No boot and random shutdown

[sleepydog]

Hi all,

First post here, I found the site looking for my problem... which is, well, the title says it all I guess. The PSU is a Liteon PS-6351-1DFS from a Dell Dimension 8400, I saw it mentioned already in this forum. It started as random shutdowns, then the PC wouldn't boot at all unless the PC was unplugged and left to rest for some time (few minutes to hours). The PSU behaves exactly the same on two other PCs: first time in it works beautifully, very silent, even under heavy load (2x Prime95 on Athlon X2 @ 2500MHz and P4 Northwood @ 3GHz). Then, the next day, nothing. I actually haven't tried it, but I suspect that if I left it unplugged when the PC is off, it might start ok (is it worth a try? would it mean anything?)
Inside, everything looks perfect: all clean, no leaks, no bulging capacitors, nothing suspicious.
The original PC has a new PSU which works ok, but by now it's a battle between me and the damn thing, I really want to see if I can fix it!

[starfury1]

Re: No boot and random shutdown

Welcome
caps can look find and still be stuffed

you will need to do some checking with a multimeter you have one ?

there is a thread by stevo somewhere who has the same type of psu and it sounds like a similar, same type of fault
one repaired by KC8 was track to a faulty cap...replace all thought is the best fix
mainly it is the DC outputs caps that fail.

Another possible area of failure is the Standby 5V circuit

the fault description is typical of failed caps.
so voltage measuring time and cap checking time.

I'll find the thread for you later

Cheers

[acstech]

Re: No boot and random shutdown

Some caps do fail without bulging or anything. Might be worth it to recap anyway.

To be honest, I really wouldn't mess with it. I would replace it and be done with it. What's the wattage on that thing anyway? Standard ATX form factor or something different?

If it's standard I might just get one of these to replace it:

http://www.hypermicro.com/product.asp?pf_id=PSAB001

Keep in mind that has a 24 pin ATX connector, and an 8 pin P4 power connector. You can use it on motherboards with the 4 pin P4 power connector if you have enough room around the connector; just plug in the appropriate side of the 8 pin connector.

[sleepydog]

Re: No boot and random shutdown

Welcome
caps can look find and still be stuffed

you will need to do some checking with a multimeter you have one ?

there is a thread by stevo somewhere who has the same type of psu and it sounds like a similar, same type of fault
one repaired by KC8 was track to a faulty cap...replace all thought is the best fix mainly it is the DC outputs caps that fail.

Another possible area of failure is the Standby 5V circuit

the fault description is typical of failed caps.
so voltage measuring time and cap checking time.

I'll find the thread for you later

Cheers

Thanks, but don't go look for it, I got it, here:

https://www.badcaps.net/forum/showth...ghlight=liteon

and here:

https://www.badcaps.net/forum/showthread.php?t=4929

stevo1210's problem does indeed sound like mine, but my PSU doesn't make any funny noises and it doesnt bootloop: it either starts and then randomly shuts down (moments or hours later) or just stays off.
The 5V standby should be ok: there is an LED on the mobo that comes on.

Update: just tried again (PSU off and out of any PC for the last 2 days), and nothing happened (except the LED on the mobo is on). Shorted green and ground on the 24-pin plug and the fan didn't come on (two drives connected as load). Did it just die completely???

I do have a multimeter (super-cheapo thing but it should do I guess), but I have no idea what to do, any help is appreciated. When you talk about the DC caps, or anything else mentioned in stevo1210 threads I just have no idea which you mean (I can recognize a capacitor, but which one is which is beyond me).

[starfury1]

Re: No boot and random shutdown

not sure on the fuuny buzz thing either could just be a noisy coil or something.

OK you got them good

Cheers

[PCBONEZ]

Re: No boot and random shutdown

Funny buzz is often +5vsb problem if it uses the flyback transformer design.
Bad flyback xfmer circuits will cause the windings to vibrate and it can sound like a hum, squeal, whine, or even a whistle.
Bad caps (among other things) in +5vsb section can do that.
(This was apparently the cause of the noisy Antec SP and TP PSUs, they used GP caps that were underrated -and- crap brand.)
-
Other symptoms also suggest +5vsb as a possible problem.

