• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Replacing an old power supply

A

anerfoc

n00b
Joined
May 28, 2009
Messages
34
Hi everybody. I've been at this for a while and I need some help. In advance of building a new PC, I'm about to install a new hard drive into my old single-core computer from way back in 2004. But before I do that I want to and may need to fix a problem that's been going on for a while, and that's the focus of this topic.

Two years ago I installed a GeForce 6600 GT into this computer (Athlon XP 2800+, Asus A7V8X-X mobo, Apex/Allied AL-A350ATX 350W PSU). Everything worked fine, but I promptly got the following message on startup:

"The NVIDIA System Sentinel is reporting that the NVIDIA-powered graphics card is not receiving sufficient power.

To protect your hardware from potential damage or causing a potential system lockup, the graphics processor has lowered its performance to a level that allowed continued safe operations."

At first I was going to fix it, then later when nothing bad happened I just got used to it. It's still going on today, every time I reboot the computer.

Now I'm wondering if it's been incredibly stupid of me to leave it for so long, and I have no idea so I'm asking you guys instead. Do you think, over the two year period, it's resulted in some kind of damage? Hardware damage, file corruption, things like that? Just a guess would help.

I looked at the stickied PSU list in this forum for replacements but I'm really not sure what'll be compatible with my now old computer. Can someone please direct me to a good, more powerful PSU that will work with the computer and that won't give me these problems? Check out the stats on the Newegg page for the current PSU details. Newegg says it has SATA connectors, but all I see are 4-pin molex ones; I have to use a SATA host controller for the HD. Newegg also describes it as AL-A350 but the picture says AL-B350.

Oh, and like I said I'm trying to upgrade the hard drive, so the new PSU should be able to support a WDC Black 1TB 7200 RPM 32MB 3.0Gb/s as well, if this computer even can support that HD.

One other thing. If I do upgrade to a non-error message-causing PSU, will it make the video card more powerful? The 6600 GT's message did say it had deliberately lowered its own performance level.

Any help would be much appreciated.
 
Last edited:
anerfoc said:
Now I'm wondering if it's been incredibly stupid of me to leave it for so long, and I have no idea so I'm asking you guys instead. Do you think, over the two year period, it's resulted in some kind of damage? Hardware damage, file corruption, things like that? Just a guess would help.
Click to expand...

There's a very good chance of that. Apex and Allied do make some spectactcularly crappy PSUs.

anerfoc said:
One other thing. If I do upgrade to a non-error message-causing PSU, will it make the video card more powerful? The 6600 GT's message did say it had deliberately lowered its own performance level.
Click to expand...

Probably. Though to be honest: Whenever I've seen a video card joined with a very underpowered PSU, the card does not downclock itself to lower energy requirements. Then again, most modern video cards probably take up just as much power at low clocks as that 6600GT at full power.

What's your budget for this new PSU? When are you planning on doing a full system upgrade?
 
I'm building a new comp this summer. But I'd like to keep the old one as a secondary. Is there some way I can determine if the key components (CPU, mobo, video card, RAM, hard drive) have been damaged? Some software, or visual cues? And, really important: if I get a new hard drive and replace the PSU but keep using other components the old PSU might have damaged, will that damage the new HD?

I was looking to spend around $50-$100 for the replacement PSU, unless there's a good reason to stretch. It is an old computer after all, and aside from a possible sharing of hard drives I intend to keep the new computer's components separate from the old one's. This was one PSU I considered: Corsair 400W. Although I can't tell if it supports my AGP video card.
 
Last edited:
Did the 6600 GT have a molex plug for additional power???

I don't remember if it did or not???

Look on the back of the GPU, it may be as simple as forgetting to plug in extra power.

I have an old Shuttle (250W PSU) with an AMD 3500+ and a GeForce 6800 GT and the 250W powers that set up just fine.

The Corsair 400W will be a good psu. For a couple more bucks, I'd get this one:http://www.newegg.com/Product/Product.aspx?Item=N82E16817139004
 
Er... yes. It did. I never plugged it in. Never thought to. Very, very good call.

Now I feel really stupid. Oh well. I'll try plugging it in and with any luck the error message will go away.

So should I still upgrade? My previous comments about AGP support and detecting component damage still stand.
 
magoo said:
I have an old Shuttle (250W PSU) with an AMD 3500+ and a GeForce 6800 GT and the 250W powers that set up just fine.
Click to expand...

To be fair, Shuttle PSUs are significantly better quality than an Apex/Allied PSU. So those Apex/Allied PSUs may actually be capable of putting out less power than the Shuttle can.

