VOGONS


First post, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

I’ve noticed the last time I googled for this that not only on the cpu it’s self but online there seems to be very little info for what settings to set these cpus at.

Correct me if I’m wrong but is it like this:

Mp6 2.8v
IDragon 2.0v

And then. You kinda just guess the mhz… 😂

Also, who knows what multipliers were supported?

Last edited by Sphere478 on 2022-03-25, 12:26. Edited 2 times in total.

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 1 of 3, by Nemo1985

User metadata
Rank Oldbie
Rank
Oldbie

I just have the iDragon, I can confirm it works with v2.0 and 200 but I was able to oc it to 250 mhz.
The original mp6 (according to wikipedia) had a frequency from 166 to 200 mhz, it shouldn't be that difficult since the pr rating was written on the cpu.

Reply 3 of 3, by Sphere478

User metadata
Rank l33t++
Rank
l33t++

Gonna update this reply with the multiplier settings supported by the 500mips labeled idragon.

On my freeway motherboard I have switches 5,6,7 I assume these correspond to pins bf0,bf1,bf2
With that assumption in hand,

1=on 0=off

Bf0, bf1, bf2
0,0,0 = 3.5x
1,0,0 = 2x
0,1,0 = 3x
1,1,0 = 2.5x
0,0,1 = edit: 3.5x
1,0,1 = 2x
0,1,1 = 3x
1,1,1 = 2.5x

It is very interesting that the 0,0,1 setting gives a post code, suggesting it may actually be a multiplier beyond the supported clock of the chip?? Unsure. I was running 66fsb

I also tried a little overclocking
After probing the multipliers, I set fsb back (up) to 100mhz. And programmed in the 2.5x setting.

It posts, and boots memtest.
I powered down and tried 3x, and powered it on, the system hangs immediately with audable beeps, so it seems that the mysterious multiplier was possibly 4x because that would give 266mhz which would make sense, because 0,0,1 got further into post than 300mhz did.

At 250 mhz in memtest I am seeing errors.

Lemme bump the voltage and see what we get

Okay, 2.3v 66mhz bus. I got the chip to post and load memtest. Turns out it is the 3.5x setting which actually shows that there are no bf2 programmed multiplier combinations on this chip. Why it crashed earlier only on that setting I am unsure.

Attachments

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)