stanwebber wrote on 2022-05-18, 18:39:

Con 2 botones wrote on 2022-05-18, 16:46:

But I did try an exotic mobile Duron, DHM0950AQS1B. This didn´t work in other two socket A boards of mine, didn´t work on this […]
Show full quote

But I did try an exotic mobile Duron, DHM0950AQS1B.
This didn´t work in other two socket A boards of mine, didn´t work on this Iwill either.
I suspect it is in non working condition though, it is a little"biten" on its die lower-right corner:

that's a huge bite. it looks even worse zoomed in. there's no hope going that deep into the core.

Got it very inexpensive, sold untested.
I would like to test a mobile Duron or Palomino on the board to see if those are supported.
As you said, this one looks defunct.

Con 2 botones wrote on 2022-05-18, 18:50:

Got it very inexpensive, sold untested. I would like to test a mobile Duron or Palomino on the board to see if those are support […]
Show full quote

stanwebber wrote on 2022-05-18, 18:39:

Con 2 botones wrote on 2022-05-18, 16:46:

But I did try an exotic mobile Duron, DHM0950AQS1B. This didn´t work in other two socket A boards of mine, didn´t work on this […]
Show full quote

But I did try an exotic mobile Duron, DHM0950AQS1B.
This didn´t work in other two socket A boards of mine, didn´t work on this Iwill either.
I suspect it is in non working condition though, it is a little"biten" on its die lower-right corner:

that's a huge bite. it looks even worse zoomed in. there's no hope going that deep into the core.

Got it very inexpensive, sold untested.
I would like to test a mobile Duron or Palomino on the board to see if those are supported.
As you said, this one looks defunct.

My chip is a mobile palomino. Aka Corvette core. You will need an AHM chip I believe for it to work.

ptr1ck wrote on 2022-05-18, 18:55:

Con 2 botones wrote on 2022-05-18, 18:50:

Got it very inexpensive, sold untested. I would like to test a mobile Duron or Palomino on the board to see if those are support […]
Show full quote

stanwebber wrote on 2022-05-18, 18:39:

that's a huge bite. it looks even worse zoomed in. there's no hope going that deep into the core.

Got it very inexpensive, sold untested.
I would like to test a mobile Duron or Palomino on the board to see if those are supported.
As you said, this one looks defunct.

My chip is a mobile palomino. Aka Corvette core. You will need an AHM chip I believe for it to work.

Ah sorry! Yes it is stated in your first post in this thread. My bad.

Finally got this thing button up to where it's really functional. CF reader is mounted in the case now and is formatted to 2GB with DOS 6.2. I played a bit of Tomb Raider 1 with the rush patch. Also fired up Redneck Rampage and Industry Giant for a few minutes. The cheap old CRT that I picked up really seals the deal.

Just to update..

I'm dual booting into DOS 6.22 and use Access as handy little game launcher when running DOS games. The machine runs everything I've thrown at it very well. Ultima 7 and Warcraft 2 even play flawlessly once the L1 cache is disabled. I handle that thru a batch file that sets setmul then launches the game, and runs setmul again once the game closes.

i hit the jackpot with a misc lot of sdram i picked up on ebay. i got 3 sticks of 512mb with infineon hyb39s2564000t-7 & hyb39s2568000t-7 chips. i can run these at cas 2 with the fastest timings (5-2-2) at a 145mhz fsb. memtest86 clocks their speed at 502mb/s. i haven't tried 150mhz as i'm not looking for the maximum overclock.

this triples the amount of ram in my kk266r-plus. the previous 512mb could barely run at cas 3 with 6-3-2 timings at 140mhz fsb (memtest86=434mb/s) and couldn't run at cas 2 period.

i think this system now tops my m7mia-r ddr board, although without ecc.

ptr1ck wrote on 2022-03-21, 16:16:

I've been wanting a KT133A board and since Abit KT7As in good shape are commanding a premium I picked up an Iwill KK266+ in need of a recap for less than 20 bucks. It has 34 new Rubycon caps on it now and looks like it's running right. The KK266 series aren't know to be compatible with Thorougbreds or Bartons (and mine isn't with the ones I have on hand) so I decided to try the low-key Mobile Athlon 4 processor for this build. I picked up an AHM1200AHQ3B for less than a happy meal, shipped to my door.
<snip>

So I got around to testing my KK266+ (re-capped) today with both a Corvette and a modded Palomino (second L5 bridge connected with conductive paint).

