Solved Possible PSU Failure?

[Britton30]

The volts have to connect up to something for current to be present. Still could be a bad GPU as well. The only true way to find a bad PSU is replace it with a known good one.

 

[ICIT2LOL]

Yep mate but I always equate it to the car battery that test up ok on volts yet doesn't have enough grunt to turn the engine and supply enough current to the ignition etc or the AA battery in a torch test ok and the thing fizzles within a second or so.

At the same time I am wondering if perhaps as it is fully modular some of the modular outlets are busy going bad??

 

[Britton30]

The PWR_OK needs only to have the volts present, it powers nothing, rather lets the PSU "know" whether to power up. All within 1/3rd of a second or so.

 

[ICIT2LOL]

The PWR_OK needs only to have the volts present, it powers nothing, rather lets the PSU "know" whether to power up. All within 1/3rd of a second or so.

Thats exactly what I meant mate it is only representative of volts not the available amps
Say you have a piece of equipment that requires 20 gallons of water (amps) per minute at 10psi (volts) and you have a pump (PSU) that can deliver the water at those rates.

Say the the pump becomes inefficient for whatever reason and can deliver the water at 10psi but not the 20 gallons per minute. So therefore you may get for example 10 gallons per minute at but still at 10psi.

 

[Britton30]

In your pump example though the PSU would completely shut down.

 

[ICIT2LOL]

In your pump example though the PSU would completely shut down.

Not necessarily because it is only testing for volts (pressure) not the amps (flow) it can "see" the volts are correct so will let the timer chip fire the machine up but once the amps are not available as the gear starts to draw the amps it requires it soon (or later) reaches a level whereby it does not have enough current to run them and you start to get failures.
I mean this isn't something that necessarily takes place slowly or fast it can be any length of time depending on what is causing the PSU to drop the current.
The drop in amps would also involve dropping the volts (a downward spike just like you would get in a brown out) and therefore the Power_Good signal would drop out That follows for a high voltage spike remembering the Power_Good signal volts have to be between 2.4 and 6 volts for the machine to maintain correct running function. So in theory if there were to be a voltage spike upwards inside the PSU the P_G signal would follow suit and drop out if it became in excess of 6 volts.
I think it is a very clever piece of engineering personally because it has probably saved more motherboards that anything else.

 

[JackALope52x]

Okay so I tested the POWER_GOOD signal today and it would constantly fluctuate between .9V and 1.2V. I don't know if it was user error or the actual power supply itself. I didn't really know what I was doing because this is obviously a learning experience for me, but I did test it multiple ways and the reading was still the same.

Honestly, I will probably just end up buying a new power supply. I am the type who likes to tinker and find the problem and fix it, but this has just been driving me insane and I have run out of patience. Also, my GPU is only a few months old so if that turns out to be the problem, it shouldn't be an issue because I believe it is still under warranty. Just pray it isn't the motherboard because I really am not in the mood to replace it.

Anyways, I just want to thank all of you for your patience and guidance with this long and drawn out process. You guys are the best!

Thanks Again,
Collin

 

[Britton30]

Yeah that looks a little fishy to me. It should be from about 4.5V-5.7V, I think. Be sure to check specs on the PSU. I think ones with a single 12V rail are best.

 

[ICIT2LOL]

Yeah that looks a little fishy to me. It should be from about 4.5V-5.7V, I think. Be sure to check specs on the PSU. I think ones with a single 12V rail are best.

Yep you're right mate the actual text I think mentions a minimum of 2.4v and a max of 6v outside of those means the PSU self test is not right and may produce voltages that would otherwise damage the motherboard. Hence the shut downs.

I am not real sure but the idea behind a dual +12v rail(s) is that one will keep the machine running until the problem is fixed but I stand to be corrected on that it is only as how I read it.

 

[Britton30]

No the total amp output is divided between multiple 12V rails. Not good and a cheap construction.

 

[ICIT2LOL]

No the total amp output is divided between multiple 12V rails. Not good and a cheap construction.

Ok I thought that was the reason So the modular rails Gary are they not a good thing to have and just more of an aesthetic thing?? Myself I like the semi modular ones like the Corsairs I've got?