@ OP
Your AGP card uses a molex connector, which the Corsair 400CX does have. So yes it will still work with your current setup. Some testing that you can do:
[Testing the RAM]
Download Memtest86+ v2.11 or whatever the latest version is, unzip it, burn the ISO file to a CD, and then boot from it, just like you would do with the XP/Vista install CD. Let Memtest+ run for at least three hours on each stick of RAM separately as well as test the RAM all together. Go for a full 24 hours if you want to be completely sure that the RAM is not a problem. If you start seeing errors, than your RAM is defective or you have incorrect settings for the RAM.

[Testing the Hard drive]
Download the CD image of Hitachi Drive Fitness Test, burn the ISO file to a CD, and then boot from it, just like you would do with the XP/Vista install CD. Test the hard drive and see if any problems are found. DFT will run on most manufacturers' hard drives. Alternatively, you can use Seagate's SeaTools for DOS to test a Seagate or Maxtor drive. For a Western Digital drive, you could use Data Lifeguard Tools for DOS to test a Western Digital drive.

[Testing the CPU]
Use Prime 95, OCCT, Orthos or Intel Burn Tool to stress test the CPU.

[Testing the GPU]
Use Furmark or ATI Tool to stress test the GPU. If you see any artifacts, the GPU could be overheating, too overclocked, or faulty.

[Testing the PSU]
Buy and use a digital multi-meter to monitor the voltages coming from the PSU. Follow the instructions for "Using a multi-meter to check voltages":
http://www.bfgpower.com/troubleshooting.html#DMM
 
Well, sure enough the error message is finally gone! Thanks for that. All I can say is that I assumed it was the PSU's fault so much that I didn't properly check the facts. I will still upgrade the PSU though because of the new hard drive.

Thanks also for the recommended damage tests. I'll take them as soon as possible.

Danny Bui said:
Your AGP card uses a molex connector, which the Corsair 400CX does have. So yes it will still work with your current setup.
Click to expand...

Lol, should have looked at the picture. For some reason I didn't expect both AGP and PCI-E support. Question: even though my motherboard doesn't have native SATA support (I use a SATA host controller for the HD), can I use the PSU's SATA power connectors to connect to my optical drives? Currently using a combination of 4-pin molex and I think 80-wire IDE cable.

On the multi-meter: Will this kinda cheap one do? Although, maybe it doesn't matter if the PSU is damaged if I'm going to replace it anyway. Might as well spend the multi-meter money bridging the gap between the 400W and 550W replacements. Unless it would help shine light on whether other components had been damaged?

Also, is there something I could use to check for mobo damage?

By the way, I did a search on Data Lifeguard Tools and was going to download it but got somewhat freaked out by this post.
 
Last edited:
As an Amazon Associate, HardForum may earn from qualifying purchases.
I think it's appropiate to say that if your current computer works..........there was no damage to it.

The nvidia message is textbook stuff. If the GPU needs additional power via PCI-e or molex, you will see that message. The ONLY thing that suffers is GPU performance, the rest of your computer should be just fine.

But rid yourself of that cheap PSU anyway.:D
 
While Dannybui has hit the nail on the head as far as diagnostics, I am inclined to go more with Magoo in that if it is working, its not broken. ( Its digital it will either work or not, for the most part.....)

Since youre going to be doing a new build very soon, I would focus on buying a PSU for the new build, and not worry too much about the old PC. ( unless you wanted to use it for a server and stuff it full of HD , which opens a whole new can of worms)
 
HOOfan_1 said:
Apparently it depends on the board. The one he linked, didn't show a slot for an ATX 12V connector, so I suppose 5V. I wonder if a modern ATX 2.0 and beyond PSU would provide enough 5V power for him.
Click to expand...

The AthlonXP 2800+ uses about 69W on the +5V so the Corsair 400CX with its 130W available on the +3.3V and +5V rail should be enough.
 
anerfoc said:
So, the Corsair 400W would be fine holding up both the GeForce 6600 GT AGP and the WDC Black 1TB? There any particular reason upgrade to the 550W?

BTW, I'm still looking for something to check for mobo damage.
Click to expand...

The 400CX will be enough for that 6600GT AGP and 1TB HDD. No reason to go with the Corsair 550W due to its high price. There are larger PSUs in the $100 price range like the Corsair 650TX and the Corsair 750TX.
 
The two ways through which cheap PSUs damage components are out of spec voltages and ripple. The former is when the voltage goes below or past the limits set on a rail (5%). Ripple is in how far the PSU produces DC power from the AC input, which translates into a waveform if you were to view the rail on an oscilloscope. An AC waveform of sufficient amplitude can and will damage DC equipment if its amplitude is high enough. An amplitude (ripple) of 50 mV on the minor rails and 120 mV on 12V are the limits set by the ATX specification.

