This seems odd...

[virtuetovice]

I have an HP Slimline desktop PC I caught for a whopping $20 with a no power issue. After recapping the mobo and MOST of the PSU[the only caps I didn't replace were two 200v 220mf, they seemed fine], it would still not power on. AS soon as I plug the PSU into the mobo the green light on the PSU cut off. Searched around a bit and ran into this thread

http://h30434.www3.hp.com/t5/Desktop.../693875/page/7

One of the suggestions was to blast the psu with a hair dryer. Seemed silly but what the hell I tried. IT WORKED!!!

Now the million dollar question...what components in a PSU might be affected by heat being applied? Is it a slight possibility that those two caps will only fire up if they're warm, while showing no obvious signs of failure? Could it possibly be something else? Forgive me if this thread is in the wrong section!!!

[mmartell]

Re: This seems odd...

I have an HP a6200n which died the other day. Suspected the power supply. Took the machine and the PS to a guy who had another. Put his in mine and same thing - pulsing light on the back of the PS. He had a tester and verified with it that mine and his PS were good.

Took it home and threw some hair dryer love at it and it came to life. So now I try to narrow down which caps on the MB are bad so I can do a quick fix but the bad one (s) are proving elusive. Turns out that was because it wasn't the MB but the PS. Sprayed a little hot air directly into it and solid green light right away.

I will be investigating soon and will post back what I find.

[c_hegge]

Re: This seems odd...

It's probably a bad solder joint. The two big 200V caps don't normally fail.

[virtuetovice]

Re: This seems odd...

The computer is an HP Pavilion s7600n slimline, by the way.

I'm investigating myself, I have the whole computer sitting on the floor, in pieces, running. I can confirm that it is indeed the power supply, I took the cover off and blasted it directly. It does indeed fire up just fine when the power supply is hot. I'm thinking about letting it cool down and possibly focusing the heat on one particular area to narrow it down a bit...


Furthermore, I swapped the single core AMD Athlon 64 processor out for a dual core version and the PSU just couldn't take the extra load of the more powerful chip. It would post and boot into Windows fine but after a short while it would just shut down. I'm almost certain there's a wattage issue going on, being that the dual core pulls more watts.

I'd love to see this little thing sitting on my entertainment center running Windows 7, hooked up to the 55 inch sammy. It's quite a capable little machine, 64 bit AMD processor AM2 socket, nVidia GeForce 6 series/NF430 chipset, but the shortfall is only 512 RAM. No mater, I have a pair of two gig sticks and a dual core to drop in it. It also has built in wifi, 5 USB's, card reader, DVD burner and a 200GB hard drive.


A part of me says screw it, just buy another PSU for 25 bucks on eBay but a bigger part of me says, what's the fun in that?!!


I was thinking the same thing c hegge, I'll go on a little iron adventure and post my results.

[Agent24]

Re: This seems odd...

Did you change the PSU startup capacitor?

[PeteS in CA]

Re: This seems odd...

What A24 posted. It's usually located near the PWM, between the two heatsinks. Look for a small electrolytic cap, perhaps 47uFor 68uF, 25V or 35V. Heating it lowers its ESR, so that the PWM can get the power to start up (once it starts, there is a winding on the transformer that provides power to the PWM).

[mmartell]

Re: This seems odd...

I tested what I believe to be the startup cap this morning and found exactly that. Heated it while the leads were on it and watched the esr fall.

There are two caps pretty much beside each other (beside the pwm) and haven't traced them out yet but that one def appears faulty.

Will replace and update.

[virtuetovice]

Re: This seems odd...

I see a 47mf 50v and a 100mf 25v

Of course the 47/50 is buried in grey glue....

I'll see if I can source those tomorrow. They're right in the middle of the PSU, where I found the heat seems to be affecting the start up.

Also, soldering it up did not work.

[Agent24]

Re: This seems odd...

I would replace both

[mmartell]

Re: This seems odd...

My startup cap was measuring about 70 ohms cold esr and about 15% its rated capacitance.

Replaced it and back in business. Thanks for the heads up!

[virtuetovice]

Re: This seems odd...

Not sure which of the two was the culprit but after replacing both, The system boots normally...

Good call to those who suggested the start up caps. I find it interesting that 5 other caps in the PSU showed signs of failure but the ones that were the REAL culprit showed no visible signs of failure.

[Agent24]

Re: This seems odd...

Not sure which of the two was the culprit but after replacing both, The system boots normally...

Good to hear

Good call to those who suggested the start up caps. I find it interesting that 5 other caps in the PSU showed signs of failure but the ones that were the REAL culprit showed no visible signs of failure.

