P4SD - case N3Q

[MXM]

Hey all, this is another p4sd i got today, the user says it sends no video, someone checked it before and didnt get conclussions on it, im gonna recap some of the caps i see blown, wich are marked in the first pic, i would like to hear opinions on those and the rest, if you consider i should recap any other please just advise... i am not touching the yellow ones cause ive been told in this site they're a special type and also cause i dont see em blown, -wel yeah i do, two out of four look bad IMO around the processor....- anyways i dont have such value/type so im not recappin those as of today...

so i got no POST, the best it was on a try with no DIMMs showed the no memory error (the beeps and red flashes) and then i installed one and got nothing again

so if theres any advise for me in this case please let me know, it'd be very much appreciated.

Thanks.

Overview... the ones marked are the ones im changing cause according to me they're blown...








Peace.

[delaware74b]

Re: P4SD - case N3Q

Recap the board. I see 3 domed caps next to the PCI slots as marked in the attached photo:

[MXM]

Re: P4SD - case N3Q

THANK YOU. I did change those caps... but, still nothing. Changed total of 7 caps, all same value.

UPDATE: Hey, i recapped those I marked and then tested, nothing, no POST, nada.

Then I used the hot air gun and heated it a little, allowed it to cool off and then tested, results still nothing.

If somebody out there has anything on this board (P4SD) that'd help it'd be appreciated. I cant just change boards I wish I could.

BTW all those caps i changed were 1000uf/6.3v TAEPO and i changed em others same values but the brand, Sanyo.

Thanks.

[goodpsusearch]

Re: P4SD - case N3Q

the yellow ones in the photos are sanyo??

[c_hegge]

Re: P4SD - case N3Q

No, the yellow ones they are actually polys. Fujitsu used to make their polys look like lytics, but not anymore.

Replace ALL teapo caps, bulged or not. they are notorius for failing without showing it. the ones which are most likely to be the problem are the ones circled here.

Also, what brand/series did you replace them with?

[MXM]

Re: P4SD - case N3Q

THANK YOU.

I will replace every TEAPO i get to see today...

Yeah the yellow ones are Fujitsu, I Think... but not Sanyo and have not touched em at all.

I have already replaced those I marked and got no good results so Im following ur advise, and will update as wel, thanks.

And wel Ive been using very crappy replacement, Im not an expert but I already know the brand/series I use are probably the worst world wide, but thats all I got, I will be able to buy better caps next month but still wont know which brand Id get til I get em :/ -third world country.- In answer to your question the 1000uf/6.3v are Sanyo, I read on em "N.E. 77 WG" the 1500uf are Ltec and sometimes I use Wendell... I barely get Rubycon or good brands, my bad...

Anyways, thanks again and i will update later on

[MXM]

Re: P4SD - case N3Q

UPDATE: I changed the few TAEPO cheapo the board had left, including the three caps marked by you c_hegge, results:

after the recap I just added the processor and its cooler with one DIMM, it DID show the POST, I turned it off and thought it wasa time to put the other DIMM and place the processor and the board base (cause the proc was just put there no base no locks) once mounted it did not show POST or anything, i tried different DIMMs and processors, and obtained a few other failing POSTs, mostly memory test fail but I dont think is the DIMMs themselves cause I swap em and one time it gave me the memory error with the DIMM it originally worked out... so im stuck with this board that now gave me one sucessful POST a few other unsuccessful and now is back again to nothing... any suggestions??

BTW i have figured that the 1800uf around the processor, the ones mixed with the yellow ones are not the originals but recapped... now they dont look bad but i was wondering if they beiing such a crappy brand might be causeing me this issue?

wel im planning on applying some heat to it and see what happens, any idea is welcome.

Thanks!

[MXM]

Re: P4SD - case N3Q

UPDATE: After applying heat and letting it cool off, i tested it under the following scenario:
No board base, No CPU base (unsecured cooler) 2 DIMMs (of which I tested before and was sending memory errors) and deviceless, results: POST... yeah it has been giving me the POST now im gonna run mtest+ for a few passes and see if it gets stuck, if it doesnt im mounting the board to the base and securing the CPU as it should, and see what happens then... i'll keep up the updates.

Thanks.