Most likely your cheap PSU is violating plenty of limits. I wouldn't want to push it past 50% load out of fear it'll fry things. An upgrade is definitely recommended.
 
HOOfan_1 said:
well the 550W is good for more 5V.
Click to expand...

Does that apply to me? Sorry, as you can tell I'm not exactly a hardware expert.

Elledan said:
Most likely your cheap PSU is violating plenty of limits. I wouldn't want to push it past 50% load out of fear it'll fry things. An upgrade is definitely recommended.
Click to expand...

As regards your 50% comment, I just realized this: I recently plugged a 4-pin molex connector into my video card as magoo suggested to give it additional power. This should mean that the PSU is now under more stress since the video card isn't nerfing its own performance anymore (see the first post in the topic). Should I unplug the molex connector to the vid card until the new PSU arrives to prevent possible damage the PSU might cause now? In fact since I've been performing some video card stress tests I'm a little alarmed now.

By the way, lots of thanks to everybody for their help and various answers to my silly questions.
 
Last edited:
anerfoc said:
Does that apply to me? Sorry, as you can tell I'm not exactly a hardware expert.
Click to expand...
Kind of. Since your system is 5V heavy, the extra 5V on the Corsair 550VX would be helpful. However, as said earlier, the Corsair 550VX has a horrible price at $100, at least on newegg.com. For $5 less than the 550VX, you can get the Corsair 650TX which has even more wattage on the +5V rail available as well as more power on the +12V rail (where most modern parts these days draw power from):
$95 - Corsair 650TX 650W PSU

anerfoc said:
As regards your 50% comment, I just realized this: I recently plugged a 4-pin molex connector into my video card as magoo suggested to give it additional power. This should mean that the PSU is now under more stress since the video card isn't nerfing its own performance anymore (see the first post in the topic). Should I unplug the molex connector to the vid card until the new PSU arrives to prevent possible damage the PSU might cause now? In fact since I've been performing some video card stress tests I'm a little alarmed now.
Click to expand...
Well if the GPU was working fine-ish while the molex was unplugged, unplug it.
 
Sorry but here I am yet again! Ran into a few problems while installing the Corsair 650W.

1. The main power connector for the Corsair has 20+4 pins, 4 too many for my old 20-pin setup. Instructions say the extra 4 should be detached, and I can see the seam where they're supposed to snap or slide off. Problem is, I can't figure how to do it. They don't seem to want to come off. Can I just put the other 20 into the mobo connector with the 4-pack still attached? Here's a pretty good stock pic.

2. Like you guys said earlier, the Asus A7V8X-X (pic) doesn't have a +12V rail so I guess the +5V is what matters instead. But uh, where is the +5V? The PSU doesn't have a +5V cable. Does it just sort of go along with the main connector? The only direct connection from my old PSU to the mobo was the main power connector in the lower right of the pic, the 4-pin +12V was unplugged.

3. The big fan on the bottom of the new PSU (it doesn't come out the back like the old one) blows right onto the CPU, which is only about an inch below it. Is this all right?
 
Last edited:
1) Yes it should detach, but if you are able to have the 24 pin connector go down into the slot and attach without the 4 pin supplement detached, then it doesn't matter

2) Yes the 5V+ that your CPU gets its power from comes from the 20+4 pin ATX connector.

3) The big fan, is an intake fan, not an exhaust fan, so it is actually pulling air from inside the case and pushing it out the back through those octagonal exhaust holes. It is not pushing any hot air onto your CPU.
 
Should have guessed, the intake fan thing makes much more sense.

Okay! Well, that's that then. One small thing: The holes for screws in my case didn't quite line up with the PSU's holes, so it's in there a little crookedly, but solidly. I imagine this happens a lot and there's not much you can do about it.

So, thank you everybody! Still working on the damage tests and looking for capacitor leakage, etc. But I've got a new PSU, and I learned a few things about power supply rails, damage testing and such along the way.
 
Last edited:
Me again. Oh well. After performing flawlessly for a day, the new PSU had big problems.

All right, so when I installed the new PSU I plugged my old WDC SE hard drive into a SATA connector for the first time. With the old PSU it was always plugged into a 4-pin molex. Just thought I should mention it. Other things hooked up to PSU: DVD drive (4-pin molex), floppy drive and motherboard.

