Buy a POST card to see if there's anything happening in the mobo

parhelia512 wrote:

Hello everybody :happy: […]
Show full quote

Hello everybody 😀

I bought a MSI MS-6168 ver2 which was in a really bad shape. Lot of dust, bad capacitors, missing inductance and maybe more that i can't detect.
My electronic knowledges are really poor (absent ?) and i would like to rebirth this board.

- For capacitors replacement, i think my job is OK : Same specifications...should be ok.
- For the two inductances, i tried to find visual approaching ones (yes this is not really science ! 😊 )

I gave it a try and...The board is powering on but nothing else.
In fact, the strange thing is that the motherboard wake up 3 to 4 seconds after shorting the good pins...strange.

All MOFSETS seems to report a correct voltage (thanks Shamino to tell me how to do this! ) but the MOFSET that is near the memory slot is reporting 1.6V (which, if this is the one for the memory is too low).

I would enjoy to listen your comments on how can i do to rebirth this little board !

Thanks 😀

Well, it's good enough, there are no invisible factors affecting a coil's properies. As long as wire length, number of turns, coil diameter, wire material, core material, and wire cross section area are the same, they will have the same inductance. Basically, if they look the same, they will be the same electrically too. Also, in that application the tolerances are quite large, they're just filtering out HF ripple.

In short, if you replaced the inductor with something visually identical, you should be ok in that department 😀

Well, you have a suspected error. I'd definitely begin there. Find some SIMM VCC/VSS pins and check the voltage.

Getting a POST card is also a good suggestion. They're cheap, just a couple of euros on Ebay, and sometimes they can be helpful.

The picture you attached is of your motherboard or did you take it from the internet? To better help you we would also need clear pictures of each area of your motherboard.

What PSU did you use? What CPU? What RAM? Is the PC speaker making any beeps?

Thanks kaputnik and Deksor ! 😀

So inductance seems right. Capacitors should be ok too...

Could you explain me more concerning the "SIMM VCC/VSS pins" ? Directly on the memory slot ?
I've found a POST card on ebay. Is there something to check before buying ?

Happy to see you here Robert !

I've updated my first post with pictures.

I have FSP power supply (300w) which is ok with my others boards.
My CPU is a pentium 2 400 (tested too with a P3 500 Katmaï)
RAM : tested with all sticks i have (32 / 64 / 128 /256MB PC100 or PC133)
Video : tested with a S3 PCI card with no success

NO speaker sound 😒

The strange thing is really the delay between the start of the computer and when i short the power on pins (3-4s)...

parhelia512 wrote:

In fact, the strange thing is that the motherboard wake up 3 to 4 seconds after shorting the good pins...strange.

Yes, it's strange. Signal from power button somehow should go to W83977EF-AW and to the south bridge PIIX4. And both chips are able to control PWRON# signal to turn ATX PSU on.

parhelia512 wrote:

All MOFSETS seems to report a correct voltage (thanks Shamino to tell me how to do this! ) but the MOFSET that is near the memory slot is reporting 1.6V (which, if this is the one for the memory is too low).

VR1 LX8384 (not a mosfet) makes VTT voltage (1.5V), so 1.6V is Ok. RAM takes power right from ATX connector (3.3v rail).
What about voltage of the mosfet bellow the Slot1 (FDD6030L)?
What is CPU VCCcore voltage? Is CMOS battery Ok? Are there hot chips (hot for a finger) on motherboard ?
Try to run the system without RAM. The motheboard should signal "no RAM" error by beeping.

Thanks Roman555 for your answer.

What do you think about this delay ? One of these 2 chips is dead ?

Ok, this is the VTT voltage. Should this voltage be stable ( this is a question for another thread : i have a slot1 board with VTT fluctuating from 1.28V to 1.72V and the board is not stable.)
I will be able to check another time the voltages this weekeend.

Tried without CMOS battery and with a not so new one.

Will check the hot spots this weekend two (should i search somewhere in particular ?)

WIthout the RAM stick there is no beep 😒 ...will try to hotflash the BIOS this weekend, i'm not sure it is good.

Thanks 😀

I have the ms6168 rev 2 too. Mine was flashed with the Award BIOS which means better CPU support but "no onboard video". Which BIOS do you intend to flash?

Try a Pentium 2 / Celeron up to 333MHz. It might not "like" CPUs with FSB100

Hellor Arctic,

I was thinking in flashing an original AMI BIOS.
Was'nt the board shipped with a P3 700 inside the Packard Bell case ?

Will try with a slocket + celeron FSB 66MHZ, thanks for this suggestion !

