Used to overclock, but now problems

Rich Sias wrote:

"David Maynard" wrote in message

You're doing this the super hard way as you KNOW you have a problem with
the Syba card. Get RID of it and get the basics running right first.
THEN try to debug that thing.


Well that is sorta what I got around to doing today.
I unplugged it and made the 3rd drive a slave to ch2 on onboard controller,
just for testing.
66 100 103 105 110
DOS 7.0 boots boots boots bts BIOS hangs
Win95 boots boots boots hangs BIOS hangs
XP boots hangs hangs notst "
OS/2 boots boots boots rough "

DOS hung on 100 @ 31 deg C, added 6" fan 26 deg C is above. Room is 59 F
today. I tried two fans and at 22 deg C BIOS still wont finish @ 110.
Probably a lost cause there no matter what.

You'd have a better chance with better heatsinks and a Vcore of either
2.1, 2.2, or 2.3. But, hey, don't take my word for it just because I'm
typing this with a pair of them OC'd to 500 at 2.1 Vcore on a BP6.

The one I gave the link to was just the cheapest thing I could find that
would be an improvement, not the 'best'.


Errors recieved with XP with SYBA card were non specific i.e. no xx.dll
identified etc.
STOP: 0x0000007F (0x0000000D 0xC0000034 0x00 0x00 )

With 3rd drive mounted as slave:
STOP: 0x0000007B (0xF894D63C 0xC0000034 0x00 0x00 )

Ugggghhh, I tried to set for 2.4 volts but I couldn't even boot to level of
quotes below. I musta changed SOMETHING, I just don't know what it was. I
tried to set for 75 Mhz but even that has problem with BIOS, it wont let it
boot, forces 66 only. It was working at 75 only 2 days ago.

Try some other things until I get it later.

Rich

"Rich Sias" wrote in message
...

"David Maynard" wrote in message
You're doing this the super hard way as you KNOW you have a problem with
the Syba card. Get RID of it and get the basics running right first.
THEN try to debug that thing.

Well that is sorta what I got around to doing today.
I unplugged it and made the 3rd drive a slave to ch2 on onboard
controller,
just for testing.
66 100 103 105 110
DOS 7.0 boots boots boots bts BIOS hangs
Win95 boots boots boots hangs BIOS hangs
XP boots hangs hangs notst "
OS/2 boots boots boots rough "

DOS hung on 100 @ 31 deg C, added 6" fan 26 deg C is above. Room is 59 F
today. I tried two fans and at 22 deg C BIOS still wont finish @ 110.
Probably a lost cause there no matter what.

Errors recieved with XP with SYBA card were non specific i.e. no xx.dll
identified etc.
STOP: 0x0000007F (0x0000000D 0xC0000034 0x00 0x00 )

With 3rd drive mounted as slave:
STOP: 0x0000007B (0xF894D63C 0xC0000034 0x00 0x00 )

Whooo Hooo, I found the problem.
1st I reset to defaults and I could once again set for 75Mhz. I checked and
couldn't identify what was different. So not sure what was off.
2nd: I finally found an old pdf file of the motherboard manual. JP45 is
"Factory Reserved" in manual removing jumper forces 100Mhz FSB. I was using
the force 100Mhz jumpers on the slotket boards. DUH ! They were automatic or
100. Leaving on auto and opening JP45 on MB and now I ran the experiments of
all OS's installed. Set vcore = 2.4, reading 2.36 to 2.38v. With big 6"
personal fan blowing on CPUs, starting temp 20c. Ran XP, Prime95 torture,
temp rose to 27c. Started outlook express and this message now at 25c,
Prime95 at Test 244.
66 100 103 105 110
112 115
33 33 34 35 36
37 38
DOS 7.0 boots boots boots boots Boots
Boots
Win95 boots boots boots btoos error
XP boots hangs boots boots blank"
OS/2 boots boots boots boots error

Win95 error was: Not enough memory in loading windows (all i can remember)
XP error was: After select XP from boot menu, screen goes black and stays
that way for 11 minutes. Rebooted.
OS/2 error was: The file OS2KRNL is not acceptable , Insert diskette &
restart.

I am speculating that the disk access is having trouble, this is the onboard
IDE with Intel chip. Should I be changing the RAS & CAS numbers from 3 to 2?
If so which item should I experimet with first ? Theres one with "delay" in
its name, "precharge?" in its name and one other. There is also 2 more
involving 16/8 bit to 32 bit delay? valuses 2 & 3. I doubt if changing these
will have any benefit for me. There are a whole lot of disabled and one
enabled on that page for things to be on or off. This is an Award BIOS 440BX
chipset.

I am able to run 75 Mhz just fine at 37Mhz PCI speeds. Cannot run 83Mhz at
41Mhz PCI speeds, periphials will not POST. I had some trouble at 112 with
POST and the on board Adaptec chip booting. Not sure if I can ever go with
115.

CPU is 366 Celery. 105 = 578 & 110 = 605 ( I may never get ALL componets to
work smoothly together at this level but desire to try in next 2 days, then
on to doing taxes and job searching.

Rich
seeking memory tweeks to push up to 110 if posible.

