Sign in with
Sign up | Sign in
Your question

New PSU wont output 12v, only getting 9-10v from rail

Last response: in Components
Share
August 24, 2009 4:22:31 PM

Hi,
Last week I purchased a new Corsair TX650W power supply unit for my PC. After taking it out the box, everything looked fine. I took out my old PSU from my case and connected up the Corsair. I did everything the manual tells you to (although in fairness I didn't need to as installing a psu is quite a simple affair). I booted up as normal and my machine seemed to be running smoothly for about an hour. Then suddenly Vista crashed and I heard the hard drives coming to a halt. The fan on my graphics card also stopped. When I rebooted after that I got a Hardware Monitor Error during the POST, asking me to check voltages in the BIOS. The 12v reading was abnormally low. It was fluctuating between 9.40v up to a maximum of 11.07v. Also the reading was shown in red in the BIOS, which indicates there is a problem.

Is this a simple case of a faulty PSU? Or could there be any other causes at play here that are worth considering? Basically it's outputting some power on the 12v rail, but not enough to allow stable functioning of my PC. Also when I put my old PSU back in, everything works again and I get a stable 12.05v reading in the BIOS.

Right now I'm only about 70% sure the new PSU is faulty. I mean if it was completely busted it wouldn't output any power at all, right? I would assume so anyway. Also the site from which I purchased the PSU has this policy where, if you ship an item back to them as defective and they decide it's not defective, they charge £15 + vat and other charges which they do not specify. Then they ship it back to you.

Any suggestions on what to do guys?

a b ) Power supply
August 24, 2009 4:41:00 PM

My guess is a bad PSU. Corsair has good quality but not even they can compleatelly eliminate bad PSUs.

Contact Corsair, they will do an RMA for you. Take pics of your BIOS reading/System and send it to them as proof.

Also can you check the readings with a DMM?
August 24, 2009 5:11:22 PM

You could always test it with a DVOM. That way you know for a fact the voltages that it is giving you.
Related resources
a b ) Power supply
August 25, 2009 1:57:04 PM

First, it certainly is possible for a PSU to develop a fault and still output power that is incorrect. In your case, it appears (according to your mobo BIOS readings) there are two things wrong on the 12v rail - low voltage, and fluctuations. Shabaa's idea is a good one - if you can, verify the BIOS screen readout with measurement by voltmeter directly on the 12v supply line from the PSU where it is plugged into your mobo.If those voltages and fluctuations are correct, there certainly is something wrong with the PSU. In fact, to verify that is is the PSU and not some odd mobo factor, replace the new one with the old one and check the same voltages.

As Shadow703793 said, contact Corsair Tech Support ant give them your story and voltages. If they confirm this is a PSU fault, you'll feel more confident taking it back to the supplier. Or, Corsair might opt to replace it for you directly.
August 25, 2009 2:06:15 PM

Bad PSU, RMA or return.
August 25, 2009 4:12:55 PM

Thank you for your responses guys.

I don't have a multimeter, but I found a cheap one on ebay. Just a cheap pocket digital multimeter, hopefully it should be enough to do the job. I should have it in a couple of days. Also went on Corsair's site and did the form, got an email back saying they are investigating and they will respond as soon as possible.

Just got a question about the multimeter. I've been looking at some tutorials online regarding how to test a PSU using a multimeter/voltimeter. They all have advice along the lines of, "be careful doing this, and make sure you know what you are doing, there is a risk of being electrocuted etc". Which I understand. Are there any particular precautions I should take when doing the tests? Haven't actually done this before.
a b ) Power supply
August 25, 2009 4:27:39 PM

Simple PSU testing:

1. Remove PSU and all connections.
2. Fix a case fan on to a Molex.
3. Jump GREEN (pin 14) on the 24pin ATX connector to any BLACK ground pin. Case fan should spin up.
4. Put the + lead of the DMM to the YELLOW connector on the Molex 4 pin. Put the - lead of the DMM to the BLACK pin on the MOLEX connector. The reading should be close to 12v.
5. Do it to the RED pin also. Reading should be 5v.

http://modtown.co.uk/mt/article2.php?id=psumod
http://pinouts.ru/Power/BigPower_pinout.shtml
August 25, 2009 4:49:23 PM

The question that comes to surface is why did you change the PSU in the first place?
If you have a short to ground somewhere or excessive load it will pull down the voltage.