[c_hegge]

Re: P4SD - case N3Q

The ones around the CPU (the ones with T-Vents) look like Panasonic FJ. They are a reliable brand and can probably be left in place. Sanyo WG are fine as replacments, aut wendell are pure junk. You can get rubycon from badcaps.net

[yyonline]

Re: P4SD - case N3Q

I've recapped a few P4SD-XX boards out of HPs. Never this particular one, but they all look fairly similar. In fact, I have a P4SD-LA board in the shop now. I'm basing this off the assumption that they are similar boards. They sure look similar, but without seeing both in person, it's hard to say for sure.

I've probably recapped about 30 boards at this point, so I've gotten the hang of things fairly well. But that P4SD-LA gave me troubles. There seems to be a lot of copper in that board that really sinks the heat away.

I don't know your experience level, so please don't take offense to this. If you're new to soldering, you may want to really go over all the solder joints carefully. In my case, the similar board required a lot of heat to make good joints. That can be a challenge if you're new with this. If you can post close-up pictures of the solder joints, members may be able to comment on them. It's sometimes hard to tell from photos, however.

My P4SD-LA was really fussy about memory voltage. It has to be 2.5V ram. A lot of PC3200 is 2.6-2.7V, so careful on that one when selecting memory.

[MXM]

Re: P4SD - case N3Q

***WORKING BOARD***

Thank you ALL for the directions.

Ive recapped many of these particular model as of today but never really done a full resoldering recapping, mainly because I dont have all the caps values.
I am not new to soldering but this board has definetly givenme particular issues even when soldering-desoldering...

All DIMMs are from other boards just like this so u can say im kinda confident those memories fit. I didnt touch those yellow ones in this board at all :P

And yeah BTW i also have a P4SD-LA (Stingray) back home and its really fussy as u say when it comes to memory voltages.-

As for my level experience -being 0 noob and 10 Guru- Id give me a fine 6 or 7 out of 10 but Ive never took caps this serious, bad about this is that i used to do this a lot, but time ago.

This board has been testing all day long and Im going to put it back to work hoping it wont fail again...

Thank you all this has been solved by you.-
delaware74b
goodpsusearch
c_hegge
yyonline

[c_hegge]

Re: P4SD - case N3Q

No problem. Glad I could help.

[MXM]

Re: P4SD - case N3Q

This board has just come back, no POST again ... I was wondering if some has any idea on what could i do to it, it worked for a few days and again, the same issues.

It was working on a wel ventilated area... when it arrived i tested it and it did boot up first time but got stuck at the middle of the OS booting, and thats it.

Any idea is welcome.

Thanks.

[MXM]

Re: P4SD - case N3Q

This is what it looks like today, i havent done nothing to it since it came back























Any idea is welcome

Thanks.

[c_hegge]

Re: P4SD - case N3Q

All I can think of is that you could be using an inappropriate series of capacitor in your re-caps. I'm assuming that the 3 big ones near the Parallel port are 1500uF 16v. In that case use these:
https://www.badcaps.net/store/produc...products_id=38

For the other 1500uF (6.3 or 10v I'm assuming) caps where you used LTEC, use these:
https://www.badcaps.net/store/produc...products_id=39

Correct me if I am wrong about te values.

[MXM]

Re: P4SD - case N3Q

Thank You.

the big ones on the parallel port are 1200uf/16v

the ones besides the DIMM slot are 1500uf/6.3v

Thanks for the info, Im checkin it out, unfortunetly I think I wont be able to order right there but I might be able to find better than what I currently have.

Peace.

[MXM]

Re: P4SD - case N3Q

UPDATE: I applied the heating method on the CPU socket w/Flux and the board has been working on mtest+ for the last 6hrs with 25 complete passes and 0 errors.

Unfortunately i wont leave it on all night since weve had some source -electricity- issues and i dont wanna risk it.

Tomorrow morning i will put it under mprime all day and see what happens, but so far, so good.

Thanks.

[Scenic]

Re: P4SD - case N3Q

well... now you know what to do first on all the other P4SD's
mine is still working fine. probably cause the PSU died early and it wasn't used/fixed since then. i'm currently using the board with a metal backplate in a standard µATX tower to prevent the cooler from bending the PCB too much and cause those socket issues..

edit: something like this.. half a milimeter thick.. got it from ebay

(not my pic)