🚀 go-pugleaf

Mick Costa <mick.costa@fmr.com.no.spam> wrote:

>Keith Rohrer wrote:
>> 
>> > Gary lake wrote:
>> > > I have a K5PR133 on the SIS-5571 based GA-586S MB. This MB supports a
>> > > bus speed of 75Mhz (and possibly 83Mhz).
>> > >
>> > > The manual is a bit confusing on what is required for this speed so I
>> > > tried the speed settings for a Cyrix 200+ but the system locks up just
>> > > after a startup msg of AMD K5PR133 at 110Mhz.
>> Read what's silkscreened on the board.  The jumper blocks over by the
>> SIMMs seem to control a PCI bus multiplier/divisor (one setting for 50
>> MHz, another setting for 75, and a middle setting for the rest of the
>> speeds), while the jumper blocks in the general vicinity of the CPU seem
>> to control the CPU bus speed.  As I recall, the K5-100's ignore the
>> clock multiplier pins; I'm not sure if the K5-PR133's do as well.
>> 
>> > > Also I thought the K5 family had a fixed multiplier but for normal setup
>> > > ie 66Mhz bus, the setting in the MB manual is for x2 ??
>> AMD's always said use the "Pentum-133" settings; apparently the chips
>> don't care.
>> 
>> > > Has anyone sucessfully run this MB/CPU combination at 75Mhz.??
>> I can run it with minimal wait states at 66 MHz, but at 75 MHz linux
>> won't load, even with an extra wait state added to all options and the
>> hard drive (Fujitsu M1638 EIDE) dumbed down to PIO 3.  Either I get a
>> CRC error, or lilo crashes.  I figure if I need to add more than one
>> wait state, I can't possibly get any performance improvement, but the
>> EIDE subsystem could be faulty somewhere.
>> 
>> > The jumper I cannot figure out is JP11 which is set to 1-2 for
>> > everything except the P200+.  Did you set it to 2-3 or leave it at 1-2
>> > for the K5PR133?  I find it strange that it get's left at 1-2 for 50,
>> > 55, 60, and 66 Mhz bus speeds and only changes for the P200+ (75 Mhz but
>> > speed).  That made me think that it has more to do with the P200+ than
>> > the bus speed or the fact that it is a Cyrix.
>> I don't remember the jumpers by number, but the ones off by the SIMM
>> sockets were changed in such a way as to seem to be a PCI speed control;
>> the others were all over by the CPU.  From my other motherboard
>> experiences, usually one "lone" jumper that was set to 1-2 is set to 2-3
>> to manage the highest speed or two the board can be clocked at.
>> 
>> > Anyone know the facts?
>> Not I; certainly it would've been nice if gigabyte had included jumper
>> settings organized by what each jumper does as well as by CPU.  Also
>> nice would've been BIOS setup entries to disable stuff like the PS/2
>> mouse port and the giving of an IRQ to my video card...the plug 'n pray
>> stuff is really heavy on the pray...
>> 
>>         Keith

>First, I don't have the answers, but I've run into very similar
>problems.  I have the GA-586S with a Pentium P166 MMX chip.  I am hoping
>to at least run this chip at 75x2.5 (187.5Mhz) - that's why I bought the
>board.  I figured for $112 it was worth the gamble.  I have had no
>success with 75 Mhz except at startup I am told the speed the CPU is
>operating at.  With different combinations of jumpers, I have seen CPU
>speeds of 150, 166, 187, 200, and 210 (only 166 and 200 actually finish
>booting up).  I don't understand the 210 since I thought the multipliers
>were 1.5x, 2x, 2.5x, and 3x.  As far as I can tell, 210 would be
>60x3.5.  I get all the BIOS startup screens, then it freezes and
>displays for different attempts (in no quotes): "P", "PC", or "PCI." 
>What could this be telling me?

>I have not played with all the jumper settings over by the SIMMS
>(JP2-4).  The manual says that the board supports PCI speeds of 25, 30,
>and 33 Mhz.  It doesn't tell you how that happens.  And yes, JP11 goes
>from a normal 1-2 to 2-3 to get the 75 Mhz speed.  I did get the 187.5
>reading from the normal Cyrix P200+ settings (except having JP8 set for
>2.5x).  I may need to keep it there and play with the other jumpers.

>From looking at the JP2-4 jumper tables, I summarize them as follows
>(including where *I* would put them for 83 Mhz (I take no responsibility
>for anyone attempting to do anything with their motherboards):

>JP4 (ON)  : 50/55/60/75 Mhz
>JP4 (OFF) : 66/(83) Mhz

>JP3 (ON)  : 55/60/66/(83) Mhz
>JP3 (OFF) : 50/75 Mhz

>JP2 (ON)  : 50 Mhz
>JP2 (OFF) : 55/60/66/75/(83 Mhz)

>I figure that 83 Mhz should go with 66 Mhz since they are most closely
>related (i.e., 33x2 = 66; 33x2.5 is approx. 83).

>I have been told that possible 83 Mhz combinations are:

>JP10 2-3, 5-6; JP11 2-3, or
>JP10 1-2, 5-6; JP11 2-3.

>My inclination would be the bottom one since that is closest to the P166
>setting. 
>However, both imply I get get anything to work with JP11 set to 2-3
>(which hasn't happened even when I try 75x2!). Could this be a
>Cyrix-specific thing?  Is this related to my strange 3.5x (210 Mhz)? 
>Too many jumpers so little time.

>Another thing I didn't do was change the ISA Bus Clock Setting PCICLK in
>the BIOS. For 60 Mhz and 66 Mhz chips, it is set to PCICLK/3, but for 50
>MHz chips it is supposed to be set to PCICLK/4.  I assume that for the
>P200+ it is set to the latter.  When going to the 75 Mhz speed, maybe I
>should have changed that.  I also didn;t change any of the POI
>settings.  I am running v1.2 o of the BIOS.  It gives me different (I am
>not sure better) results with CTCM.  Faster main memory but slower
>DOS/WIN.  There are also a coupe of added settings in the BIOS, but they
>don't seem todo much for me (and of course I have no documentation on
>them).

>This weekend means another attempt to hit the higher speeds.  Otherwise,
>I may just have to go with the new ABIT or AOpen boards...

>Mick Costa

Mick:

FWIW - I'm running a P150 at 166 on a GA586S.
I have had success running at 187.5 by changing the voltage.  JP9
controls the voltage, apparently, even though the official settings
control only dual voltage settings.  187.5 runs fine at JP9 3-4 EXCEPT
that the chip starts running too hot unless I leave the case open, and
I start getting crashes in Win95.
Give it a try, but watch the heat.
Steve