VOGONS


K6-2/3+ Success on Asus P5A 1.06 - simple mod

Topic actions

Reply 80 of 109, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

Okay, so we know that we can pull wt/wb high or low to switch between write back and write through. But which is which? I read through the thread and no one seems to have mentioned which is which.

Vcc3 pull up is faster than vss pull down.

From what I can read online it seems write through is the slower one and write back is the faster one.
https://www.geeksforgeeks.org/write-through-a … -back-in-cache/

So pull up is write back
Pull down is write through?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 82 of 109, by debs3759

User metadata
Rank Oldbie
Rank
Oldbie
Sphere478 wrote on 2022-03-31, 21:44:
Okay, so we know that we can pull wt/wb high or low to switch between write back and write through. But which is which? I read […]
Show full quote

Okay, so we know that we can pull wt/wb high or low to switch between write back and write through. But which is which? I read through the thread and no one seems to have mentioned which is which.

Vcc3 pull up is faster than vss pull down.

From what I can read online it seems write through is the slower one and write back is the faster one.
https://www.geeksforgeeks.org/write-through-a … -back-in-cache/

So pull up is write back
Pull down is write through?

According to the post at Re: K6-2/3+ Success on Asus P5A 1.06 - simple mod pull up is WB, pull down is WT, so you surmise correctly

See my graphics card database at www.gpuzoo.com
Constantly being worked on. Feel free to message me with any corrections or details of cards you would like me to research and add.

Reply 83 of 109, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
debs3759 wrote on 2022-04-01, 00:58:
Sphere478 wrote on 2022-03-31, 21:44:
Okay, so we know that we can pull wt/wb high or low to switch between write back and write through. But which is which? I read […]
Show full quote

Okay, so we know that we can pull wt/wb high or low to switch between write back and write through. But which is which? I read through the thread and no one seems to have mentioned which is which.

Vcc3 pull up is faster than vss pull down.

From what I can read online it seems write through is the slower one and write back is the faster one.
https://www.geeksforgeeks.org/write-through-a … -back-in-cache/

So pull up is write back
Pull down is write through?

According to the post at Re: K6-2/3+ Success on Asus P5A 1.06 - simple mod pull up is WB, pull down is WT, so you surmise correctly

Doh!

Good job as usual Snufkin, and Debs3759!

I missed that when I skimmed through. Thanks!

Okay that answers that. I’m updating the silk screen to differentiate between the two modes.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 84 of 109, by dj_pirtu

User metadata
Rank Member
Rank
Member

Struggling with mine P5A v.106 with mod and K6-2+/570. Random reboots in windows and nothing seems to help. Took all the cards out, swapped memory sticks, swapped PSU...

Machine seems to be stable, no memory errors, I can run 3DMarks with no problems but when in Windows desktop doing nothing, it just reboots like pressing reset-button.

Any hints?

Reply 85 of 109, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
dj_pirtu wrote on 2022-05-03, 07:28:

Struggling with mine P5A v.106 with mod and K6-2+/570. Random reboots in windows and nothing seems to help. Took all the cards out, swapped memory sticks, swapped PSU...

Machine seems to be stable, no memory errors, I can run 3DMarks with no problems but when in Windows desktop doing nothing, it just reboots like pressing reset-button.

Any hints?

-Try a lower power, older gpu.
-Have you done the k6-2+/3+ mod?
-Latest bios?

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 86 of 109, by dj_pirtu

User metadata
Rank Member
Rank
Member
Sphere478 wrote on 2022-05-03, 07:56:
-Try a lower power, older gpu. -Have you done the k6-2+/3+ mod? -Latest bios? […]
Show full quote
dj_pirtu wrote on 2022-05-03, 07:28:

Struggling with mine P5A v.106 with mod and K6-2+/570. Random reboots in windows and nothing seems to help. Took all the cards out, swapped memory sticks, swapped PSU...

Machine seems to be stable, no memory errors, I can run 3DMarks with no problems but when in Windows desktop doing nothing, it just reboots like pressing reset-button.

