gerwin wrote:

@Glitter: Did you try that new MII-366GP?

Well, I have an issue with the delivery of the heatsink that was lost by the carrier. I will go to see my brother to look in his old stock what I can find.

And of course, now that I bought nearly everything there is a very good offer on Ebay:
http://www.ebay.fr/itm/Sockel-7-System-K6-III … OUAAOSwdGFYzoLV

That is a good deal. The seller must really need the money! The non-plus K6-III chips weren't that common place.

I thought it was the III+ chips that were the rare ones.

For 450 MHz, I think the rarity is as follows:
K6-III AFX - 2.2 V
K6-III AHX - 2.4 V
K6-III+
K6-II+

Of course, K6-III-475, K6-III+ 500, or K6-III+ 550 top that list.

What's the deal with the Cyrix MIIv 333 with the 4X multiplier? You mentioned in a PM that to get the 4X multiplier working you had to update your BIOS. Why is that exactly? Also, have you tried overclocking this model?

Looking back at my emails from Jan, here are a few excerpts.

"Your BIOS has support for Cyrix 6x86MX/MII multipliers up to 3.5x, but no Cyrix MII 4.0x multiplier support
...
[after his BIOS mod]
The 4x multiplier detection on the Cyrix MII works now. I took care of all the possible Cyrix multipliers (2.0x, 2.5x, 3.0x, 3.5x, and 4.0x) such that the detection and the calculation of the Busspeed (FSB) is correct so the chipset programs its timing register correctly in each case."

My AZZA PT-5IT and Biostar 8500TTD boards, which are based on the 430TX chipset, both needed this BIOS mod for the 4x multiplier to work. I don't remember what the symptoms were prior to the mod except that 4x didn't work at all. Either the computer did not power on, or it hung at POST, or it kept running at 3.5x - don't recall which.

So on the MII, the 4X multiplier was set by CPU registers and not by a dedicated pin on the CPU? Correct?
Assuming I could get my system to boot up, would it be possible to alter the register settings via software?

Did you look at the MII's databook? I do not beleive the multiplier is set via registers. They seem to be set by the CLKMUL input pins on the CPU.

The Clock Multiplier inputs are sampled during RESET to determine the MII CPU core operating frequency.
CLKMUL2 CLKMUL1 CLKMUL0
If = 100 core/bus ratio is 2.5
If = 101 core/bus ratio is 3.0
If = 110 core/bus ratio is 2.0 (default)
If = 111 core/bus ratio is 3.5

For 4.0
CLKMUL2 CLKMUL1 CLKMUL0
010
the rest are "reserved"

"The CLKMUL pins have internal pull-up and pull down resistors to define the default ratio. Therefore the default setting indicates which mode the CPU will operate in if the CLKMUL are not driven and left floating."

From what Jan wrote, it seems the BIOS needs to be aware of the CPU's operating speed to set some timing correctly.

That's pretty weird. Why didn't any other CPUs have this issue with 4X?
Cyrix sure made a big deal about 4X multipliers. (cough cough...5x86)

Well, if it doesn't work at 4x, you could request Jan to mod your BIOS for 4x support.

Sadly, it's not an option because the board I'm using is based on some kind of weird homebrew Acer BIOS from a first generation Pentium. I had already asked him earlier about adding support for K6+ chips.

Have you tried any Cyrix 6x86 CPU in this board?

No. I don't own any. This MII will really be my first experience with a 6x86.

Also, keep in mind that I'm running on a very early Socket4 platform with a Powerleap converter.

When will you be able to test it? I'm kinda looking forward to this.

Mid-late June.

Hi all

I haven't forgot about you, don't worry. 😀

So, I had some diffilculties with the first MB I bought, an ATC-5220 which refuse to boot with FSB above 66MHz, with a Cyrix MII-366 or a K6-2 400. Either it has a real problem either I'm a total idiot.
I decided to buy another SS7 MB, and discovered the brand TMC (not distributed in France obviously) with the TI5VG+.
The K6-2 400 worked perfectly ...... but not the MII-366 🙁

I could use a MII-333 but decided to try to find a better CPU and found the advice to forget about ebay and check offers on cpu-world. And quickly, I was able to buy a MII-400 running at 2.2V for a cheap price (32$ vs around 75$ for such Cyrix cpus on ebay).
Of course, now, the main reason of the tread is obsolete 😜
BTW I tried to overclock the cpu from 3*95 to 3*100 and I coudn't boot if I don't change the voltage 😁

And then I ran into the main difficulty to build that PC, to find 2 IDE ribbon with 40 holes. Seriously, the first 3 sellers promised me it was 40 holes before sending me a 39 holes ribbon. 😵
So, special thanks to techdnauk for brand new IDE cables 😀

I manage to install Win98SE even if I'm really rusty with such old system, except that I still haven't manage to make USB work properly. I haven't tried Beos/Haiku yet.

I haven't decided the case for it, I would like a desktop for ATX motherboard and I'm surprised to see that there isn't much model available.

What voltage are you running the MII at with 300 MHz? And what voltage with 285 MHz?

Solution for the 39 hole IDE ribbon cable: Use a very small flat head screw driver to bore a hole into the plastic connector with the covered hole. Sometimes they actually just use plugs that you can remove with the edge of a sharp knife.

feipoa wrote:

What voltage are you running the MII at with 300 MHz? And what voltage with 285 MHz?

Normal voltage at 285 MHz, 2.2V. It's true that I could try 2.1V but with such a small margin for OC, I fear that I have no margin neither for undervolting.
I tried at 2.4V (not the 2.3) and stopped at POST, I haven't tried to use windows with these settings yet.

I would give 2.1 V a shot. I sorta recall these CPUs not liking higher voltages.