Gigabyte GA-7VAXP voltage issues

[wij the king]

hi everyone
i discovered this forum 2 days ago so please be indulgent if it's adouble post (searching was fruitless):

well, first here are my specs:
Gigabyte GA-7VAXP rev 1.2 F15 BIOS
256 MB TwinMos 333MHZ
AthlonXP 2000+ box
2x80GB ATA133 maxtor
Skystar2 DVB Card
GeForce 4 MX 440 AGP 8X
Motorola SM56 modem
52x CD ROM Drive
24X CD-RW drive
380 W PSU Unit
WIn 98 SE French

All settings in bios to auto, no raid enabled

Dram V : 1,05V
V core : 1,74 V
+12V :12,49V
+5V : 4,56V
CPU FAN : 4900-5100 RPM
System T° in bios (idle mode) : 52°C
Cpu T° in BIOS : 58-65°C

and here is my problem
Since dec 2005 i was unable to install win xp (or 2k). it all begun after a modem installation, os kept crashing randomly if i surf the web, open explorer, movies or mp3, Visual studio...i unplugged so what i thought faulty (this modem) but problem reamained.all that last 15days

after that i managed to borrow a Winxp cd and did a fresh and clean install, but os kept giving me BSODs with IRQL_not_less_or_equal messages during installation, same thing for win 2k. copying files always runs fine, but after first reboot i get either BSOD at "34 minutes left" or if CHKDSK attempt to scan a disk.
i tried to test my disks in an other PC, nothing suspect, no virusus, no corrupted system File, no problems with partition, MBR or anything else. i backuped all data, reformatted, repartitionned but no way, problem remained and it is always the same.

after several month of discussion on hardware forums, someone told me that my DDR voltage were way way way too low for the quantity of RAM i had, normally, i would expect 2.45-2.5V and i'm having only 1.05V!!!

what i first did is changing the PSU, no luck...voltage remain unchanged.
i tried then to changes these voltage in BIOS, no luck..still unchanged.
i then tought that i had some faulty components, so i pulled the board out of its case and looked carefully for bad caps, and here's what i noticed:
1- a component (maybe a diode...see n° 1 in the photo) had some char under and around it, on it i can find this writing : PJ2D SR32
2- a capacitor was slightly bending (n° 2 in the photo)

i proceeded immediately to the replacement of the capacitor and let the other component because i haven't found how to replace it, but nothing changed.

yesterday, i pulled again the mobo for a new verification on the caps, and i found the same first component near Firewire pins and it was apperantly still working fine (nor char around) (see n°3 in photo) so i removed component 1 and 2 and switched their position!!!!(i know that's wrong but i was hopeless)

after that t re-cased the mobo and booted, in BIOS i found that DDR voltage jumped to 1.32V and when i launched win98 i was able to let it work until this morning without problems (never occured in the past!!!!)

so finally here is my questions:

--> do you know what is the component named PJ2D SR32, where can i find it?
--> can the capacitor placed near that element (see n°4) be faulty, may i replace it? it is a 6.3V 1000 mFarad ...
--> do you know for this MOBO waht capacitors can cause voltage issues, as the MOFSETS are apprently working fine

thanks in advance for replies and sorry for the photo quality and the lenght of the post...

here is the official photo :
http://tw2005.giga-byte.com/MotherBo..._GA-7VAXP.htm#

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

sorry for the double post but that ten minutes limit..
here is another photo I found were element 1 is better shown

[kc8adu]

Re: Gigabyte GA-7VAXP voltage issues

replace the bulging or leaking caps and all of the same type with high quality(read japanese) low esr caps.
lately i recommend replacing all 470mfd and up.any on the bad list gotta go.gigabyte has a habit of using a mix.some good some crap.

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

finally a reply...
well i have no more bulging caps on that board, besides i first want to know if replacing the mosfet could resolve the problem as i'm having power regulation issues only for memory, for CPU all is going fine...
and can someone with that same board indicates me what are caps that influe on meory voltage?
thanks Kc8adu for the help...

[willawake]

Re: Gigabyte GA-7VAXP voltage issues

actually i saw your post earlier but its a little difficult to answer your questions this time. anyway welcome to the forum and thanks for providing a lot of information in your post.

the way i see it is that you replaced the mosfet with another one which came from a circuit with probably nothing to do with vdimm and that mosfet is maybe in good condition. so perhaps when you replace with a new one it will have the same result that you have now.

on the other hand, maybe the caps is the issue but i think maybe also you have another unknown part which is defect.

