Reply 60 of 73, by Sphere478
- Rank
- l33t++
Exciting!!! I dug out my 266 2.0 btw if you have any questions.
Oh, btw.
Re: Cyrix 400gp multiplier settings
Exciting!!! I dug out my 266 2.0 btw if you have any questions.
Oh, btw.
Re: Cyrix 400gp multiplier settings
Sphere478 wrote on 2022-08-30, 23:14:Exciting!!! I dug out my 266 2.0 btw if you have any questions.
Oh, btw.
Re: Cyrix 400gp multiplier settings
Thanks!
Well it posted right up at stock clock's and voltage. Left the FSB at 75mhz as 83mhz has never been stable on my PC Chips board. Slowly increased the multiplier and it booted at each speed. I stopped at 75x4.0 as 300mhz was my initial target speed. The bios identifies it as a MII 366 at this clock, so I might have to play around with that.
But I ran through a handful of DOS benchmarks and had zero stability issues. I threw in a Wondows 98 drive that had correct drivers, and with my Voodoo 3 it is currently running a 3dMark 2000 default benchmark. I don't have 3dMark 99 installed on that drive for whatever reason.
But all in all the chip is running super cool and all seems pretty well so far!
Edit: 3dmark 2000 default benchmark: 458!
Sweet!!
So your mobo supports bf2 but not 83fsb? That’s odd. Or are you using a tweaker/bf2 mod?
I think your gp rating will go up if you raise the fsb. So 366gp is probably a decent guestation on the bios’s part.
Sphere478 wrote on 2022-08-31, 03:21:Sweet!!
So your mobo supports bf2 but not 83fsb? That’s odd. Or are you using a tweaker/bf2 mod?
I think your gp rating will go up if you raise the fsb. So 366gp is probably a decent guestation on the bios’s part.
It technically "supports" 83mhz FSB but it is as I understand it, and unofficial overclock of the chipset if it is used past 75mhz.
I've already set little heatsinks on the chips. They don't seem to get warm so I'm guess it's just a stability issue.
It worked about 80% stable with the K63 that I previously had in the board, but was way happier at 75mhz.
I will do some more playing with it when I get the chance. Cheap crappy PC Chips boards are always a new adventure.
Well, I didn't have much time to mess with it. But I did try a few different multiplier settings and try for 83mhz FSB.
I could get the chip to post and boot into Windows at 333mhz, but it wasn't stable. I tried a handful of different sticks of ram and few different PCI video cards. But just wasn't very stable.
I don't really think it is the CPU. The K6-3+ did the same thing, and that chip runs happily at 100x6 in my other two boards.
I just think it is the PC Chips board or the SiS chipset not liking that FSB setting.
But I am happy enough with the chip running at 300mhz. That was my personal goal. Maybe one day I will nab another super 7 board and try to hit some higher numbers.
First chance I get I will dig out my spare case and try to actually build the system. Might even throw up some pics, who knows?
Try replacing the caps.
But semi stable at 333 isn’t unusual for a 2.2v chip.
Try this:
Benchmark the difference between 75x4 and 83.3x3.5
I went with a P200MMX for my Socket 7. Not that much slower than a 233 but Turbo pascal compiled programs don't runtime 200 on it (at least not on mine). Tried a 233 and it would occur in my build.
Sphere478 wrote on 2022-09-02, 04:36:Try replacing the caps. […]
Try replacing the caps.
But semi stable at 333 isn’t unusual for a 2.2v chip.
Try this:
Benchmark the difference between 75x4 and 83.3x3.5
Yeah I tried that. It doesn't really seem to matter the multiplier settings. Any frequency at 83mhz FSB is unstable.
According to what I read while going through this struggle
PC Chips M571 K6-3+ Support - UPDATED WITH BENCHMARKS
It seems that only certain revisions of this board officially support the 83mhz FSB, the other revisions might or might not do it. Just luck of the draw.
I assume it is a voltage thing.
Replacing the caps might get it more stable.
But I'm not too bothered, I plan eventually on picking up another SS7 board, likely a ALi chipset model as I don't have one.
Then I can use one of my VIA boards with this chip and retire the PC Chips guy, or use it to build a 233MMX rig.
gen_angry wrote on 2022-09-02, 13:31:I went with a P200MMX for my Socket 7. Not that much slower than a 233 but Turbo pascal compiled programs don't runtime 200 on it (at least not on mine). Tried a 233 and it would occur in my build.
Yeah I have a 166mmx chip and a Pentium Classic 100 that I could use. Plus I have two K6+ builds if I need more speed.
It is mainly about playing around with a platform that I haven't before.
I do eventually plan on trying to put together a MMX 233 system though.
I would also like to try a Rise MP6 if I can ever find one cheaply enough.
What about a dual 233 mmx? 😀
gen_angry wrote on 2022-09-02, 13:31:I went with a P200MMX for my Socket 7. Not that much slower than a 233 but Turbo pascal compiled programs don't runtime 200 on it (at least not on mine). Tried a 233 and it would occur in my build.
Some mmx200 pentiums have 3.5x multi btw.but most it seems are locked.
Sphere478 wrote on 2022-09-02, 18:03:What about a dual 233 mmx? 😀
gen_angry wrote on 2022-09-02, 13:31:I went with a P200MMX for my Socket 7. Not that much slower than a 233 but Turbo pascal compiled programs don't runtime 200 on it (at least not on mine). Tried a 233 and it would occur in my build.
Some mmx200 pentiums have 3.5x multi btw.but most it seems are locked.
I threw up benchmarks of the mostly completed build. Figured it deserved it's own thread.
A nice modification is two switches for FSB/multi reclocking. One toggle on JP5B to go from 50 to 66 mhz FSB (switches to 1-2 common from 3-2 common), another toggle on JP7A to switch 3-2 common to 1-2 common again. This gives you 100 mhz, 133 mhz, 175 mhz, and 233 mhz. On top of that, you get all the TR12 settings to play with. Going this route yields one of the most versatile machines I have ever tested.
All hail the Great Capacitor Brand Finder
gdjacobs wrote on 2022-09-08, 22:45:A nice modification is two switches for FSB/multi reclocking. One toggle on JP5B to go from 50 to 66 mhz FSB (switches to 1-2 common from 3-2 common), another toggle on JP7A to switch 3-2 common to 1-2 common again. This gives you 100 mhz, 133 mhz, 175 mhz, and 233 mhz. On top of that, you get all the TR12 settings to play with. Going this route yields one of the most versatile machines I have ever tested.
I've done this, and I concur. That said, I find that the multiplier switch really doesn't do much; once you start disabling the most crippling stuff, it really doesn't matter what the internal clockspeed is, in my experience.
I actually don't have any great data on that -- mostly on FSB manipulation. I should take my P55 time machine off the shelf and do some testing with multipliers.
All hail the Great Capacitor Brand Finder