Announcement

**ss627** · 05-25-2009, 09:50 AM

Re: Gigabyte GA-7VM400AM

Did you try putting in a new bios battery?

**Scenic** · 05-25-2009, 10:56 AM

Re: Gigabyte GA-7VM400AM

any "Choyo" caps on that one..?
25V 330uF (iirc) were common on older gigabyte boards. (socket A and 478 (P4) era)
they pretty much always leak from the bottom and dry out..

**lucky13** · 05-25-2009, 09:00 PM

Re: Gigabyte GA-7VM400AM

Originally posted by ss627

Did you try putting in a new bios battery?

Actually, I took out the CMOS battery prior to test. I put it (checked and has 3V) back in but still won't POST. Gee, I thought that would have been too easy.

**lucky13** · 05-25-2009, 09:07 PM

Re: Gigabyte GA-7VM400AM

Originally posted by Scenic

any "Choyo" caps on that one..?
25V 330uF (iirc) were common on older gigabyte boards. (socket A and 478 (P4) era)
they pretty much always leak from the bottom and dry out..

Nope, anything of larger size are all KZG except two 1200uF/6.3V which are HD(M).

I borrowed a ESR meter and gone thru all the KZG and HD(M) caps and they are not quite low in terms of ESR.

Anyway, I powered it up again and check the voltages on the VRM FETs:

IN Out
4.28 1.65
6.05 0.02
4.27 1.66
7.44 0.02

So, it appears two of the FETs are not turning on. However, all 4 FETs give me similiar ohms so I don't think they are shorted.

Also, I can turn the power to the board on and off using the Power button connected to the mobo so the board is not totally dead.

Anyone has experience on the voltages on the VRM FETs?

Anyone still suspect the 3300/6.3V KZGs even they look fine and are not high on ESR?

Thanks.

**lucky13** · 05-25-2009, 09:33 PM

Re: Gigabyte GA-7VM400AM

Ah... found the instruction by willawake on how to attach picture. Many thanks!

So the green markings are the two have zero volt out.

Attached Files

**lucky13** · 05-27-2009, 12:20 AM

Re: Gigabyte GA-7VM400AM

OK, maybe two of the FETs are dead.
The two that have no output (and cold) have lower ohms than the other two.
The markings on all 4 are:
07N03L
AV340

I Googled and found there are 4 different varieties of this 07N03L:
IPB07N03L
IPD07N03L
IPP07N03L
IPU07N03L

Does anyone know how to tell which is the correct one?
Are there any subs for this?

Thanks.

**Gianni** · 05-27-2009, 03:31 AM

Re: Gigabyte GA-7VM400AM

Hi lucky13

it is hard to say because the marking type it is the same for two MOSFET types: 30A and 80A.

I think probably it is IPD07N03L package TO252 -3-11, let see if other members have a definite answer.

I attach here the datasheet I found, you can search for other MOSFET with same parameters and case.

I wonder how they can sell different devices with the same marking on the package.

Ciao
Gianni

Attached Files

**kc8adu** · 05-27-2009, 05:58 AM

Re: Gigabyte GA-7VM400AM

the mosfets are not a problem.
that is a 2 phase vrm and it it running.one of each pair is the switch and the other a syncronous rectifier.
measure across the big caps close to the cpu and you will also find your 1.65v reading there.
there is a chance the vrm output caps are bad but your only way to prove that is to replace unless you own an esr meter or a scope.do you have a post card?
http://cgi.ebay.com/USB-PC-Desktop-Analyzer-Diagnostic-POST-Test-Card-PCI_W0QQitemZ370153103748QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item562ed82184&_trksid=p3286.c0.m14&_trkparms=72%3A1546|66%3A2|65%3A12|39%3A1|240%3A1318|301%3A1|293%3A1|294%3A50

**lucky13** · 05-27-2009, 06:27 PM

Re: Gigabyte GA-7VM400AM

Originally posted by kc8adu

the mosfets are not a problem.
that is a 2 phase vrm and it it running.one of each pair is the switch and the other a syncronous rectifier.
measure across the big caps close to the cpu and you will also find your 1.65v reading there.
there is a chance the vrm output caps are bad but your only way to prove that is to replace unless you own an esr meter or a scope.do you have a post card?
http://cgi.ebay.com/USB-PC-Desktop-Analyzer-Diagnostic-POST-Test-Card-PCI_W0QQitemZ370153103748QQcmdZViewItemQQptZLH_DefaultDomain_0?hash=item562ed82184&_trksid=p3286.c0.m14&_trkparms=72%3A1546|66%3A2|65%3A12|39%3A1|240%3A1318|301%3A1|293%3A1|294%3A50

Well, that's good to know about the MOSFETS. I was about to unsolder the legs to ohm them again.

Funny I just borrowed a POST card from someone today. Although it does not look like the one you quoted, this one has ISA and PCI edges. It does have 4 digits readout like the one you shown. I will plug it in and try and post back results.

Is the one you quoted newer and has better indications?

**lucky13** · 05-27-2009, 06:30 PM

Re: Gigabyte GA-7VM400AM

Originally posted by Gianni

Hi lucky13

it is hard to say because the marking type it is the same for two MOSFET types: 30A and 80A.