Pay close attention when replacing Dell PSU's.
Some used the standard ATX plug but with a proprietary non-standard pin-out.
Same plug as everyone else but the wires were rearranged. (For no good reason.)
[I think yours is not one of those but check anyway.]

.

[stevo1210]

Re: No boot and random shutdown

stevo1210's problem does indeed sound like mine, but my PSU doesn't make any funny noises and it doesnt bootloop: it either starts and then randomly shuts down (moments or hours later) or just stays off.
The 5V standby should be ok: there is an LED on the mobo that comes on.

Yeah, I still have the same issue with my PSU.
Now it doesn't bootloop but instead it's doing the exact same symptoms as yours. It will randomly shut down whever it feels like.

For me it also depended on the motherboard I used it with. If I used this PSU with my brothers system which is based on an Intel DP965LT motherboard, it will bootloop, but with an Asus P4P800 series board, it will just turn off and act dead. I currently have it in my Asus PC.

Update: just tried again (PSU off and out of any PC for the last 2 days), and nothing happened (except the LED on the mobo is on). Shorted green and ground on the 24-pin plug and the fan didn't come on (two drives connected as load). Did it just die completely???

I've that happen to me as well. I thought it was dead.... but I left it out unpowered for a few hours and it started up again.

Thanks.

[starfury1]

Re: No boot and random shutdown

(For no good reason.)

bloody good reason for them
you have to go back to them to get basically a standard psu at 3 times the price! (with the plug wired differntly)

or even better you use a standard one and it all goes BOOM

Thats if memory serves me

there is a similar trap with some Apple PSU's

yeah 5+stand by runs 24/7 if psu is plugged in and turned on (at the back).
This is as pointed out "again above by bonez" a very good possibility for issues if the psu has been running for a long time.

so I think a recap of DC outputs wont go astray regardless
but theres possibly a more subtitle fault with this.....and possibly crap caps of lower value

There is an older thread somewhere were it was discussed in a more in depth manner.
(dont think it came to a final resolve...)


this is sounding like a common fault so there is probably a common type cure...sorry but I don't know what it is.

you will need to do some detective work on the psu

Cheers

[starfury1]

Re: No boot and random shutdown

The comment posted by PCbonez on DELL using different connector pinouts is very valid and something to take note of and watch out for with older Dells
that is my reply to that comment as to probably why...
(with a bit of tongue in cheek)

I think this particular model in the thread is a standard pinout one
if Stevo has been running it ok on different MB's

just so we are clear on this

[Bookworm]

Re: No boot and random shutdown

I have a Dell power supply adapter - it's made by a third party company, and lets you put in a standard power supply.

Mind you, I ALSO found out that they made those power supplies mirror imaged. So if you try to fit it in, you find that they set the screw holes so that the fan portion of a standard power supply would be against where the plug is supposed to go. That required a hacksaw and file to fix.

[hardwareguy]

Re: No boot and random shutdown

There was also an ATX like connector on really old Dell PSUs that did not have 3.3V..... the leads were switched up a bit and 5V was in place of the 3.3 lines. It's easy to spot..... just look at the colors.

Then there was the otherwise ATX supply with the leads switched around...... that's the one that pisses me off.

[sleepydog]

Re: No boot and random shutdown

For me it also depended on the motherboard I used it with. If I used this PSU with my brothers system which is based on an Intel DP965LT motherboard, it will bootloop, but with an Asus P4P800 series board, it will just turn off and act dead. I currently have it in my Asus PC.

I've that happen to me as well. I thought it was dead.... but I left it out unpowered for a few hours and it started up again.

