I found (totally by accident) that you can run k6 CPUs at 2x w/o it turning into 6x. I did this on two different MVP3 boards while incorrectly setting the 5.5x jumper setting. This will run your CPU at 2x100MHz or 2x66 (depending on what FSB you set). Most of my boards do 60MHz FSB as well, so the lowest I got was 120MHz on a 450MHz k6-2. Try some random multiplyer jumper combinations and see what happens - I pretty sure there's no risk of hurting the board or the CPU.

P.S. At 2x multiplier all my boards report CPU speed as 211MHz

According to your settings, i've tested a second time with all possible multiplier settings at my VIA Apollo MVP3 board:
With multiplier high to low:
68.5Mhz(FSB) * 5.5 (multiplier setting acc. to PBC) = 377Mhz
68.5Mhz * 5.0 = 343Mhz
68.5Mhz * 4.5 = 308Mhz
68.5Mhz * 4.0 = 274Mhz
68.5Mhz * 3.0 = 206Mhz
68.5Mhz * 2.5 = 171Mhz
68.5Mhz * 2.0 = 411Mhz (This multi will interpreted as 6.0x)
68.5Mhz * 1.5/3.5 = 240Mhz (This multi will interpreted as 3.5x)

I cannot get the results you have got with your MVP3 boards...

Well, more random multiplier settings are not possible! There are three jumper (1/2/3 which can be switched On or Off for the FSB ; and 4/5/6 which can be switched On or Off for the multiplier). Overall there are 8 possibilites for the jumpers 4/5/6 , or not?
I think that i've tried all possbile settings for this CPU. Next week I'll get also an Intel P54CS CPU which i can test.

I got these clocks on a Lucky Tech P5MVP3 and on an Aopen AX59PRO. I'll retest when I have the time

AFAIK the motherboard can't do anything about multipliers. Those jumpers are mapped directly to pins that set the multiplier in the CPU socket itself. What could change from motherboard to motherboard is that it detects the wrong multiplier (when it doesn't know the processor model) and displays the wrong text, but it wouldn't really change the speed at which the processor is running.

Can 1.5 and 2 be selected on K6-2+ and III+ CPUs using SETMUL?

^^ 1.5x doesn't work, but 2x works with Setmul, in addition to 6x.

That's how Phil achieved that clock, then.

Regarding the situation that the AMD K6-2 interprets the multiplier 2x with 6x, and with the setting '1.5x/3.5x', the post-screen will show anyway the clock with '3.5x * selected FSB':
It is possible to install a pentium 150MHz (P54CS) (original clock: 2.5x * 60MHz FSB) in my super-socket 7 board? Well, in my manual only the four pentium's with 66MHz FSB are expained: 100MHz, 133MHz, 166MHz and 200MHz.
According to my last test's, i was not able to set a 66Mhz FSB. Only 68MHz was possible and with the original multiplier, the P150 is then overclocked to 170Mhz.
What would be with the settings '2x' and '1.5x/3.5x'? Will the P150 be set with the correct multiplier, e.g. 1.5x and 2x to underclock a little, or it's possible to destroy the cpu with 3.5x and 6x?

The multiplier is interpreted by the CPU. The Pentium Classic only recognizes 1.5x, 2.0x, 2.5x and 3x (and I believe early variants max out at 2x), no way to set it above or below that. Super Socket 7 specify one additional pin for selecting the multiplier so processors like the K6-2 can recognize more multipliers, but with a Socket 7 processor they will be ignored and a corresponding multi from 1.5x to 3x will take effect.

alexanrs wrote:

The multiplier is interpreted by the CPU. The Pentium Classic only recognizes 1.5x, 2.0x, 2.5x and 3x (and I believe early variants max out at 2x), no way to set it above or below that. Super Socket 7 specify one additional pin for selecting the multiplier so processors like the K6-2 can recognize more multipliers, but with a Socket 7 processor they will be ignored and a corresponding multi from 1.5x to 3x will take effect.

This explains it all basically.

In other words, Pentium 1 MMX was made when there were 2 jumpers for determining it's multiplier.
K6-2 had 3, but Pentium 1 MMX simply doesn't know it, it will ignore it completely, as for it, it doesn't exist 😜