Any hints?

-Try a lower power, older gpu.
-Have you done the k6-2+/3+ mod?
-Latest bios?

I have K6-2, will try that
I have done the K6-2+ mod (and noticed I have used 2,2kohm resistor, not 1k, can it be the problem...)
Latest BIOS

Edit: took motherboard out from the case and now it works...

Reply 87 of 109, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
dj_pirtu wrote on 2022-05-03, 08:02:
I have K6-2, will try that I have done the K6-2+ mod (and noticed I have used 2,2kohm resistor, not 1k, can it be the problem... […]
Show full quote
Sphere478 wrote on 2022-05-03, 07:56:
-Try a lower power, older gpu. -Have you done the k6-2+/3+ mod? -Latest bios? […]
Show full quote
dj_pirtu wrote on 2022-05-03, 07:28:

Struggling with mine P5A v.106 with mod and K6-2+/570. Random reboots in windows and nothing seems to help. Took all the cards out, swapped memory sticks, swapped PSU...

Machine seems to be stable, no memory errors, I can run 3DMarks with no problems but when in Windows desktop doing nothing, it just reboots like pressing reset-button.

Any hints?

-Try a lower power, older gpu.
-Have you done the k6-2+/3+ mod?
-Latest bios?

I have K6-2, will try that
I have done the K6-2+ mod (and noticed I have used 2,2kohm resistor, not 1k, can it be the problem...)
Latest BIOS

Edit: took motherboard out from the case and now it works...

I don’t think the p5a mod is fully understood. Perhaps it could be the 2.2k resistor. The original modder and subsequent people to copy it used 1k resistors. Including myself all to successes in regards to the wb/tb issue so far as I know.

I however still had many stability issues on my 1.06, which I Believe to be not related to the mod. I think much of it had to do with specific gfx cards I tried in it. There were also issues by my recall related to older bios versions.

I think Jan once reported that sometimes the latest bios actually would cause errors, a report I was unable to replicate when trying different versions.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 88 of 109, by dj_pirtu

User metadata
Rank Member
Rank
Member

Been testing alot past two days. Random rebooting happens even at @105FSB, but if I disable external cache in bios it seems to be stable @115FSB so only AMD K6-III+/633MHz internal caches are in use.

Maybe my mobos cache is busted, don't know. Strange is that I can run Quake3 benchmark and 3DMark99 and 3DMark2000 with no problems but the random rebooting happens when sitting idle in Windows 98SE desktop.

No problems in DOS. But, I'm happy if it works without L3-cache, speed difference is minimal.

Reply 89 of 109, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
dj_pirtu wrote on 2022-05-04, 10:24:

Been testing alot past two days. Random rebooting happens even at @105FSB, but if I disable external cache in bios it seems to be stable @115FSB so only AMD K6-III+/633MHz internal caches are in use.

Maybe my mobos cache is busted, don't know. Strange is that I can run Quake3 benchmark and 3DMark99 and 3DMark2000 with no problems but the random rebooting happens when sitting idle in Windows 98SE desktop.

No problems in DOS. But, I'm happy if it works without L3-cache, speed difference is minimal.

It’s possible you may have dried out capacitors.

Try recapping it.

Bad caps give line noise, give poor stability.

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 90 of 109, by dj_pirtu

User metadata
Rank Member
Rank
Member
Sphere478 wrote on 2022-05-04, 10:31:
It’s possible you may have dried out capacitors. […]
Show full quote
dj_pirtu wrote on 2022-05-04, 10:24:

Been testing alot past two days. Random rebooting happens even at @105FSB, but if I disable external cache in bios it seems to be stable @115FSB so only AMD K6-III+/633MHz internal caches are in use.

Maybe my mobos cache is busted, don't know. Strange is that I can run Quake3 benchmark and 3DMark99 and 3DMark2000 with no problems but the random rebooting happens when sitting idle in Windows 98SE desktop.

No problems in DOS. But, I'm happy if it works without L3-cache, speed difference is minimal.

It’s possible you may have dried out capacitors.