... and in fact today the little bugger started again .
I also agree on the fact that some machines seem to have more problems than others: my guess (totally unsubstantiated by any hard fact) is that for one reason or another the PSU can't cope with the current demand, so it either fails to start or shuts down, and of course different machines put different loads on it.
Whether this is a problem fixable by changing capacitors or something else... I have no idea.

starfury1: YES, it is a completely standard ATX 24 pinout.

Also, which ones are the DC capacitors? There's quite a few, some almost inaccessible, and I honestly don't know a thing about PSUs.

Bookworm: It's not the PSU: size is standard and the screw holes are in the standard places, it's THE CASE that has metal in the PSU area, so that only PSUs with no switch and the socket in that position will fit (there are many, not only this one). Standard cases have a large hole to accomodate ANY standard PSU. Anyway, nothing that half an hour of sawing couldn't fix (maybe more, try doing that by hand in such a tight space!).

[PCBONEZ]

Re: No boot and random shutdown

Like I tried to say,
The no restart indicates a +5vsb problem.
-
Many PSU manufacturers have a bad habit of using under rated caps there so it's something you want to look at.

As you haven't worked on them before don't go touching heat sinks if you power it with the cover off.
- They have power. - OUCH!

.

[starfury1]

Re: No boot and random shutdown

To add to what pcbonez has said
THEY ARE DEADLY these things DIRECTLY rectify the Mains Voltage (there is no isolating step down transformer like in linear supplies) and chop up DC Voltage at around the 340V mark so yeah all "Cautions" are advised......

Those main caps can hold a nice belt too even when power unplugged
(normally they will discharge but a fault could leave them in a charged state)

I think bonez is on the money here but I don't know the psu so no idea how its put together.

Wish I had the "change Cap X Y Z or Resistor A B C diode whatever for you but I dont

for a brief description read peteS FAQ on how they work
I would start there to can an idea of the basic areas of the PSU and what there purpose is.
General PSU Discussion and FAQ contributions

the DC output caps are the caps used across the DC supply rails
(part of the filter circuit)

So there will be 1 for 12V one for 5V one for 3V3 etc
(does depend on actual design)

The replacement caps must be low ESR types with good ripple rejection Rubycon ZL panasonic FC (FM) ETC.

the 5V SB rail is powered differently So its somewhat a different/ separate circuit to the main supply setup
here is another few links.

On the DC output caps Area "E" in the photo is were they will be located on that photo. (if you look at the PSU BUILD Quality photos thread you will see most have the same very basic layout areas.

http://en.wikipedia.org/wiki/Switched-mode_power_supply

for a more in depth read

http://www.hardwaresecrets.com/article/327

you will find a few schematics on this page but don't think that yours is going to be exactly the same cause there are many ways to build the same thing
and although some may follow a basic layout the details can change greatly.

http://www.smps.us/computer-power-supply.html

You will need to back trace the 5V+ SB rail and see what caps are hanging off associated circuits but while your at it you may as well recap it
(DC output caps on all rails.)

Stevo has already stated his 12+ is floating around to start with.

hopefully it is just a purely a crappy cap fix but I don't know for sure.

Be very careful like I said these do have a potential to kill you.

Make sure you check you work!,
a mistake with these can lead to fireworks.

Stevo you done anything yet with regard to yours?

HTH Cheers

[stevo1210]

Re: No boot and random shutdown

Stevo you done anything yet with regard to yours?

Yeah, thanks for asking. I haven't done anything to fix it yet because for some reason it's fine again. Also could be because I've got it in a low end Celeron system....
But sooner or later, I will have to do a cap replacement or similar on it because Iw ant it to power my P4 Prescott system.

Thanks.

[starfury1]

Re: No boot and random shutdown

Risky
as you say its possibly light load but still there is something amiss with it
I'd defiantly at lest recap the DC output side of things.

Cheers

[sleepydog]

Re: No boot and random shutdown

Hi guys, thanks all for all the suggestions / info.