All of them were the culprits. The 5 others that showed visible failure may not have stopped the PSU from powering up if you had only replaced the startup capacitor, but they would have been letting dirty power into your system if you had used them in that state.

That's why the saying here goes: "All bulging caps have failed, but not all failed caps bulge"

They can simply dry out which shows no outward visible signs but will at the very least cause a significant reduction in capacitance - Sometimes even go completely open-circuit.

I think this may be due to the smaller size (less can pressure?) or since they are not doing a 'hard' job and are probably a general purpose series than low ESR, have a different chemical composition which tends toward the 'silent' failure modes. That said - some brands - for example OST often fail without visible signs, even in the tough areas where other capacitors would vent.

[virtuetovice]

Re: This seems odd...

It kinda sucks that I had to roll with Jackcon and NTE replacements but I guess I will see how they hold up...

I only have ONE other rant about this comp besides the inferior caps. It seems the hard drive manufacturer decided to use the cheesiest, cheapest, crappiest plastic on the face of the planet when they made the SATA connector. It snapped with absolutely no hesitation. Thankfully the pins themselves are fine and the hard drive works but it looks like I'll have to glue the SATA cable to the hard drive LMAO

[momaka]

Re: This seems odd...

I think this may be due to the smaller size (less can pressure?) or since they are not doing a 'hard' job and are probably a general purpose series than low ESR...

Exactly this.

Most small caps usually just don't enough "juice" inside to vent or blow like the bigger ones, so instead they have a silent death.

It kinda sucks that I had to roll with Jackcon and NTE replacements but I guess I will see how they hold up...

Jackon and NTE are general purpose, with the former also being quite shoddy in quality.

If the PSU runs hot, don't expect them to last too long. Also, you may be stressing the capacitors on your motherboard a lot more if the caps you put in the PSU are not doing a good job filtering out the ripple.

For that reason, it is best to stick with Japanese replacements from Digikey, Mouser, Badcaps, and etc. Don't get me wrong, Priestcain Electronics is a nice place, but not the right place when it comes to caps. It does appear to me that some of their parts come straight from Digikey, so it may be possible to do an order through them. That said, their prices will be marked up (it's a store after all, though, so that's expected). If you live close by them and need just one or two common parts that they probably have in stock, it's worth going there. But if it's something like caps, just buy in bulk online (again, only Digikey, Mouser, Badcaps, and other reputable places - NOT ebay!).

Thankfully the pins themselves are fine and the hard drive works but it looks like I'll have to glue the SATA cable to the hard drive LMAO

LOL.
Just FYI, in case you come up with something cool, we do have a ghetto mod thread here :
https://www.badcaps.net/forum/showth...tto+mod+thread

BTW, congrats on the fix! Don't you love it when a 10 cent part fixes a piece of equipment worth at least 100 times that ammount?

[mmartell]

Re: This seems odd...

I need clarification on something...

A cap can go open or have reduced capacitance or have a high esr.

Is there another failure mode which causes ripple which can be seen with a scope or is that a function of capacitance/esr which can both be measured with other equipment ?

[Agent24]

Re: This seems odd...

The capacitor is used in the filtering circuit to short the AC component (the ripple) to ground. Any failure mode which prevents this from happening, or from happening properly, will mean the ripple will be present or measure higher than normal.

Though there are other reasons aside from a faulty capacitor that can cause ripple to be seen on a scope when testing, including (but probably not limited to) some other fault in the PSU (Even such as using a capacitor with ESR that is too low, in some designs), or interference on the scope input due to incorrect testing procedure\setup.

[mmartell]

Re: This seems odd...

What I meant was, if a cap measures at its stated capacitance and rated esr can it not be ruled out as imparting ripple on the circuit ?

I ask because I ONLY replaced the offending cap. The others measured good. Replacing a good cap with a good cap buys you more time but it doesn't clean up your power any better.

[Agent24]

Re: This seems odd...

If the capacitor has the correct capacitance and ESR, then it is unlikely to be faulty and it should be working in the circuit as intended. Whatever that intention is. For example, the ATX specification allows for up to 120mV of ripple on 12v.

Yes, there is no point in replacing perfectly good capacitors with new ones - with the exception that if they are known crappy brands\series then you would be wise to replace them anyway for preventative measure to ensure longer life of your PSU and anything powered from it.

[lti]

Re: This seems odd...

Jackon and NTE are general purpose, with the former also being quite shoddy in quality.

Jackcon does make low ESR caps. They were the reason why Abit got sued.