Try recapping it.

Bad caps give line noise, give poor stability.

Maybe caps are the reason. Been testing whole day and it's not stable even on default clocks. Now I know how to test it, if I play MIDI-file in windows through Aureal Vortex2 it will reboot in 10 seconds.

And no, it will do random rebooting without Vortex2 so that's not the one to blame here.

One thing that bothers me is that if I google 'Asus P5A K6-2+' I can find some conversation of the exact same problem from over 20 years ago...

Reply 91 of 109, by dj_pirtu

User metadata
Rank Member
Rank
Member
Fenyo wrote on 2021-03-30, 10:06:

I have another P5A rev 1.06 like this, but it has 1 MB onboard cache and tag chip (I've checked with wpcredit and strangely both tagchips are active), the northbridge revision is the same (G) but little bit older ('99 vs '00) , sadly this board is only stable in Dos and Win95/OSR2.X, newer op's like 98, 2000 or Xp are resetting (like someone mentioned before) if the system is idle (under Linux even during the system startup), under load (and normal cpus) it seems stable. Stability is little better if prevent irq sharing (manaully add irq in bios) and increase the pci latency timer to 64, but not solve it, so the mods are half success 1 off 2 prefectly stable.

My chipset reads 9935 (year/week) and exact same problems.

Reply 92 of 109, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie

I think the problems of the P5A rev 1.05 and 1.06 are because of asus and not the chipset. Unlucky I'm not aware of any other Aladdin V motherboard with such revision of the chipset so hardly we will know the truth.

Reply 93 of 109, by Fenyo

User metadata
Rank Newbie
Rank
Newbie
dj_pirtu wrote on 2022-05-04, 14:36:
Fenyo wrote on 2021-03-30, 10:06:

I have another P5A rev 1.06 like this, but it has 1 MB onboard cache and tag chip (I've checked with wpcredit and strangely both tagchips are active), the northbridge revision is the same (G) but little bit older ('99 vs '00) , sadly this board is only stable in Dos and Win95/OSR2.X, newer op's like 98, 2000 or Xp are resetting (like someone mentioned before) if the system is idle (under Linux even during the system startup), under load (and normal cpus) it seems stable. Stability is little better if prevent irq sharing (manaully add irq in bios) and increase the pci latency timer to 64, but not solve it, so the mods are half success 1 off 2 prefectly stable.

My chipset reads 9935 (year/week) and exact same problems.

Since then, I’ve come to the conclusion that bios also affect things. I've reverted the bios to 1010.002 back and it goes perfectly with that Linux (which doesnt even start with latest bios) as with win9x too, but it's still not stable with 2k and xp (but not tested too much). Let's say, of course, the processor doesn't recognize this older bios and you have to move on with f1 key, so you could patch this bios or even older ones, but cbrom can't read the cpu microcodes from this special Asus bios, maybe Jan from The Unofficial K6- The owner of 2+ / K6-III + page can help you with this/us ...

Reply 94 of 109, by dj_pirtu

User metadata
Rank Member
Rank
Member
Fenyo wrote on 2022-05-08, 18:28:

Since then, I’ve come to the conclusion that bios also affect things. I've reverted the bios to 1010.002 back and it goes perfectly with that Linux (which doesnt even start with latest bios) as with win9x too, but it's still not stable with 2k and xp (but not tested too much). Let's say, of course, the processor doesn't recognize this older bios and you have to move on with f1 key, so you could patch this bios or even older ones, but cbrom can't read the cpu microcodes from this special Asus bios, maybe Jan from The Unofficial K6- The owner of 2+ / K6-III + page can help you with this/us ...

I actually tried that bios and made it made no difference. I'm quite sure there was something wrong with CPU power supply...
Sold the board away already because I lost my nerves with it and now I'm playing with Asus TX97-XE and Asus P55T2P4.

Intel did something right, TX97-XE isn't much slower @83FSB than P5a @115FSB.
with Voodoo3, Quake3 demo001 -benchmark:
P5A K6-III+/633MHz: 40fps
TX97-XE K6-III+/500: 36fps.