KC8 did not say to only replace the bulging caps.

replace the bulging or leaking caps and all of the same type with high quality(read japanese) low esr caps.

you probably have GSC caps on the board, they are one of the worst. perhaps you have choyo also, they are bad too. maybe more fail in the future and then you have another problem, ie no post or fet/vrm chip damage and then cpu killer board.

but anyway i guess you are taking this into consideration and you want to figure out which caps are in the memory circuit to see if replacing those will correct the low vdimm. and then to do the full job of replacing all bad caps when you see the board is going to be good.

problem is that it is not possible to determine really all the caps in the memory power circuit. the vdimm output caps yes, they would read 1.32v, but the input is what you need. if i was doing the job i would use esr meter to test all the caps and replace only those bad and then see if the vdimm problem is correct before using many new caps and getting nowhere.

and damn that twinmos is tolerant of low voltage if stable at 1.32v. i can think of several that would be annoyed at anything less than 2.5v.

[MD Willington]

Re: Gigabyte GA-7VAXP voltage issues

GSC is very common on GA boards...

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

thanx willawake and MD willington for the care...
well no improvement since yesterday, the whole thing is damn stable at 1.32V and i had no random crash at all.
just to remember you my MOBO is GA-7VAXP ver 1.2, and on this version i read yesterday that caps are often good quality ones, so here is what i found opening the case:

--> All the 6.3v x 1000uF (14) are......Rubycon, even the one that bulged and that i replaced with an unbranded capacitor (oops...)
--> All the 6.3v x 1200uF are (2) are.....Rubycon also!!!
--> Four 6.3v x 3300uF are KSG or KZG (i'm not sure..) and they seems physically good (no bulging, no electrolyte, no char...)

so here i'm really getting confused, Rubycon are the best one i thought...do i have to replace them? besides they seems good!!!
the 4 KSG (or KZG) seems to me of no replacement, they are physically far from memory circuit so...
finally i have no ESR tester, a multimeter but that's all...so i think i'm gonna replace the two rubycon caps of 1200uF coz they are the nearest to the one that bulged in the past...
if nothing happens i'll proceed with another cap: the nearest (n°4) to the memory circuit and the ATX connector..
if nothing happen, I think i'll switch two mosfets that seemed suspicious to me...
and if nothing happens....i'll throw that board in the garbage : ....
no no seriously, this board is inevitably on its way to death, so i'll continue with it till it die completely (i think that it wouldn't damage the CPU coz VCore are practically = 1.756V and AMD indicates that it should be 1.75V ) and then i'll buy a new nforce3/KT800 based one and an AThlon64 3500+...

anyway, i'm going to purchase the caps this afternoon and i'll keep you informed with evolutions...

EDIT: you said it WillaWake that twinMos accept anything...it used even to run fine (even when playing games or wathching video...) with only 1.05V !!!

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

again that 10min Limit...
well here are the two rubycon to be replaced (i found the photo in the forum, it's not my MOBO...) and please i have the same char aroud these NEC K3405 25M, does it sound normal to you or do i have to try replacing them?

[Per Hansson]

Re: Gigabyte GA-7VAXP voltage issues

The "char" around the VRM legs are fine, it is solder paste

Your memory would not operate at 1.32v, I think your mobo is reporting the voltages incorrectly, get a multiemter and measure the voltage at an empty DIMM socket

http://pinouts.ru/Memory/dimm_184p_pinout.shtml

And when you are done with that test some more points, like on VRM's for the chipset etc

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

Ok Per Hanson, i'll do it, but just to verify that i have well understood the pinout you shown me, for example i'll test voltage between Pin n°3 and Pin n°7 (VSS and VDD)??

and for VRM's, what voltage may i expect ,i need areference to know if they are OK or not!

and another thing, the machine should be running when i di these tests or just plugged in the AC, please note that i have a DDR led that lights when you plug the power cable into the PSU, does it mean that there is current flowing into RAM slots ? i just don't want to blow anything or short any component by inattention !!!

[Per Hansson]

Re: Gigabyte GA-7VAXP voltage issues

No, only measure between VDD and chassi ground.

For the CPU VRM's you should get very close to CPU VCORE (1.75v in your case)
Unless you BIOS lists what default chipset voltage your chipset should get you will have to look it up, or ask here, but do the measurement first...

Also be careful when measuring the VRM's, if you short it you might end up sending 12v directly to the CPU, with abvious results...

[Fatal0E]

Re: Gigabyte GA-7VAXP voltage issues

again that 10min Limit...
well here are the two rubycon to be replaced (i found the photo in the forum, it's not my MOBO...) and please i have the same char aroud these NEC K3405 25M, does it sound normal to you or do i have to try replacing them?