The computer had been on for about six hours straight when the monitor suddenly went dead. As it did I seem to remember hearing a sound, probably the regular "schew-hoo" my comp makes when it turns off. The green "on" light on the front of the comp was off. But you could tell by the sound that a fan was still on, inside the case. I turned the computer off, back on, same problem. No light, no image on the screen, fan noise. So something in there was still working.

Unfortunately at this point I didn't bother to determine WHAT was still working (I have 4 fans, one on the PSU itself, one each on the CPU and GPU, and a rear vent fan hooked up to the mobo). I had been working on something and reconnected the old power supply in a panic to make sure the hard drive hadn't been erased.

A few hours later I tried the new PSU again, with what I'm 100% certain was the same SATA connector I had been using before. This time it worked. Green light, monitor, hard drive booted up, everything.

I'm not sure what to do at this point -- RMA? Wait for it to fail again so I can tell which fans, and therefore components, were working? I'm worried about electrical damage to other components if I go on testing.
 
Last edited:
There isn't enough information to diagnose the issue. It could have just been a random fluke. I would just keep doing business as usual, unless it happens again in which case you should pay attention to exactly what you were doing at the time and what changes (if any) you made to the PC.
 
Well, at the time I was watching a streaming video on the net (pretty high CPU usage, and it had been on for a few hours) and typing in notepad.

Really doubt it's a fluke; new PSU, second day? I've never had anything like that happen before, and like I said after I turned it off and on again the problem persisted. PSU aside, I haven't made any changes except from a 4-pin molex to SATA hard drive connector as stated above.

But we'll see what happens next time. So I don't have to worry about damaging other components? Or the contents of my hard drive?
 
Last edited:
anerfoc said:
Well, at the time I was watching a streaming video on the net (pretty high CPU usage, and it had been on for a few hours) and typing in notepad.

Really doubt it's a fluke; new PSU, second day? I've never had anything like that happen before, and like I said after I turned it off and on again the problem persisted. PSU aside, I haven't made any changes except from a 4-pin molex to SATA hard drive connector as stated above.

But we'll see what happens next time. So I don't have to worry about damaging other components? Or the contents of my hard drive?
Click to expand...

There is an extremely low chance of your HDD or other components being wiped or killed. Though try using the 4Pin molex instead of the SATA this time.

And some other tests: if you want to assuage your mind
[Testing the RAM]
Download Memtest86+ v2.11 or whatever the latest version is, unzip it, burn the ISO file to a CD, and then boot from it, just like you would do with the XP/Vista install CD. Let Memtest+ run for at least three hours on each stick of RAM separately as well as test the RAM all together. Go for a full 24 hours if you want to be completely sure that the RAM is not a problem. If you start seeing errors, than your RAM is defective or you have incorrect settings for the RAM.

[Testing the Hard drive]
Download the CD image of Hitachi Drive Fitness Test, burn the ISO file to a CD, and then boot from it, just like you would do with the XP/Vista install CD. Test the hard drive and see if any problems are found. DFT will run on most manufacturers' hard drives. Alternatively, you can use Seagate's SeaTools for DOS to test a Seagate or Maxtor drive. For a Western Digital drive, you could use Data Lifeguard Tools for DOS to test a Western Digital drive.

[Testing the CPU]
Use Prime 95, OCCT, Orthos or Intel Burn Tool to stress test the CPU.

[Testing the GPU]
Use Furmark or ATI Tool to stress test the GPU. If you see any artifacts, the GPU could be overheating, too overclocked, or faulty.

[Testing the PSU]
Buy and use a digital multi-meter to monitor the voltages coming from the PSU. Follow the instructions for "Using a multi-meter to check voltages":
http://www.bfgpower.com/troubleshooting.html#DMM
 
Yep I know, I was doing some of those tests already. Funny, now I'll probably have to do them over again.

Update: I was in the next room at the time, but the person who was actually watching the screen (headphones on) says that there was a loud pop (not at all the normal comp turning off noise) and the screen suddenly flickered to full white, staying that way for maybe a second. Then he thinks he remembers the screen going full black, with a small white dot maybe half the size of a cursor in the center. This lasted for perhaps another second, before the screen went all black.

However I wouldn't be sure of anything except that the screen went bright white and there was a loud pop.
 
Loud pop?

Start sniffing around the power supply. See if there's burning smell or anything like that.
 
There is a slight smell. Sort of like lightweight burning rubber, like if you have a blender with rubber wheels and you run it too fast. After turning the comp off and sniffing up close the smell is much more noticeable, and 20 minutes after shutdown it's still there. It's either coming from the PSU, the CPU or the GPU, I can't tell. In fact, I think it's a combination.

