VOGONS


First post, by Nemo1985

User metadata
Rank Member
Rank
Member

Hello dear vogonians, i'm fighting against this motherboard that I suppose it has something wrong but I can't understand what precisely.
I know the board can have issues with power hungry agp videocard, in fact i'm using a pci voodoo 3 2000, change the ram installed (1 dimm 128mb) gave no result, the bios version is the latest one (R2.36), the card is in pristine physical condition and I can't really find what the problem is, the revision is 1.32 and it is the 1mb version.

I checked the jumper of the motherboard and the only wrong setting was the ram frequency bound with agp clock instead of bus, when I changed the jumpers I noticed that the instability gone worse.

In Dos the system is apparently fully stable, while on windows it crash randomly, it was able to run 1mb super pi with a k6-3+ but it isn't stable with an higher voltage cpu (IDT Winchip v3.52), after the first step it says: not exact in round.
Also testing the ram with memtest gives a general protection always at the same time (around 3 minutes), there are 4 capacitors rated V6,3 1000uf and another 4 rated V6,3 1200uf and they seems in good shape, no bulging or leaking, i'm almost giving up on this mb, any advice is welcome.
Pictures: https://imgur.com/a/vOiudDg

Thanks!

Last edited by Stiletto on 2020-08-16, 06:07. Edited 1 time in total.

Reply 1 of 16, by ph4nt0m

User metadata
Rank Member
Rank
Member

The VRM looks like a good synchronous buck. I would replace 4 output capacitors solid polymers 6.3V or 4V, 10mm diameter, 1500uF most likely. 2 input capacitors with 6.3V, 8mm diameter, 1000uF to 1500uF; also all other 8mm capacitors on this board with these to be safe.

My Active Sales on CPU-World

Reply 2 of 16, by Nemo1985

User metadata
Rank Member
Rank
Member

Thank you ph4ntom, I just need some clarifications the 4 output capacitors are the ones near the socket (rated v6,3 1200uf)? The 2 input are those next to the coil? The 8mm capacitors are the smallest ones like the one on the first picture black and grey near the agp slot?

Reply 3 of 16, by ph4nt0m

User metadata
Rank Member
Rank
Member
Nemo1985 wrote on 2020-08-15, 14:56:

Thank you ph4ntom, I just need some clarifications the 4 output capacitors are the ones near the socket (rated v6,3 1200uf)? The 2 input are those next to the coil? The 8mm capacitors are the smallest ones like the one on the first picture black and grey near the agp slot?

Yes. Use them in all important places. +3.3V for AGP, SDRAM, CPU Vio, north bridge. +5V for PCI and ISA.

My Active Sales on CPU-World

Reply 4 of 16, by Nemo1985

User metadata
Rank Member
Rank
Member

Thank you for the clarification, I decided to go with those ones for the biggers capacitors (4 near the cpu + 4 around the mb): https://www.tme.eu/it/en/details/ulr1200_6.3/ … 28m0jf1arrx0cr/
But i'm undecided about the smaller ones (I counted 11 of them), i'm undecided between these alternatives:
https://www.tme.eu/it/en/details/ulr100_16/th … n/ulr107m1ce11/
https://www.tme.eu/it/en/details/urz1c101mdd1 … itors/nichicon/
Which one do you advise?

The x-con are polymer while the rubicon are more old fashioned and smaller, sorry if I sound dumb but i'm worried about the body dimensions and even more about terminal pitch of the capacitors, what's if it is different?
I counted the 1000 uf the terminal pitch is 4mm, the 1200 uf is 5mm and the smaller is 3mm.
Thanks

Reply 5 of 16, by ph4nt0m

User metadata
Rank Member
Rank
Member

X-CON and Samxon are brands of Chinese Man Yue. They may not be better than Japanese capacitors, but they are clearly a win over noname Chinese ones.

ULR337M0JE08RRX0CR
Capacitor: polymer; 330uF; 6.3VDC; ESR: 10mΩ; THT; ±20%; -55÷105°C
Ø6.3x8mm
terminal pitch 2.5mm

These are better replacements for the smaller ones. Low voltage low ESR liquid electrolytic capacitors are EOL mostly these days. Physical size isn't really important except for terminal pitch which is better matched. It's also better to match diameter with a board's layout, but that's only for the looks and not mandatory.