The rev1 (with i440ZX) was shipped with late P2 and early P3 Katmai CPUs, the rev2 (with i440BX) was shipped with Katmai and later Coppermine P3s. Every MS-6168 would have supported 100MHz FSB, so if it only does 66MHz, it's got problems. That said, if a component isn't working, trying it at lowest clock is always a good idea. You don't even need a Celeron for that, just set it to 66MHz with the P2 in it. In terms of BIOS, a P2 Deschutes would be the least challenging option (Mendocino is a newer CPU).

That said, I'd be highly surprised if a rev2 had trouble with Mendocino or Katmai, but you never know... I have two rev2 boards. One boots fine with any Coppermine up to 1GHz, the other - identical in every aspect I've been able to figure out - works perfectly with Katmai but utterly refuses to POST with Coppermine. Unfortunately the video memory on the Coppermine-capable board is dying, so I'm stuck with a P3-550 on the other one for now.

parhelia512 wrote:

What do you think about this delay ? One of these 2 chips is dead ?

I don't know. There is some glue logic on-board that corresponds to these signals and others (pwrok, pwrgood), сrystal oscillator 32kHz. Also I've never seen such behaviour.

parhelia512 wrote:

Ok, this is the VTT voltage. Should this voltage be stable ( this is a question for another thread : i have a slot1 board with VTT fluctuating from 1.28V to 1.72V and the board is not stable.)

Yes, it has to be stable. As datasheet sais:

1"VTT must be held to 1.5 V ±9%. It is recommended that VTT be held to 1.5 V ±3% while the Pentium II processor
2system bus is idle. This is measured at the processor edge fingers." 

parhelia512 wrote:

Will check the hot spots this weekend two (should i search somewhere in particular ?)

First of all - chipset, super I/O controller. It's recommended to check temperature of another big and little chips.
But be careful not to damage component by static electricity.

parhelia512 wrote:

WIthout the RAM stick there is no beep 😒 ...will try to hotflash the BIOS this weekend, i'm not sure it is good.
Thanks 😀

I would try to run motherboard without BIOS chip, without CPU to see whether the starting delay has been changed or not.

Arctic wrote:

I have the ms6168 rev 2 too. Mine was flashed with the Award BIOS which means better CPU support but "no onboard video". Which BIOS do you intend to flash?
Try a Pentium 2 / Celeron up to 333MHz. It might not "like" CPUs with FSB100

I uploaded the latest AMI BIOS for this mainboard. It should support Coppermine, IMO.

The lower coil/inductance definitely looks very different from the stock one.
The stock one has only got one wire, yours has three. The amount of windings is different as well. I think it's ten windings on the original.

Is there somebody here who has got the board and can make good pictures of the coils including diameter measurements?

Roman555 wrote:

Arctic wrote:

I have the ms6168 rev 2 too. Mine was flashed with the Award BIOS which means better CPU support but "no onboard video". Which BIOS do you intend to flash?
Try a Pentium 2 / Celeron up to 333MHz. It might not "like" CPUs with FSB100

I uploaded the latest AMI BIOS for this mainboard. It should support Coppermine, IMO.

Both the retail AMI and PB OEM Awards BIOS work fine with Coppermine.

dionb wrote:

Both the retail AMI and PB OEM Awards BIOS work fine with Coppermine.

dionb,I've always thought it's vice versa: AMI is for PB and Award is for retail 😀
Doornkaat, here it is (it's not mine)
http://tdfx.de/pic/ms6168_07b.JPG
https://youtu.be/V_ezB6kH4Q4 (at 20sec it's seen clearly 10 windings)
According to schematics it should be these chokes:
choke1 - 4µH
choke2 - 1.6µH
Yes, I see now and I'm agree that installed choke1 is different from that is supposed to be.

Thanks for all your replies.

So, todo list for this weekend :
- Change choke1 / try
- Hotflash with uploaded AMI BIOS / try
- Check all the voltages

Small hope 😀

Before change of choke1 it's better to measure voltage VCCcore on this choke. Maybe it's ok...
How many windings does your choke1 have? Where did you get this choke (shop, another motherboard) ? - I'm just trying to understand its inductance.

OK Roman555. How can i measure voltage on the choke ? From the MOSFETs ?
I'm sorry it's a newbee question 😒

Maybe I don't understand your question. You can measure voltage using voltmeter (or multimeter in voltmeter mode).
Black probe is connected to GND (black wire of molex connector of PSU or rings around mounting holes of mainboard), red probe - to point of measurement. You posted early that "the MOFSET that is near the memory slot is reporting 1.6V" - so you measured it somehow 😀 Just be careful not to short something with probes.