After running PCISET to change register 55 so that the multiplier change will go through, I found that although multiplier changes were now successful, the system became completely unstable afterwards.

This was the case with both the Corvette and the Palomino.

I also observed that the modded Palomino did not accept multiplier changes through the BIOS (would not POST), but that's probably related to another bridge needing to be connected, and it's not really that important to me.

Are you able to use SETMUL properly with your setup?

I don't have this problem with any of my other KT133A boards and Barton chips.

mockingbird, It didn't work for me, but I may have did it wrong when I tried. It's been a few weeks since I've had it plugged in since I have limited space for my computers anymore (like 7 and counting). 😁

Can you post exactly what commands you ran so as I can try the same?

ptr1ck wrote on 2022-07-14, 12:54:

mockingbird, It didn't work for me, but I may have did it wrong when I tried. It's been a few weeks since I've had it plugged in since I have limited space for my computers anymore (like 7 and counting). 😁

Can you post exactly what commands you ran so as I can try the same?

Hi, it's:

pciset 0305 1106 55 00000100 00000100

And then "setmul 4" or any number.

thx

I finally got around to messing with this thing some more. I swapped the Voodoo 3 for a 4200ti and a single Voodoo 2.

mockingbird, It seems like the command you shared works fine on my system. I did a couple tests in Win98 and ran 3dMark99 at 400mhz (3x133) and at 1400mhz (10.5x133).

400mhz - Stock Clocked 4200ti:
4333 3D Marks
8942 CPU Marks
52.0 FPS average test 1
37.1 FPS average test 2

1400mhz - Stock Clocked 4200ti:
11002 3D Marks
24704 CPU Marks
136.2 FPS average test 1
92.0 FPS average test 2

I also have this I want to share. I can't remember where I sourced it, but I ran it also before the command to change the multiplier.

1@echo off
2
3rem -----------------------------------------------------
4rem works on VIA KT133 chipset (vendor 1106, device 0305)
5rem sets in register 70h the bits 1 and 2 to 0
6rem and in 75h the bits 0..3 to 0
7rem -----------------------------------------------------
8
9
10rem -----------------------------------------------------
11echo Disable 'PCI Delay Transaction' and 'PCI Master Read Caching'
12echo on VIA KT133 and KT133A
13
14pciset 0305 1106 70 00000110 00000000
15
16
17rem -----------------------------------------------------
18echo Set 'PCI Bus Master Timeout' to 0 Clocks
19echo on VIA KT133 and KT133A
20
21pciset 0305 1106 75 00001111 00000000

ptr1ck wrote on 2022-08-18, 01:44:

mockingbird, It seems like the command you shared works fine on my system. I did a couple tests in Win98 and ran 3dMark99 at 400mhz (3x133) and at 1400mhz (10.5x133).

Fantastic, thanks... I will conduct some further testing... I wonder if it might have to do with my older board revision though.

I've been playing with this more lately. The machine doesn't like to restart fully when I do it from Windows if the multiplier is lowered. I haven't tested this extensively as it's just a minor annoyance.

I swapped in my V2 SLI setup.

Managed to get GTA running beautifully in DOS at 400mhz (3*133) with Glide. Used the OVL from my tribute driver set for Voodoo 2s.

ptr1ck wrote on 2022-09-04, 18:52:

I've been playing with this more lately. The machine doesn't like to restart fully when I do it from Windows if the multiplier is lowered. I haven't tested this extensively as it's just a minor annoyance.

