make sense of it. I do know that this happens a lot to motherboard where you can only use a certain clock after flashing bios. Nothing physical has changed, just the "code" that governs it.

Thanks for sharing.

 

[SunnyD]

I got the message. Repeatedly. Dozens upon dozens of times during my troubleshooting.
And the PSU was not the culprit. I thought I had explained that though.

You did. It's just the usual "I didn't read through the entire post/thread" syndrome. Junior Members. Dontcha love em!

 

[dajeepster]

i installed windows7 business 64 bit on it and the problem seems to have corrected itself. It had vista ultimate 64bit on it before. Didn't do the bios update... i'm sure I will eventually.

doh... my problem just crept back in... ok.. time to update the bios now.

 

[Keysplayr]

Don't Brick it!!! Let us know how you make out.

 

[dajeepster]

Don't Brick it!!! Let us know how you make out.

I'll do it tomorrow morning when I get off of work... it's sad too... i have lots of hope for this system as a cruncher. The processor (i860) is going strong at 3.66Ghz under full load at @55C. The gpu and cpu are on the same h2o loop.

 

[Nemesis 1]

I have lost only one PS , But there is something I will never forget about it . The smell of burning paper. Keys probably got it on the bois update.

 

[Keysplayr]

I have lost only one PS , But there is something I will never forget about it . The smell of burning paper. Keys probably got it on the bois update.

I've lost two PSU's that I can recall. But they were the type that came with the cheapy cases that I sometimes buy. In both cases, they would just "give out" and the PC would shut down out of nowhere. I don't think I've ever had a "higher end" PSU give out on me.
::: Knocks on Wood (head) :::

 

[Friendo]

Wow this sucks, ppl who have these cards & doesn't know how to flash bios to videocard and out of warranty support are getting a fuck in the ass by nvidia cause of their failures.
Glad I only paid $50 for my 285.

 

[Keysplayr]

Wow this sucks, ppl who have these cards & doesn't know how to flash bios to videocard and out of warranty support are getting a fuck in the ass by nvidia cause of their failures.
Glad I only paid $50 for my 285.

How eloquent.

 

[dajeepster]

I've lost two PSU's that I can recall. But they were the type that came with the cheapy cases that I sometimes buy. In both cases, they would just "give out" and the PC would shut down out of nowhere. I don't think I've ever had a "higher end" PSU give out on me.
::: Knocks on Wood (head) :::

I've had two cooler master psu's go out on me two weeks ago... they were each only two years old

 

[heyheybooboo]

Just a thought ...

Most of the time I get a consistent, clean 123-125v off my 'computer' circuit. Since the 'cold snap' I've dropped in the 117v range.

I've lived in much worse places, and actually worked for an electric utility in a suburban 'boom' area where during peak power periods we felt lucky to deliver 110v 'per leg' to the home at the far ends of our grid.

So if someone gets fancy with the math and all the assumptions that go with it, you are dealing with the possibility of reduced amperage on a single rail pushing a (65nm?) GTX280 (minimum of maybe 17 amps at stock?) and that could be part of the problem if this has been a recent occurrence.




--

 

[Seero]

Just a thought ...

Most of the time I get a consistent, clean 123-125v off my 'computer' circuit. Since the 'cold snap' I've dropped in the 117v range.

I've lived in much worse places, and actually worked for an electric utility in a suburban 'boom' area where during peak power periods we felt lucky to deliver 110v 'per leg' to the home at the far ends of our grid.

So if someone gets fancy with the math and all the assumptions that go with it, you are dealing with the possibility of reduced amperage on a single rail pushing a (65nm?) GTX280 (minimum of maybe 17 amps at stock?) and that could be part of the problem if this has been a recent occurrence.




--

I would have OC 200% more if i have 10% of your knowledge.

 

[heyheybooboo]

I would have OC 200% more if i have 10% of your knowledge.

LOL --- I'm just old.

PSUs have some fancy voltage regulation - some better and more efficient than others.

Keys has thrown some great PSUs at his problem so, as he noted, this is probably not his specific issue. Others may not be so fortunate.

And my power at the wall just dropped to 116v




--

 

[dajeepster]

ok... i need to flash my gtx280 with new bios ( i think).. well.. anyways... I've never flashed the bios on my video cards before and i'm having a little bit of technical difficulties. I went to mvtech.net website... got the utilities ... didn't work... they are fairly old utilities too.

so... can someone help me with right links and everything.
help is greatly appreciated.

i'm running windows7 64bit