 

[Britton30]

While my HX850 may have 10 places to plug in the cables, each with a 12V line, it still has but one 12V rail which powers all the connections. Rails themselves aren't modular, it's the cables.

I saw a dual 12V rail, 650W the other day which 12V-1 rated at 20A and 12V-2 at a mere 6A.

 

[ICIT2LOL]

Yep mate the Thermaltake I have in the tester at the moment has dual rails one at 24A and the other 22A but I think the 24 pin is on the larger rail.

Anyway I suppose we have hijacked this thread a bit eh

 

[madcratebuilder]

Voltage is only part of the psu output equation. You have to know what it is putting out in current, or amps. A psu may test with good voltage readings but fail on amp output causing all sorts of problems. You can test for amps with a multimeter but it well require a shunt. Good meters may have a internal shunt capable of up to 10 amps or so, but nothing close to what a pc psu can supply. Having a "test" psu available can save hours of troubleshooting. It doesn't need to be a real expensive high end unit, several good options in the sub $50 area.

 

[Britton30]

doesn't an amp meter need to be placed in the circuit?

 

[Solarstarshines]

Most of the time when there is a TDR in the mix driver crashing and recovering it is a issue with the GPU

I had that same issue for weeks till it came down to the GPU the PSU was fine

I recommend sending the GPU back http://www.sevenforums.com/graphic-cards/224746-6870-display-driver-has-error-every-3-mins-3.html

 

[ICIT2LOL]

Voltage is only part of the psu output equation. You have to know what it is putting out in current, or amps. A psu may test with good voltage readings but fail on amp output causing all sorts of problems. You can test for amps with a multimeter but it well require a shunt. Good meters may have a internal shunt capable of up to 10 amps or so, but nothing close to what a pc psu can supply. Having a "test" psu available can save hours of troubleshooting. It doesn't need to be a real expensive high end unit, several good options in the sub $50 area.

Which is what I have been saying all along Mad and yes I have had some more expensive multimeters that will read up to 20A in DC and 10A in AC. There are to ones that do "clamp" around the cable but are usually quite expensive and not what we would probably want for our purposes because they measure AC current only How does a dc clampmeter work? | Electrotechnik - The Website on Electrical Engineering

 

[JackALope52x]

Anyway I suppose we have hijacked this thread a bit eh

Haha that's not a problem. The more you guys got back and forth, the more I learn.

Most of the time when there is a TDR in the mix driver crashing and recovering it is a issue with the GPU

I had that same issue for weeks till it came down to the GPU the PSU was fine

I recommend sending the GPU back http://www.sevenforums.com/graphic-cards/224746-6870-display-driver-has-error-every-3-mins-3.html

Hey Solarstarshines, thanks for the reply. I was thinking that in the back of my mind but I was trying to convince myself it wasn't. I don't really want to have to return something to Asus and deal with the whole process. If the problem still persists when I replace the PSU, then that will be my next step.

 

[JackALope52x]

Hi All,

Just a quick update. I ordered a new PSU from newegg.com yesterday morning and it should be arriving here sometime today. Again, that's the best thing about living so close to their warehouse, standard shipping feels like overnight shipping. Anyways, I will install it tonight and report my findings. If the problem still exists, I will look into returning the GPU to the manufacture.

Thanks,
JackALope

 

[Indianatone]

Just to clarify something which some of you may not be aware. The power good signal is generated by the motherboard to let the PS know the motherboard is happy with the supplies it is getting, if that signal is out of range the PSU is 99% of the time defective 1% of the time the motherboard is the issue.
Gary I do not agree on you 12v single rail. In the event of a short there is no current limiting and bang goes all your stuff with 100 amps through it. I always use Enermax PSU. I put this baby in my upgrade I have done in the time this thread has been going....ENERMAX ERV1000EWT-G 1000W ATX12V v2.3 / EPS12V v2.92, v2.8 SLI Ready CrossFire Ready 80 PLUS GOLD Certified Modular Active PFC Power Supply - Newegg.com