My Active Sales on CPU-World

Reply 6 of 16, by Socket3

User metadata
Rank Member
Rank
Member
Nemo1985 wrote on 2020-08-15, 06:28:
Hello dear vogonians, i'm fighting against this motherboard that I suppose it has something wrong but I can't understand what pr […]
Show full quote

Hello dear vogonians, i'm fighting against this motherboard that I suppose it has something wrong but I can't understand what precisely.
I know the board can have issues with power hungry agp videocard, in fact i'm using a pci voodoo 3 2000, change the ram installed (1 dimm 128mb) gave no result, the bios version is the latest one (R2.36), the card is in pristine physical condition and I can't really find what the problem is, the revision is 1.32 and it is the 1mb version.

I checked the jumper of the motherboard and the only wrong setting was the ram frequency bound with agp clock instead of bus, when I changed the jumpers I noticed that the instability gone worse.

In Dos the system is apparently fully stable, while on windows it crash randomly, it was able to run 1mb super pi with a k6-3+ but it isn't stable with an higher voltage cpu (IDT Winchip v3.52), after the first step it says: not exact in round.
Also testing the ram with memtest gives a general protection always at the same time (around 3 minutes), there are 4 capacitors rated V6,3 1000uf and another 4 rated V6,3 1200uf and they seems in good shape, no bulging or leaking, i'm almost giving up on this mb, any advice is welcome.
Pictures: https://imgur.com/a/vOiudDg

Thanks!

I have this exact board and I love it - one of the best super 7 boars I've ever used for several reasons. A few months ago I got the EXACT same symptoms you're describing, so I took the PC it was in apart and replaced it with another MB. Last weekend I started working on it - it turns out the caps were way off when measured with an ESR meter, even tough capacitance was within spec and the caps were OK visually. So I recapped the board with panasonic low-ESR caps, and the instability went away - at stock speeds at least. When overclocking it's a different story. Usually I can get my K6-3+(e) to 550MHz on most motherboards @ 2.0v, but on the AX59 it was unstable at any voltage - running OK only at stock speed (400) and at 450MHz 1.8v. After further testing I noticed the two MOSFETs feeding the CPU (at the top of the CPU socket) was getting really hot with some CPUs (K5, pentium, high clocked K6-2s over 2.2v) so I replaced them with a similarly specked ones off a dead socket A motherboard, and the AX59 now runs a treat. I even got my PR133 running @ pr166 speeds perfectly stable. Took a couple of days of work and intensive testing, but it was well worth it.

Reply 7 of 16, by Nemo1985

User metadata
Rank Member
Rank
Member

Thank you for your precious advice. I agree it is a very good board mainly thanks to the wide voltage selection.
I will test the mb again before order the capacitors, as a matter of fact, I was able to overclock my k6-3+ to 600 mhz, but I just tested it with 1mb super pi without issues, while with higher voltage cpus (IDT) I had many issues.
Did you recap the whole board or just the bigger capacitors?
Do you have a link with the capacitors you used?
Thank you

Reply 8 of 16, by Repo Man11

User metadata
Rank Member
Rank
Member
Socket3 wrote on 2020-08-16, 19:26:
Nemo1985 wrote on 2020-08-15, 06:28:
Hello dear vogonians, i'm fighting against this motherboard that I suppose it has something wrong but I can't understand what pr […]
Show full quote

Hello dear vogonians, i'm fighting against this motherboard that I suppose it has something wrong but I can't understand what precisely.
I know the board can have issues with power hungry agp videocard, in fact i'm using a pci voodoo 3 2000, change the ram installed (1 dimm 128mb) gave no result, the bios version is the latest one (R2.36), the card is in pristine physical condition and I can't really find what the problem is, the revision is 1.32 and it is the 1mb version.

I checked the jumper of the motherboard and the only wrong setting was the ram frequency bound with agp clock instead of bus, when I changed the jumpers I noticed that the instability gone worse.

In Dos the system is apparently fully stable, while on windows it crash randomly, it was able to run 1mb super pi with a k6-3+ but it isn't stable with an higher voltage cpu (IDT Winchip v3.52), after the first step it says: not exact in round.
Also testing the ram with memtest gives a general protection always at the same time (around 3 minutes), there are 4 capacitors rated V6,3 1000uf and another 4 rated V6,3 1200uf and they seems in good shape, no bulging or leaking, i'm almost giving up on this mb, any advice is welcome.
Pictures: https://imgur.com/a/vOiudDg

Thanks!

