Hardware

Ask something
Question Locked

PSU problems, no video output

0 helpful
11 replies

by martin

Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.

I've seen several questions like this, but my experiences so far may provide some more detail for this specific issue.

The problem has many facets. Primarily I will lose video output (monitor goes black) during or a while after startup. Sometimes my PC won't properly boot at all giving a long-short-short-short beep code. Other times it will give the OK beep just to hang a moment later giving out continuous short beeping. During any of these problems my PC does not shut down.

The weird thing is, once I have the PC running it will continue to run for hours unless I really stress the hardware (by playing certain games).

My first instinct was to suspect the video card (GeForce 7900 GTX). However my PSU (Antec 450W that came with the Sonata II) was also giving a high pitched sound which made me decide to return it. I received a new one and the problem was gone.

Several months later I tried the Crysis demo and my screen went black again. Since this moment, all issues described above are happening again consistently.

Again I'd suspect the video card if it wasn't for the fact that the symptoms are exactly the same as before and that replaced PSU fixed everything.

Has anyone experienced problems like this? I have seen a lot of people complaining about the Antec PSU's, but I can't help feeling that there is more to it.

I will buy a new PSU of another brand tomorrow and update you with the results.

PS I'm pretty sure it's not a temperature issue. The cpu, case and video card never get near critical temperatures, even after heavy stress.

System:
Intel Pentium D 3.0 GHz
Asus P5LD2 Deluxe
Geforce 7900 GTX
Maxtor 300 GB SATA II 7200 Diamond Max 10
Antec True Power 450W PSU
2GB RAM (Corsair VS2GBKIT667D2)
Windows XP Mediacenter SP2

11 replies

by mjd420nova

For that system and the video card, I don't think 450 watts is big enough. It should be over 500 watts to get reliable operation. Some mfgr's only give the peak power specs for their power supplies. Look closely at the specs and select one with at least 500 watts continous ratings.

by thechas

While I would agree that your power supply is on the low side for your system, I think your primary problem is too much heat.

Check the chip-set fans and make sure they run all the time. If there are no chip-set fans, consider installing a set.

Asus used to have a nice system monitor utility on the driver CD. If you have it, install it and see what the reported system temperatures are.

I typically install 2 case fans that blow air into the case, and rely on the power supply fan for exhaust.

For your system, I would consider adding in a slot fan to assure that the video card had a source of fresh cool air.

You need at least 2" of free space behind the PC for the fans to work at removing hot air. More if the case is in closed area.

Chas

by martin

I figured that I may need a more powerful PSU. However, this very system has worked fine for several months on 450W, shouldn't I have issues at a consistent rate if it was the wattage?

As for heat. I have been monitoring my system on and off. Also with Asus Probe and Geforce's temperature graph. The system is blowing straight into an open space and the 120mm rear fan is keeping the air moving at its high setting. I doubt that it's the problem, but I'll double-check that.

by rahouseholder

Electronics degrade over time. Heat is the biggest problem. The more strain you put on your power supply, the hotter it will get. The components in it will begin to fail. They will not always fail catastrophically right off the bat, but their ability to perform will be degraded. This can cause intermittent errors where the computer will fail but if you turn it off for a few hours, it will run fine for a while again. I see this on a daily basis. I test electronics in an oven at 300F and a lot of times it will fail at that temperature but run fine at ambient.

by thechas

300 degrees F!

What kind of electronics are you testing?

That is 149 C.

It is no wonder you are finding (creating) failures.

Most commercial electronics components are only rated for up to 70 C (158 F)

125 C rated parts are about the best you can find in Automotive / Military grade parts. And they are darn hard to come by anymore.

I work for an avionics company, and we very seldom test anything active above 70 C. We may perform limited "storage" condition testing at elevated temperatures as part of design qualification. But, those parts NEVER get used in a plane.

Testing at 300 F, you are damaging and or degrading every part you test. Even those that survive will have a shortened service life.

Chas

by martin

All the hardware is little over 1 year old and the airflow in the Sonata II case has always kept the temperature at acceptable levels. Even the video card wouldn't come near critical temperatures under heave stress. I've had temperature problems with previous machines, so I've been careful not to repeat that.

However the degrading hardware problem you mentioned may apply as the PSU has been powering the hardware badly for several weeks, which can't be a good thing.

by martin

Well, I've just bought and installed a Zalman 500-HP PSU to replace the Antec 450W PSU.

Just after installation the machine booted instantly and did so the rest of the night. No more black screens or beep codes.

The next morning the machine seemed to boot fine but without video output (no beep codes however). Sub sequential bootups went fine, there was video output, everything.

Not a good start but I'll see where this leads.

by martin

Well there's a brand new Zalman PSU in the machine but it hasn't changed the situation.

All the video problems are still there. I have switched the GeForce 7900 GTX with a 8600 GS which seemed to please the PC. I wasn't able to reproduce the problems with that card.

I'm sending the GeForce 7900 GTX back and hope they can confirm the problem and get me a new one.

I have also been monitoring the temperatures closely. In idle state the CPU is 45C, the case is 30C and the card's GPU is 36C. Playing a recent FPS game raises the GPU to 55-60C. Nothing out of the ordinary I guess, no weird spikes at all.

by martin

Well a conversation with the nVidia support desk later it turns out that it wasn't the hardware, but the driver.

Although I was told this was not a documented problem, my card in combination with my system and the latest certified drivers were causing these problems.

Recent drivers as well as drivers I used several months ago caused these problems. When I uninstalled and cleaned my current driver properly and I installed the beta driver, all problems were gone.

I couldn't have guessed this was it. I didn't expect the drivers to have influence beyond the OS.

To anyone that experiences loss of video output I highly recommend either installing beta drivers, or older driver versions to see if the problems go away.

by rahouseholder

Sorry it took a while to reply. I work in the oilfield service industry. Our tools are designed to work at those temperatures and a little hotter. We have some that operate at 185C and a little above. The pressure underground causes things to heat up and our sensors have to be able to operate reliably.

This discussion is closed to new replies.