SV25/FV25 problems after recap

[ArOs]

I bought a SV25 a while ago. After some time I discovered that many FV25 MB's have had problems with bad caps... but mine was still running just fine. But to be safe I decided to recap my board anyway.
I ordered a kit from Topcat and then recapped the board with the received Rubycon caps. Everything seemed fine - nice soldering, correct polarization and no visible damages.

Now to the problem... after the recap I now have frequent problems with that the computer hangs in the "Verifying DMI Pool Data" state.
BUT, if I disable the L2 cache - the computer boots (read that tip in
a forum). The person/persons that had this problem SOLVED it by changing
caps...

I have tried booting with an other PSU, and with an other CPU - but the problem persist... I even resoldered the caps.

Maybe the problem is one of the new caps? What suggestions do you have?

The capacitance of the new caps differ from the old caps. Maybe thats a problem?

The new caps:
3x 4700uF 10v 12,5mm
3x 3300uF 6.3v 10mm
5x 1000uF 10v 8mm

The original caps:
3x 3900uF 10v 12,5mm
3x 2700uF 6.3v 10mm
5x 1000uF 10v 8mm

[Per Hansson]

Re: SV25/FV25 problems after recap

That's about 20% less capacitance, it could be a problem. I'm not too sure about that though... I'm assuming those are for CPU VCore? Are there any unpopulated capacitors spots there that you could try populating?

[Topcat]

Re: SV25/FV25 problems after recap

I've recapped hundreds of FV25's with those values of caps, never had an issue. Your problem lies elsewhere.

[ArOs]

Re: SV25/FV25 problems after recap

No, actually it's more capacitance now than before. I think you misread the figures.
To upper the values shouldn't be a problem as far as I know... but maybe that's wrong?

I have actually ordered new caps (Samxon) with "correct" capacitance and voltage - we'll see if that solves the problem.

[Topcat]

Re: SV25/FV25 problems after recap

unless changing the caps out corrects a soldering error you made, it's not going to help. sorry. There is something else wrong with the board. Most likely a through hole tube was damaged in the process. I have seen that cause similar issues on other boards.

[ArOs]

Re: SV25/FV25 problems after recap

unless changing the caps out corrects a soldering error you made, it's not going to help. sorry. There is something else wrong with the board. Most likely a through hole tube was damaged in the process. I have seen that cause similar issues on other boards.

OK - can that be fixed?

[Rainbow]

Re: SV25/FV25 problems after recap

It can be fixed using a piece of wire but you need to find which one is damaged.

[Per Hansson]

Re: SV25/FV25 problems after recap

Yes ArOs, sorry, I misread your post... The new caps should work just fine, more capacitance is good (up to a point)

[ArOs]

Re: SV25/FV25 problems after recap

Now I've done some more testings... I've tested the MB with several different CPU's - with different results. Every CPU is known to be working before the testings.

The CPU's I've tested with is two different tualatin celerons (1.2 and 1.3GHz), one tualatin pentium iii-s (1.4GHz) and one coppermine pentium iii (733MHz).

The celerons do pass the "Verifying DMI Pool Data" message (and starts to boot the OS) maybe 2 out of 8 trials. The 1.2GHz (SL656) seems to be slightly better on this compared to the 1.3GHz (SL5ZJ).

The P-III-S 1.4GHz are better than the Celerons on passing the "Verifying DMI Pool Data" state. It do succeed about 8-9 out of 10.

The coppermine P-III passes every time.

In the end of the testings the MB seemed to pass the DMI message more often than before I begun... but I'm not completely sure - but it do feels like it. Can that have any explanation?

Now I'm going to test the stability of the system with the different configurations. Of course I will post the results here.

[ArOs]

Re: SV25/FV25 problems after recap

Now I've run prime95 for about 1,5 h with the 1.4GHz P3 without errors. Seemed good. Then I rebooted the computer... but than it hanged again in the Verifying DMI state... Rebooted - and then it was OK... rebooted again - still OK. Very strange...

If you google for "sv25 dmi" then you will get very many hits. If you read through the found pages then you'll see that this problem seems to be derived from bad caps. So probably my problem have the same cause... we'll see if it's so when I get the new ones. Maybe I've found the very first bad Rubycon cap...

[gonzo0815]

Re: SV25/FV25 problems after recap

No, i for shure dont belive that this issue can be solved by changing freshly changed Rubycons, but i can understand that you belive in this solution .

I would grab a magnifying glass and would do a visual inspection of ALL solderpoints and for scratches.

I do not know your equipment nor your soldering skills, but belive it or not, those multilayered boards are not that easy to solder, and need a lot of heat to be soldered (meaning completely filled with liquid solder) right.

I have read on some TV tech forums that even those fear the doublelayerd (!) pcb of some digital picture prozessing modules in modern 100Hz digital CRT TV sets.
And i had the pleasure to see, how an very very experienced TV tech replaced a few caps on a HP SF system, with a very expensive Weller 80W (!) station and a semiprof. vacuum desoldering station.

Now you see, even very experienced and equiped people have some problems, if they haven`t figured it out.

Nex i would try to replace the small caps too, if the board is equiped with them.
This is usually not necesary, but if i would have a recaped board with similar problems, i would try it.

[kc8adu]

Re: SV25/FV25 problems after recap

i would doublecheck polarity of every cap you changed.
i would also resolder all joints on caps you replaced.
use a good hot iron to make sure the solder flows well.
add more to get the flux.
i have seen traces cut by careless use of the common xcelite flush cutters too.
check and carefully recheck your work and you will find the goof.

[linuxguru]

Re: SV25/FV25 problems after recap

1) Check the DIMM slots - the contacts lose their springiness easily after a few insertions. Clean them with isopropyl alcohol.
2) Also swap the RAM and check - bad RAM sometimes cause these symptoms.
3) Examine the cap(s) right next to the DIMM slots and recap if necessary.
4) If the PSU is a Channel-Well, it will most probably need to be recapped.