I swapped in my V2 SLI setup.

Managed to get GTA running beautifully in DOS at 400mhz (3*133) with Glide. Used the OVL from my tribute driver set for Voodoo 2s.

I'm pretty certain that the first revision of the KK266+ (and the vanilla KK266 - though I have one waiting for re-capping so I can test it at some later point) has this issue and revision 2 is ok.

I say this because I was wrong about the two chips at the bottom left corner on the revision 2 board being the SMD version of the GAL chips on the REV 1... They are in fact a hard-wired logic version of replacement for the GAL on the revision 1...

It's either that that's causing the issue trying to lower the multiplier on the Revision 1 or perhaps an old revision of the KT133A -- which I doubt.

So definitely only go with a revision 2 of this board of you plan on using it with a Corvette or a modded unlocked Palomino.

mockingbird, just to clarify, everything works right until I restart. The machine shuts down, but when it would normally show the POST screen, it is just like no signal to monitor. I can run PCISET and SETMUL from within Windows and it will drop to 400mhz without issue and run benchmarks just fine. I haven't tried 300 yet because I'm too lazy to pull the cover and switch the jumper. I plan on wiring up a toggle switch to the outside of the case for that eventually.

ptr1ck wrote on 2022-09-04, 20:16:

mockingbird, just to clarify, everything works right until I restart. The machine shuts down, but when it would normally show the POST screen, it is just like no signal to monitor. I can run PCISET and SETMUL from within Windows and it will drop to 400mhz without issue and run benchmarks just fine. I haven't tried 300 yet because I'm too lazy to pull the cover and switch the jumper. I plan on wiring up a toggle switch to the outside of the case for that eventually.

Yes I have that issue as well sporadically on my KT7A, no big deal. Just hit reset.

Ah ok, please do let us know.

I think I did get 400Mhz working properly, but it would always crash when trying to go to x3 multiplier or maybe I'm misremembering things and it also crashed at x4 multiplier.

Added an Audigy 2 OEM to this build for EAX goodness in Windows; perhaps a few more FPS to be gained as well over the ES1869. I have some major consternation over picking the Audigy 2 over my beloved Turtle Beach Santa Cruz. I'm still on the fence... I've never been a fan of Creative for many reasons, but it seems to be the best option other than overpriced Vortex 2s. I read a ton of conflicting reviews of EAX implementation on the Santa Cruz.

I also forgot to try true 300mhz out when I had it open. /facepalm

Update: 300mhz works fine, even from a batch file sitting on the desktop.

I wonder if there is some utility that would let me control it from Windows?

this thread encouraged me to mod my own kk266-r plus bios. i was hoping that unlocking the 'pnp bios installed' setting would help get my dos soundcard drivers working, but the mandatory intel pnp configuration manager still locks up my system regardless of the pnp os setting. in any case, i updated the internal highpoint 37x raid bios to v2.351 (the last version released and harder to find than v2.35) and unlocked the following settings:
- report no fdd for win95
- agp ws write
- agp ws read
- resume state on power loss
- pnp os installed
- disabled 2.5v on pc health status

i attached the bios (KKPR0826_2351_MOD.7z) here for any that care to use it. it is ONLY for the plus and raid version of the board--the non-plus and non-raid revisions use a different bios.

edit: performed the same modifications for the kk266 plus non-raid bios and attached here (KKP0826_MOD.7z).

That's awesome stanwebber. I need to update this thread with some pictures since I rebuilt the system in a Lian Li PC-7B Plus II. I've been too consumed with other obligations to be able to enjoy this machine.

attached kk266 plus non-raid bios with all hidden menu options enabled regardless of whether it is advisable to change them or not. all un-hidden settings are still set to the original defaults, but i have not flashed this bios myself so fair warning.