Step one at this point is to isolate the PSU and test independently like Shadow posted.
a b ) Power supply
August 25, 2009 5:30:32 PM

A computer PSU will NOT electrocute you unless you open up the case and reach inside. There may be higher voltages inside, but the highest voltage available on the wires that come out is 12 volts - same as a car battery. You could put your fingers across the car battery's terminals and never feel the electrical current.

The real danger in testing the PSU is to the PSU itself. If you mistakenly short out two lines with your probes you could damage the PSU. So just be careful to touch only one wire with each probe.
August 25, 2009 6:41:42 PM

I agree with knotknut. If you test the supply and the readings are normal then you will have to test the PSU with one component attached at a time. This testing will isolate a component that is drawing too much or may have a short to ground.
a b ) Power supply
August 25, 2009 7:20:48 PM

Note: It's NOT the VOLTAGE that kills, it's the AMPS that kill.

Quote:
You could put your fingers across the car battery's terminals and never feel the electrical current.

Try that with wet hands.
August 25, 2009 7:36:35 PM

Shadow703793 said:
Note: It's NOT the VOLTAGE that kills, it's the AMPS that kill.


Very true, but at 12v, even with wet hand, you cannot get the resistance of the body low enough to electrocute yourself (which is why you are allowed to submerge 12v light in a bath if you like). So generally you can go by the voltage to determine if it is safe, although technically it is the current that kills.
August 25, 2009 8:50:57 PM

Shadow703793 said:
Note: It's NOT the VOLTAGE that kills, it's the AMPS that kill.

Quote:
You could put your fingers across the car battery's terminals and never feel the electrical current.

Try that with wet hands.

Even with wet hands, 12V isn't enough to drive a significant current through your skin. It's true that current kills, but you need sufficient voltage to overcome the body's resistance before the current becomes an issue.
a b ) Power supply
August 26, 2009 1:07:01 AM

&Wait but doesn't car battery's actually provide any where from 12-15V? And don't thay have like 50-60+A?
a b ) Power supply
August 26, 2009 1:44:36 PM

A car battery is capable of supplying 100 to 150 amps through a very low-impedance load like a starter motor trying to turn over a stiff engine. BUT the current is determined entirely by the voltage available (yes, at low currents the car battery will output about 14.5 volts) and the resistance of the load across it. Even with wet hands to give improved electrical contact to your skin, the resistance across your hand limits the current to a few milliamps. It ususally takes currents of several amps to cause permanent damage to body parts. The most dangerous situation, though, is a current flowing through the chest where it can interfere with normal nerve signals (millivolts!) regulating heart action and cause the heart to fall into fibrillation. This can happen with a through-the-chest current as low as 10 to 30 milliamps. So if you connect to the voltage source with two wet hands you could get a potentially lethal shock from a source as low as 50 volts. But not 12 to 15 volts.

Anyway, back to OP's original problem. Knotknut and Shabaa have very good points to consider even after the PSU itself is checked. Was it overloaded by a faulty component? That needs to be answered before connecting another PSU and potentially damaging it in the same way.
a c 144 ) Power supply
August 26, 2009 2:03:57 PM

Test in stages. Do the PSU with fan load. Safest place to check is one of the drive molex connectors. Remember thois represents a minimal load - less than an amp.

If you get 12 volts, plug in the main power connector (24 pins) and check. If it's low, either the PSU cannot support a 3 - 5 amp load, or there's something wrong with your motherboard.

If you get 12 volts, plug in the CPU power cable. This will represent the first real load - depending on the CPU, 6 to 10 amps. If low, either motherboard or PSU.

Continue with video card and the rest of the components.
August 26, 2009 9:02:44 PM

Ok guys I did the test where you jump the green wire to a black wire on the 24 pin connector. It gives me a reading of only 11.20v to 11.24v on the multimeter. It fluctuates. That is with fan load only.
a b ) Power supply
August 26, 2009 10:13:06 PM

^Yup. I'd say bad PSU.
!