BTW when the comp crashed the GPU was connected to the PSU via 4-pin molex, for almost the first time ever. Right now it's not connected. Since the crash I haven't tested the GPU's supplemental power yet.

However frankly it probably doesn't matter. I did a lot of close-up examination, and I think I've found the problem: this thing has just had it. I finally got my digital camera back from a friend and took some photos, which will probably seal this machine's doom.

GPU:
1: Black smear on the side of one of the cylinders (capacitors?). I think it's been there for a long time.
2: Some sort of residue of around the same area.

CPU:
1: Large amount of weird dirt-like junk inside the CPU that I have noticed but never really cared about for a reason I can't even imagine.
2: Slightly different camera angle.
3: Corrosion or something like it on the exterior that I didn't notice even when putting in the PSU because of lighting and angles on my desk. Seems to have spread over to the mobo, defying gravity.

At this point I seriously think that my best bet is to cash in on that EVGA GeForce 6600GT lifetime warranty, RMA the PSU in case it's been damaged, and buy identical or slightly upgraded versions of the CPU and mobo I already have. Shouldn't cost too much since they're now heavily outdated componentry, if I can find them. That is if it's even worth it.
 
Last edited:
For the CPU and mobo, that's just dirt and dusts that's literally baked/caked onto the system. Get a can of compressed or equivalent and blow out the dust and dirt.

It's not even worth it to get the old AXP CPUs. Just save your cash for a new CPU, mobo and RAM. If that GPU is dead, see if eVGA will let you get a PCI-E card instead of a AGP card.

Oh and make sure that you test the video card in another system.
 
Actually, I originally got the AGP intentionally because I don't have a PCI-E slot.

I don't really have another system to test the video card on since my other one died a while ago. Everybody else has laptops, storebought rigs or Macs.

But I'll give some compressed air a try. Is the smear on the yellow cylinder and surrounding residue on the video card just dirt/dust as well?
 
Last edited:
anerfoc said:
But I'll give some compressed air a try. Is the smear on the yellow cylinder and surrounding residue on the video card just dirt/dust as well?
Click to expand...

The residue yes, the smear not too sure.
 
So at this point is there a way I can tell what caused the crash/smell? I'll try to replace the video card and clean the CPU/mobo. Should I RMA the PSU?

As for buying new modern components, if this comp is still functional and I can get the PSU working without crashes or weird smells, I'd rather save up the money for my new rig.
 
anerfoc said:
So at this point is there a way I can tell what caused the crash/smell? I'll try to replace the video card and clean the CPU/mobo. Should I RMA the PSU?
Click to expand...

Physically remove the PSU away from the PC and smell it at distance to avoid any influence that the smell of the PC may have. If there's a weird burnt smell coming from it, time to RMA.
 
I considered that, but you have no idea how long it took me to screw it in right.

Oh well, I did it anyway. The PSU is definitely the source of the smell, with the GPU and CPU inconclusive. I think they do have a slight odor about them, but possibly just dust, and certainly nowhere as bad as the PSU. RMA time.
 
Last edited:
Argh! As usual, the only thing anyone sells in physical stores is "compressed gas", which will probably do more harm that good. The one I got is "Dust-Off - JR - The Original Compressed Gas Duster." Similar to this, with different packaging. Who knows what it'll do to the sensitive innards of the CPU. Some people have good reports, some bad. Maybe I should get this instead. What's the usual stuff to use?
 
Last edited:
As an Amazon Associate, HardForum may earn from qualifying purchases.
anerfoc said:
Argh! As usual, the only thing anyone sells in physical stores is "compressed gas", which will probably do more harm that good. The one I got is "Dust-Off - JR - The Original Compressed Gas Duster." Similar to this, with different packaging. Who knows what it'll do to the sensitive innards of the CPU.
Click to expand...
Dust-Off is normal compressed air and will not harm any components. I've used it hundreds of times.
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
I thought you would say that. I got a little alarmed after hearing an account of a negative reaction with plastic.

One last thing. Before I RMA the PSU, is it safe to use it to set up (install Windows etc) my new 1TB backup HD? Can't use the old PSU 'cause it's SATA power connectors only. I'd like to make some backups, but I don't exactly want my backup drive to be damaged by an electric shock. It's probably not worth the risk.
 
I wouldn't risk it. You won't be able to use the new drive anyway until you get a replacement PSU, so it shouldn't be too much of a problem for you to just wait until then. The other alternative would be to get a molex to SATA power adapter to use with your old PSU, which is a perfectly valid option.
 
You must log in or register to reply here.
Back
Top