Too bad there's no AGP.

Reply 95 of 109, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

I also was very dissapointed with my 1.06 my 1.04 however... good mobo. but for me the freeway mobos are where it's at.not the best overclockers though. but stable and awesome features

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 97 of 109, by Repo Man11

User metadata
Rank Oldbie
Rank
Oldbie
dj_pirtu wrote on 2022-05-13, 04:19:
I actually tried that bios and made it made no difference. I'm quite sure there was something wrong with CPU power supply... Sol […]
Show full quote
Fenyo wrote on 2022-05-08, 18:28:

Since then, I’ve come to the conclusion that bios also affect things. I've reverted the bios to 1010.002 back and it goes perfectly with that Linux (which doesnt even start with latest bios) as with win9x too, but it's still not stable with 2k and xp (but not tested too much). Let's say, of course, the processor doesn't recognize this older bios and you have to move on with f1 key, so you could patch this bios or even older ones, but cbrom can't read the cpu microcodes from this special Asus bios, maybe Jan from The Unofficial K6- The owner of 2+ / K6-III + page can help you with this/us ...

I actually tried that bios and made it made no difference. I'm quite sure there was something wrong with CPU power supply...
Sold the board away already because I lost my nerves with it and now I'm playing with Asus TX97-XE and Asus P55T2P4.

Intel did something right, TX97-XE isn't much slower @83FSB than P5a @115FSB.
with Voodoo3, Quake3 demo001 -benchmark:
P5A K6-III+/633MHz: 40fps
TX97-XE K6-III+/500: 36fps.

Too bad there's no AGP.

Be sure to try benchmarking the memory bandwidth (Speedsys, Sandra) with and without the motherboard's cache enabled - I've a TXP4 with a K6-3+ and I found the motherboard's cache actually lowered the memory bandwidth when enabled. The TX chipset only caches 64 megabytes of RAM, so I at first assumed that this slow down was caused by my having 256 megs of RAM installed, but I then tried it with only 64 megs of memory, and the result was the same

"I'd rather be rich than stupid" - Jack Handey

Reply 98 of 109, by Sphere478

User metadata
Rank l33t++
Rank
l33t++
dj_pirtu wrote on 2022-09-16, 14:30:

Got myself another P5A v1.06 today (with K6-2/500 and 128MB RAM), need to mod it and run some tests with K6-III+. Hope this one is OK.

2 hasn’t been known to mod to 3 or 3+

Only 2+

You can make a 2 with a 2+3+ though by moving the second resistor

Anyway, my personal experience with 1.06 hasn’t been that great. Recommend a 1.04

Pulling wt/wb to vcc3 using a 1k reisitor seems to have worked with others. On 1.05 1.06

Sphere's PCB projects.
-
Sphere’s socket 5/7 cpu collection.
-
SUCCESSFUL K6-2+ to K6-3+ Full Cache Enable Mod
-
Tyan S1564S to S1564D single to dual processor conversion (also s1563 and s1562)

Reply 99 of 109, by debs3759

User metadata
Rank Oldbie
Rank
Oldbie
Sphere478 wrote on 2022-09-16, 15:58:
2 hasn’t been known to mod to 3 or 3+ […]
Show full quote
dj_pirtu wrote on 2022-09-16, 14:30:

Got myself another P5A v1.06 today (with K6-2/500 and 128MB RAM), need to mod it and run some tests with K6-III+. Hope this one is OK.

2 hasn’t been known to mod to 3 or 3+

Only 2+

You can make a 2 with a 2+3+ though by moving the second resistor

Anyway, my personal experience with 1.06 hasn’t been that great. Recommend a 1.04

Pulling wt/wb to vcc3 using a 1k reisitor seems to have worked with others. On 1.05 1.06

I think he was talking about modding the motherboard to run the a K6/3+

See my graphics card database at www.gpuzoo.com
Constantly being worked on. Feel free to message me with any corrections or details of cards you would like me to research and add.