Those arent Rubycon's, you should replace all caps that are same brand as whatever those are.

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

FatalOE these are Rubycon caps, don't believe the photo these are caps put on a rev1.1 GA-7VAXP, mine is ver1.2, practically all capacitors in that board are rubycon ones, only 4 KSG or KZG are suspicious for me...

[willawake]

Re: Gigabyte GA-7VAXP voltage issues

if they are all rubycon then dont replace them. was the one that was leaning actually bulging or just leaning.

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

it was bulging...and please WillaWake do KSG or KZG brand exists?

[tiresias]

Re: Gigabyte GA-7VAXP voltage issues

FatalOE these are Rubycon caps, don't believe the photo these are caps put on a rev1.1 GA-7VAXP, mine is ver1.2, practically all capacitors in that board are rubycon ones, only 4 KSG or KZG are suspicious for me...

The KZG should be no less reliable than the Rubycon... At least, if they're real United Chemicon KZG.

...

PS. If your machine's actually running, then I'm pretty sure that your Vddr is not 1.05volts!

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

Ahh i see now, so KZG are United Chemicon ones...it sounds logical for me as all the other caps are good quality ones...besides why will they try to mix good and bad caps, loosinge hence the improved reliability of the good ones...
EDIT: maybe the BIOS report wrong info but that's what it is shown in PC helath section.., and now he reports 1.312V

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

AHHH i must be damned!!!!!
why do i have to live in such a country, you can never find branded things here

anyway, her i am again with bad news, i haven't found ANY rubycon, sanyo, panasonic, nichicon chip here!!!!
the worst is that i live in the capital city and i passed more that 3 hours from a shop to another in the greatest "electronic avenue" without luck !!!!

AHHH I hate gigabyte....and tunisia sometimes....

[wij the king]

Re: Gigabyte GA-7VAXP voltage issues

Hi everyone
i'm here again for good and bad news....
let's start with good ones, well as i haven't found any good quality caps i decided to ask to a freind of mine to search for some used (but branded) ones that he may find in 3 motherboards that he told me he had..these boards are no more used (don't know why...) and i hope he can find something useful for me within tomorrow

meanwhile, i tried saturday to reinstall winXP and guess what...it installed without a single problem
All drivers had also been installed successfully, all PCI cards (modem and DVB card) also, no crashes, no restarts....

so that was good news, let's go for bad ones now..
so as i was testing my components i tried to watch some TV using my DVB card, and i was surprised coz the program couldn't tune any channel
the problem appeared to be that the signal was absent, after some trials and tests on an external digital reciever i found that the cable was OK, it was hence something in my PC that was causing troubles.
rebooting to Win98SE gave me same results, i couldn't tune anything !!!
after a couple of manipulation (changing PCI slot, removing all other components, changing PSU...) and some readings (manufacturer FAQ, some DVB forums...) i found that, to tune a channel, the DVB card has to "lock a Tuner" by sending him a certain amount of power (amperage i think) and that tuning requested that 5A can flow in the PCI slot to the card

so to resume, i'm not having these 5A!!!
and that's probably due to voltage issues in my mobo...so now i'm left with my only solution, find and recap my mobo and see what happens..
keep you informed as usual....

[gonzo0815]

Re: Gigabyte GA-7VAXP voltage issues

Keep in mind, that those DVB cards work like networkcards, therefore any firewall will lead to the same problems you experiencing. Therefore make sure any firewall is disabled and som like Norton or nvidia should be deinstalled.
Normaly for a LNB anntenna there you only need about 0,5 -2A, but 12 or 18v. if the card is only feed with 5 v it need those 5 A.
I don`t think your problems are cap related in any way. IF you have Rubycon and Chemicon there shouldbe no problems. I suspect your psu could have some bad caps or even can?t deliver mutch power....

EDIT:

on the pic there are a lot of silkscreen spaces for big vrm caps. If there are realy that few caps on that board, i wonder how it can perfrom in any way. i wouldn`t hesitate to solder some big 3300uf MCz Rubycons in there. All decent boards i saw have at least above 12000uf (even Asrock uses 13200uf (4x 3300uf KZG), which is not a conservative number, e.g more should be better.) on the vrm output side. With anything less i doubth that an amd Athlon Xp will running fine on the long run.
And as far as i tink it is only a dual phase vrm, those needs very mutch input and output capacity to runing reliable and not overstress the inputcaps.