I think probably it is IPD07N03L package TO252 -3-11, let see if other members have a definite answer.

I attach here the datasheet I found, you can search for other MOSFET with same parameters and case.

I wonder how they can sell different devices with the same marking on the package.

Ciao
Gianni

I found Digikey in the USA carries IPU07N03LAIN-ND. It is a N-CH 25V 30A TO-251 MOSFET.

I might not have to order them as kc8adu said they are probably OK as this is only a 2 phase VRM!

Thanks for your effort!

**lucky13** · 05-27-2009, 08:39 PM

Re: Gigabyte GA-7VM400AM

OK, this looks like the POST card I borrowed:

http://cgi.ebay.ca/ws/eBayISAPI.dll?...0054557&_rdc=1

I plug this card into one of the PCI slot. No memory installed. Just CPU and speaker. K/B & on board video connected. Now, here are the results:

Press power button.
Memory LED on mobo lights up
CPU fan spins
No beep, nothing on the screen.
The POST card shows 4 dashes.
These LEDs are lit:
All 4 on the left hand column, -12V, +12V, +5V, +3.3V
Only one light on the right hand column lights up: CLK
The other 3 are off: TRDY, FRAME & RESET

Apparently, the 4-digits should go from 00 to FF?

Anyway, just for fun, I removed the CPU, fan and tried again:

Same result EXCEPT the RESET on the right column now lights up too.

So what does this all mean? Dead CPU?

Should I go get another supported CPU to try? Any chance this board is a CPU killer?

**lucky13** · 05-29-2009, 11:41 PM

Re: Gigabyte GA-7VM400AM

OK, borrowed a working Duron 750 and popped it in. Same result. No post.

Putt he Duron back into the donor motherboard, still works. Yeah, so at least this board is not a CPU killer.

Anyone wants to chime in?

**kc8adu** · 05-30-2009, 08:34 AM

Re: Gigabyte GA-7VM400AM

put your chip into the donor.report results.
of course make sure the board can support yours.and your board can support the donor chip.
must configure the boards right too.

**lucky13** · 05-30-2009, 10:00 PM

Re: Gigabyte GA-7VM400AM

OK, recap those 5 KZG 3300uF and 2 HD 1200/6.3V.

I was really hoping for a miracle....

but NO GO.

Same for the original Sempron chip and the Duron chip I borrowed.

The POST card does not show anything other than what I've reported previously. So I guess the board just doesn't get to POST.

I was holding out hope for this board because it seems to function other than no POST. The board will shut off if I leave the CPU heatsink loose so that tells me the temp sensor to shut off still works. The NB and SB do not get really hot. The NB which has the heatsink definitely runs much warmer than the SB.

Anyway, I put the Sempron CPU into the donor board to see if the CPU is dead. It got to the Compaq logo and I smell something burning so I guess the CPU is alive.. I quickly yank the ac cord and looks like no damage done to the donor board (not sure about the Sempron). it was a jumperfree board.

Anyway, the POST card shows many different numbers on the 4-digit display as the Compaq board boots up. So my conclusion is that until the board takes the 1st step, nothing will happen and the board remains in veggie state.

I will see what the owner wants to do with this board. Maybe I can keep it and I might further learn how the boot cycles works and find something...

Thanks for the pointers everyone.

**lucky13** · 06-03-2009, 12:24 AM

Re: Gigabyte GA-7VM400AM

I was reading the post "Testing VRMs" made by AK0R.

He said

d. Measure resistance from output of VRM to ground (should be greater than 1 megohm, usually much greater).

Well, just for fun, I ohmed the VRM 3300uF output caps (replaced already) and they have like 84 ohms!

So, they are not a dead short but perhaps explain wht this board won't POST.

Now, where do I look again for this partial short?

**PCBONEZ** · 06-03-2009, 01:58 AM

Re: Gigabyte GA-7VM400AM

Check for solder hairs or spatter trails.

**lucky13** · 06-03-2009, 09:14 PM

Re: Gigabyte GA-7VM400AM

Originally posted by PCBONEZ

Check for solder hairs or spatter trails.

There isn't anything wrong with soldering... looked with 30x magnifier.

Using the picture I have posted above, this is a 2 phase VRM as per kc8adu.

I've used my DMM to measure the dc resistance from Drain-Source on the 1st MOSFET (the one that has 7.44 in and 0.02 out, see 1st table). This as I understand, is the Vcore voltage.

Here is a somewhat interesting/surprising result:

with CPU in the socket, the dc resistance is 0.2ohm

if I remove CPU out of socket, it is 32.2ohm

Remember the MOSFETs are not hot (see temperature taken with IR temp gun). If anything, 2 of them are pretty cold.

Tested with 2 known working P/S (SPI brand).

So what does this mean ???

**kc8adu** · 06-04-2009, 06:02 AM

Re: Gigabyte GA-7VM400AM

if vcore is reading correct for your cpu look elsewhere.
normal to read some resistance on vcore to ground.
you probably killed your procs running without the hsf.
amd athlons dont like that!by the time you smell it or the board shuts down its too late.

Announcement

Gigabyte GA-7VM400AM

Gigabyte GA-7VM400AM

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment