Intel D845GVSR Recap - No boot

[cheebster]

Hi. I've been trying to recap an Intel D845GVSR board. It had failed Nichicon NM(H) 820UF 6.3V X 6, they were all popped.

All i had off the shelf was some Tayeh (?!?) 10V 1000UF. I've installed the 6 of them, and the board doesn't even try to boot in any way.

I'm shorting pins 6 and 8 of the J9G1 header, it should be ok, checking from



Do i need to replace other caps ? I'm unsure if the board should even try to boot or not, at this point. I tried 2 cpu's, 2 psu's 2 ram chips.

Thanks in advance for any help ! Vincent.

[Wizard]

Re: Intel D845GVSR Recap - No boot

wrong caps.

Have to be *specific* capacitors like polymer or ultra low ESR as others have recommended here.

Cheers, Wizard

[370forlife]

Re: Intel D845GVSR Recap - No boot

A: Tayeh are definitely a terrible choice for caps (not to mention we need the series, too)
B: you should have kept in the same voltage (moving to 1000uf is fine though,)

[cheebster]

Re: Intel D845GVSR Recap - No boot

Ok. I've been reading a bit, and some say, if they are 6.3V it means they are on a 5V rail and not 12V (which is the next voltage possibly present on a mobo) so i tought 10V was ok to use, as well as 1000uf.

I will find better caps. Thanks

[willawake]

Re: Intel D845GVSR Recap - No boot

10v is ok substitute for 6.3v
if these are new working caps then even if they are not low esr then should get a post or some life at least.

board doesnt post before?
now it still doesnt post

maybe something else than the caps then?

[cheebster]

Re: Intel D845GVSR Recap - No boot

Pretty good question, willawake. I took the caps from an old mobo i had here. They seemed ok visually, and i have a ESR meter (handmade blue one from ANATEK) and they read ok.

I actually bought 4 p4 motherboards as-is on ebay to build ghetto boxes for various projects, had cpu's laying around... i don't know if this one even tried to boot up before recap. I should have tried, but the caps were literally popped on the top, so i changed them right away.

I have another one just like this one, that has a diff. behavior, when i plug the psu on it, the green light flashes on and off and the psu kicks in and out, all by itself. Weeeeird..

Gotta love having problems, eh

[willawake]

Re: Intel D845GVSR Recap - No boot

well if you have esr meter then check all the existing caps on the dead board which are over 330uf
try clear cmos on the dead board.

[kikkoman]

Re: Intel D845GVSR Recap - No boot

are you confident in your soldering? no offense.
pics!!

make sure you have the MB speaker connected. beep codes can be a great help. or a port80 PCI card.

[i4004]

Re: Intel D845GVSR Recap - No boot

probably the fets....
were gone even before that mobo reached cheebster....

>wrong caps.
Have to be *specific* capacitors like polymer or ultra low ESR as others have recommended here.

nope, nope, nope....
here's the correct answer from this very thread
https://www.badcaps.net/forum/showpo...92&postcount=5

[cheebster]

Re: Intel D845GVSR Recap - No boot

i4004, ok. So, i got some defective fets on the one that flashes on and off, or on the one that wouldnt even try to post ?

Thanks, Vincent.

[cheebster]

Re: Intel D845GVSR Recap - No boot

Hey - what's a very low esr, btw ? I get readings all well below 0.1 on these TAAAYEEEEEH caps. Is that good ?

kikkoman - yes my soldering is ok, i use a good weller station as well. Should be good on that part

[Wizard]

Re: Intel D845GVSR Recap - No boot

Caution about i4004, does only offer ill-advised comments. Take a search and read threads about it.

Um, 10v is fine but note that is too much for Vcore which is 2V and less. 10V makes caps bulky in size and can get 6.3V and 4V capacitors to help you get largest uF and lowest ESR possible while keeping physical size small to fit properly.
Polymer allows you to stay small and use 820uF 4V in 10mm or 8mm diameter.

We *do* research for better caps and check on the mainboard voltages to see what their voltages is really are.

Because, for example Intel is infamous for putting 25V caps on very low voltage supplies like 5V, 1.5V etc. Head shaking.

Cheers, Wizard

[willawake]

Re: Intel D845GVSR Recap - No boot

Hey - what's a very low esr, btw ? I get readings all well below 0.1 on these TAAAYEEEEEH caps. Is that good ?

below 0.1 like below 0.06 or something. top quality low esr caps will measure like 0.01 or 0.0

failed is usually over 1.0ohm (i had some bulging at 0.26 though)

[willawake]

Re: Intel D845GVSR Recap - No boot

i4004, ok. So, i got some defective fets on the one that flashes on and off, or on the one that wouldnt even try to post ?

Thanks, Vincent.

how can you say you have defective fets without testing them?

[cheebster]

Re: Intel D845GVSR Recap - No boot

willa; it was a question ! i was wondering if i4004 meant i had bad fets on the board turning on off by itself or the dead as a nail one. He said i had bad fets, i didn't know about that myself. Thanks for the helpful replies !

[i4004]

Re: Intel D845GVSR Recap - No boot

>Caution about i4004, does only offer ill-advised comments. Take a search and read threads about it.

caution, wizard here and there doesn't know what's he talking about...he doesn't care about esr of the caps cheebster has put, while saying he put "wrong caps".
really wizardish of him...he has a crystal ball and knows badcaps even when their esr is fine...uhm...

>Take a search and read threads about it.

search about me?
where pcbonez and i fought?
lol!

you're aware that by saying this you're also saying willawake doesn't know what's he talking about as we have same standpoint in this thread, right?
otoh, could just be you're "offering ill-advised comments" because you're telling the bloke to buy new caps without really knowing if it's caps, because nothing is pointing at the caps thus far...
for all our intentions and purposes he put caps that should enable this mobo to turn on...

in the end, i would politely ask you to point me to one such "ill-advised" comment, as you've put it. also with explanation why is it bad.
thank you very much!


and oh yes...

>Because, for example Intel is infamous for putting 25V caps on very low voltage supplies like 5V, 1.5V etc. Head shaking.

nonsense...
they could put 50, 100, or 200v if they were made for such capacitance, and if they could fit them, and if they had esr that satisfies...
voltage doesn't matter as long as it's higher than the voltage circuit runs on...

just as willa said, 10v can replace 6.3...if your crystal ball doesn't show that, try it for yourself...

-------------------------

cheebster, i said
"probably the fets...."
in a sense that if caps you've put are ok(and it seems they're, no matter what wizard "feels") and all other caps you test are fine(as suggested by willa) then next thing to check are fets...there's thread here somewhere about it, but the procedures in it are not really that easy to do...(if i remember correctly)
what you could do is do a ballpark estimate , ie measure fets on few ok boards, and then on suspicious board and see if you can see any difference...(measuring resistance between the pins)
offcourse, only reliable way of testing them is to desolder them and test outside the circuit...
(and offcourse if fets are gone it's not like you're 100% sure it's just them and not some other semiconductors there too...like chips etc.)

i was thinking about the board that won't power on...

and as willa said, check other caps too...

another thing that steers away from caps is the fact they were
"literally popped on the top"
as you've said it...and fets can't take infinite beating (and venting caps do beat them)...
ie i think you GOT these bards for cheap because they're in such condition..

you say you have another board like the dead one and it goes on/off?
well, that probably has better chances to live, because if it turns on it means fets are still alive...or at least some of them...heh
use that one to check fets against...

>Hey - what's a very low esr, btw ? I get readings all well below 0.1 on these TAAAYEEEEEH caps. Is that good ?

yes, that's good for now(for testing..as you don't have better) but nobody would really recommend relying on tayeh for longer periods of time...
ie boards work when esr of caps is "well below 0.1"....
they develop problems on higher esr values, as will said...

btw. do you have a way to test capacitance of those tayehs?
there's some small chance esr is ok, but capacitance is not...but that occurs rarely...

and oh yeah, have you disposed those badcaps you've took out?
if not check its esr.
ans also check esr of caps on that other board(one that goes on/off), as that should be borderline case and their esr is especially interesting for us...

[Wizard]

Re: Intel D845GVSR Recap - No boot

ESR do matter and current ripple rating.

Higher voltage caps get bulkier in size and the insulation get thicker, and ESR rises and impendance changes too much to be useful. Otherwise we'd love to have mainboard with very few magical capacitors instead of little caps all over.

Cheers, Wizard

[Wizard]

Re: Intel D845GVSR Recap - No boot

Oh, your mention about capacitance, that is only amount of capacitor that can hold number of electron charges in coulombs (C). By itself it is just a number of potiential energy stored and not very useful unless combined with ESR (the circuit's view) to do work which needs circuit the capacitor is part of to cause electrons to flow into or ebb from a capacitor as storage or couple through.

Capacitance is measuring amount of cap in farads ONE THING, we don't measure for that, we measure ESR to determine the overall health of electolyte function containing within electroytic capacitor as this is what determines function of the circuits. High ESR capacitors from circuit's view appears like high resistance with high farads aka very floppy capacitor action that can't handle any current rather at all.

Our main intent here is low ESR for proper function. This ESR is what we measure for which there are tools for this, and this is NOT function of capacitance. Relablity in long term (that is reason we demand quality capacitors, can't have acceptable or low quality capacitors, had to be best) as we do care about data (not lose the data) and return in time invested and investment on our equipment insured against damage that bad capacitors causes. Down time is expensive for certain situations.

We know i4004 is promoting fragmentary knowledge without full understanding of what really is involved. I have been there many times where I seen the problems with bad caps caused even acceptable ones, (i4004 implied intention of using medium quality capacitors) were not doing the job at all in most electronics.

In digital application we need low ESR to keep hash on DC supply thereby limit radiated noise back out to the AC and RF emissions to a minimum as a consquence to a minimum on the voltage supply so digital circuit function correctly and cleanly, and keep errors to vanishingly small rate.

Cheers, Wizard

[linuxguru]

Re: Intel D845GVSR Recap - No boot

The Vcore rail (VRM output) caps have to be ultra-low ESR. Generally, the best choices here are polymers (Nippon Chemicon PS/PSA/PSC/PSE/PSF/.., Fujitsu, Sanyo OSCONs, ...) or Ultra-Low ESR Aqueous Electrolytics (Panasonic FJ/FL/FM, Rubycon MBZ/MCZ, Nichicon HD/HE/HM/HZ, Nippon Chemicon KZE, ...), Samxon Gx, etc.

The Tayehs (now a division of Teapo) probably weren't too bad when new, but if they're pulls from a dead mobo, all bets are off. They might work OK in low-current bypass applications, but not on Vcore.

You can check if the FETs are shorted by measuring the in-circuit source-drain resistance (using both polarities of the probes). If both are zero or very close to zero, the FET is probably shorted. Also, Gate-Source should be non-zero, usually in the mid kilo-ohms to 100s of kilo-ohms. If it's below a kilo-ohm, it's suspect but not necessarily dead.

Note that FETs in circuit can exhibit very low resistances from drain to source if there's some residual gate charge. Try shorting the gate to ground before measuring drain-source resistances.

[PCBONEZ]

Re: Intel D845GVSR Recap - No boot

Tayeh ESR is no where near as low as HM.
Tayeh is more like FC [or maybe up to KZE on a good day].