I have this exact board and I love it - one of the best super 7 boars I've ever used for several reasons. A few months ago I got the EXACT same symptoms you're describing, so I took the PC it was in apart and replaced it with another MB. Last weekend I started working on it - it turns out the caps were way off when measured with an ESR meter, even tough capacitance was within spec and the caps were OK visually. So I recapped the board with panasonic low-ESR caps, and the instability went away - at stock speeds at least. When overclocking it's a different story. Usually I can get my K6-3+(e) to 550MHz on most motherboards @ 2.0v, but on the AX59 it was unstable at any voltage - running OK only at stock speed (400) and at 450MHz 1.8v. After further testing I noticed the two MOSFETs feeding the CPU (at the top of the CPU socket) was getting really hot with some CPUs (K5, pentium, high clocked K6-2s over 2.2v) so I replaced them with a similarly specked ones off a dead socket A motherboard, and the AX59 now runs a treat. I even got my PR133 running @ pr166 speeds perfectly stable. Took a couple of days of work and intensive testing, but it was well worth it.

It's such a great feeling when a repair goes well. But it's mostly a private victory - trying to tell most people abut how happy you are because you were able to repair an old computer motherboard...

Reply 9 of 16, by Nemo1985

User metadata
Rank Member
Rank
Member

So I tested the mb, I ran memtest with the IDT winchip, it went fine for less than 10 minutes, than the pc rebooted, I took off the power and touched the mosfets they aren't hot, I really wonder what's wrong with this motherboard, it seems a ram issue so is it worth to recap it?
With the ram at 100mhz I can't load windows while at 66 mhz it loads but it is unstable.
Maybe the Socket 7 cpus don't like the 100 mhz bus? Is it possible? Socket3 sorry for asking, were you able to run the k5 with bus 100 mhz?

Reply 10 of 16, by ph4nt0m

User metadata
Rank Member
Rank
Member

Intel Pentiums, IDT WinChips and Rise mP6 don't consume much power. AMD and Cyrix CPUs are another story. The K6-3 was infamous for burning mainboards with crappy VRMs.

I have purchased a bag of Vishay SUB75N03 FETs back in the early 2000's and used them to upgrade many Socket 7, Socket 370 and Slot 1 mainboards. There are better FETs available now, but those are good enough.

My Active Sales on CPU-World

Reply 12 of 16, by ph4nt0m

User metadata
Rank Member
Rank
Member
Nemo1985 wrote on 2020-08-17, 14:44:

Thank you phantom, do you have any good advice for some fets, even if i'm not sure the problem will be the capacitors or the fets...

Salvage them from a dead Socket A or 478 mainboard. If I had to buy new, SUB85N03 or something like that would do great. Just don't buy them from China 😀

My Active Sales on CPU-World

Reply 13 of 16, by Nemo1985

User metadata
Rank Member
Rank
Member

sorry for the bump, I wasted some more time to troubleshoot the mb.
I benched the system with speedsys, at 66mhz I get 101,47MB/s while at 100 mhz I get 131,55MB/s as memory throughput
I restored the 66mhz settings and tried to boot on windows, before it loaded I had a bsod: irreversible error: 0E in 0167:BFF72170
Before spend the money on capacitors and other stuff, is there anything else I can do to know where the problem is?
Obviously the ram I tried works fine for sure, I also tried another ram stick, same result, memtest has been able to run for like an hour and then I got an hard reset, even windows 98 when boots in safe mode I get an hard reset after a brief moment.

Ok I'm almost sure that I found where the problem is! I disabled the l2 cache and the system is fully stable, now the problem is, since it is the 1mb version where I can find any cache chip? From utsource or mouser? The one used is rated 5ns (ut6164c64aq-5).
It's running memtest without issues for 2 hours and counting!
Please help!

Reply 14 of 16, by LunarG

User metadata
Rank Oldbie
Rank
Oldbie

This may be a dumb question, but do you have the Via 4-in-1 driver package installed?
I know this should be obvious, but I remember back in the day, when this board was brand new. I was not aware of this, and these drivers did not come with the board. I had SO many issues until I figured that out.
Once I got those installed, I had no further issues.
I have one AX59 Pro board today as well, and no issues with that either.

WinXP : PIII 1.4GHz, 512MB RAM, 73GB SCSI HDD, Matrox Parhelia, SB Audigy 2.
Win98se : K6-3+ 500MHz, 256MB RAM, 80GB HDD, Matrox Millennium G400 MAX, Voodoo 2, Waveforce 192XG.
DOS6.22 : Intel DX4, 64MB RAM, 1.6GB HDD, ELSA Winner 1000PRO, GUS 1MB, SB16.