I'm stressing the PSU right now on yet another PC (old athlon Xp 2200+... I really should start to throw away some stuff, at least P3 era components!! ). The 12V is, according to the monitoring utility, a bit out of spec: 11.264V, without fluctuations. 5V is 4.86V, 3.3V is 3.28V, so within 5%. With a meter, 5VSB was 4.85V and 12V is 11.4V (I don't know what to trust more, the crappy multimeter or the internal probe), 5V also was slightly more, all very stable (but again, I'm not using an oscilloscope). Other wires are hard to reach (it's a shuttle), but I'm happy reading the data via windows.
Still no crash, but then the PSU passed the same torture test on an Athlon X2 2.5GHz. Let's see if anything happens.

Bad capacitors should give unstable output, right? But a low voltage like I'm seeing on the 12V? Bad PWM regulator? Bad transistor (nah...)?

[PCBONEZ]

Re: No boot and random shutdown

The crappy meter will likely be more accurate.
The sensing built into a motherboard is pretty weak/simple.

By the meter everything is within 5% and in spec just barely.
5% of 12v = 0.6v

As all outputs are consistently low and not jumping around I think it's an overall regulation problem and not bad caps in so far as ripple filtering goes.
(Although the filter caps may be bad too.)
-
It's probably just due to some component(s) shifting value with age.

.

[sleepydog]

Re: No boot and random shutdown

As all outputs are consistently low and not jumping around I think it's an overall regulation problem and not bad caps in so far as ripple filtering goes.
(Although the filter caps may be bad too.)
-
It's probably just due to some component(s) shifting value with age.

You explanation makes perfect sense, and I am not at all happy with those low voltages. The PSU is 2 years old though, Dell must have specified it incorrectly for it to go off spec already. Summers are pretty hot and humid in Japan (where the PSU originally was), but hey, still too early!

Anyway, it's done it again!!

Stress testing ok for a couple of hours, constant temperature (just 42°C for the CPU, 30 something°C for the other sensors), the voltage increased a bit when i quit prime95 (regulation not so good under load, who to blame?).

Turn it off, then on, then off (a couple of times, every 10-15min, to test), until right now I was in the BIOS checking voltages and it shut down, and now it refuses to restart. 5VSB a perfect 4.9V, fans start when grounding the green wire, voltages ok (11.6V, 4.9V, 3.3V) but not enough juice to start the PC.

By the way, and I don't think this is good, there was still tension on the rails with the PC off: 12V = 0.2V, 5V = 0.11V. Leakage somewhere? Fucked transistors?

I'm afraid the thing will go to the recycling center, it's unusable like this and not knowing what exactly is wrong, it doesn't make much sense changing parts randomly to fix it.

[PCBONEZ]

Re: No boot and random shutdown

>> 12V = 0.2V, 5V = 0.11V. <<
... is probably just residual stored in the filter caps.
To discharge them 100% there would have to be a dead short across them and that wouldn't be a good thing.

There may be a trimmer pot in there to tweek the feedback signal.
That would recenter to regulation point.
If you move ANYTHING then mark where it is before you start and go in tiny steps.

Also, if you haven't, then unplug and replug the PSU-MoBo connectors a couple of times. You may have oxidation built up in the connections and doing that is usually enough to help things.

On your restart.
The fans and voltages are coming up so the PSU is starting but the system isn't starting.
That suggests that the pwr_good signal may not be passing.
That signal must go before the CPU itself has 'permission' to start up.
-
It works like this.
The chipset gets power then continually sends a reset signal to the CPU until it (the chipset) gets the pwr_good signal from the PSU. Then the chipset stops resetting the PSU and the PSU starts up.
.
Thing is it is a one shot deal, once the CPU is running the pwr_good signal doesn't affect a thing whether it's there or not (I have cut that wire on a running system to test this and nothing happened, it kept right on running) so that doesn't offer an explanation as to why it shuts down in the first place.

I don't know your PSU but some pass the pwr_good from the same IC the PWM is in.

A lot of things are pointing toward the controller IC(s) malfunctioning.
Might be a supporting component and not the actual IC though.

Anything look like it's been too hot?
Bad solder joints?

.