"Rich Sias" wrote in message

66 100 103 105 110
112 115
33 33 34 35
36 37 38
DOS 7.0 boots boots boots boots Boots
Boots
Win95 boots boots boots btoos error
XP boots hangs boots boots blank"
OS/2 boots boots boots boots error


I have experimented further. Of some reference web sites I found out that
there is some preference to adjusting CAS/RAS lines.
1. CAS Latencey
2. RAS-CAS Delay
3. CAS Precharge Delay

I first tried tweeking on 110 Mhz. Things just didn't work out to well
there. I eventually resorted to tweeking 105 Mhz and booting to DOS and
running Checkit and running a memory test. The thorough test failed most of
the time with THREE different memory sticks. The failures were random. They
occured with varying amounts of "2" and "3" in above items. The only sure
fire way was to set system to 100Mhz then it passed totally 15x tests of
thorough memory testing. The memory is 124Mhz 8ns, 133Mhz 7.5ns, and 133Mhz
7.8ns. I found this from AIDA test. Can anyone confirm or deny that the
CPUs are pusing their limitations at the 105 and up level? I rather suspect
that the memories are OK based on the above numbers.

I am actually running in XP at 105 (578Mhz) and don't seem to notice errors
anywhere. Is there some kind of parity error repair going on to keep the
errors away from interference ? If so is there a way to review the XP logs
of "error hits" somewhere ? I would like to push it higher than 550Mhz, but
now looks like that may be the limit I have to live with.

The only alternative is to purchase a matched couple of Tualatin 1.4Ghz and
put them into the Powerleaps I own but haven't put to use yet. I will
probably need to get with eSupport.com to update my bios to get them to work
though. Running @ 140 could get 1.47Ghz or @150 - 1.575Ghz. Do these chips
run this high ? My other choices would be 1.26 @140-1.33, @150-1.425. Hmmm
that is barely up to the 1.4@133 and that is as high as my MB goes. Anyone
reccomend one over the other ?

Rich Sias

Rich Sias wrote:

"Rich Sias" wrote in message


66 100 103 105 110

112 115

33 33 34 35

36 37 38

DOS 7.0 boots boots boots boots Boots

Boots

Win95 boots boots boots btoos error
XP boots hangs boots boots blank"
OS/2 boots boots boots boots error



I have experimented further. Of some reference web sites I found out that
there is some preference to adjusting CAS/RAS lines.
1. CAS Latencey
2. RAS-CAS Delay
3. CAS Precharge Delay

I first tried tweeking on 110 Mhz. Things just didn't work out to well
there. I eventually resorted to tweeking 105 Mhz and booting to DOS and
running Checkit and running a memory test. The thorough test failed most of
the time with THREE different memory sticks. The failures were random. They
occured with varying amounts of "2" and "3" in above items. The only sure
fire way was to set system to 100Mhz then it passed totally 15x tests of
thorough memory testing. The memory is 124Mhz 8ns, 133Mhz 7.5ns, and 133Mhz
7.8ns. I found this from AIDA test. Can anyone confirm or deny that the
CPUs are pusing their limitations at the 105 and up level? I rather suspect
that the memories are OK based on the above numbers.

I am actually running in XP at 105 (578Mhz) and don't seem to notice errors
anywhere. Is there some kind of parity error repair going on to keep the
errors away from interference ? If so is there a way to review the XP logs
of "error hits" somewhere ? I would like to push it higher than 550Mhz, but
now looks like that may be the limit I have to live with.

The only alternative is to purchase a matched couple of Tualatin 1.4Ghz and
put them into the Powerleaps I own but haven't put to use yet. I will
probably need to get with eSupport.com to update my bios to get them to work
though. Running @ 140 could get 1.47Ghz or @150 - 1.575Ghz. Do these chips
run this high ? My other choices would be 1.26 @140-1.33, @150-1.425. Hmmm
that is barely up to the 1.4@133 and that is as high as my MB goes. Anyone
reccomend one over the other ?

All the PIII-S 1.4Ghz chips I've tested (about 10) had no problem
running at 1.575Ghz on default voltage, but I've never managed to get a
pair of them stable past 140Mhz FSB. The FSB limit also applies to
PIII-S 1.26Ghz pairs IME.

I suspect the problem is due to noise on the bus request signals between
processors - the Intel specification for the maximum distance these
signals should travel is virtually impossible to achieve on a Slot-1
motherboard with adapters installed, given there has to be space between
slots for the heatsink and fan, and increasing FSB would tend to reduce
signal quality.

I have a couple of dual processor systems running stable at 140Mhz FSB
using Slot-T adapters, but the same hardware could only manage 133Mhz
FSB with Powerleap adapters. The bus request signal paths appear to be
slightly longer on the Powerleaps....

BTW, what *is* your motherboard?

P2B

"P2B" wrote in message


BTW, what *is* your motherboard?

P2B


Shuttle HOT-649a
Dual slot - 1 with MSI slotkit so far. Good only for up to 600Mhz chips. I
have a pair of PowerLeaps waiting on sidelines but 1.4s are still expensive.
If I have to live a limit of 133 or most 140 I may rather have the 1.4s.
FSBs 66, 75,83, 100, 103, 105, 110,112, 115, 120, 124(1/3&1/4) 133, 140, &
150

Rich

Rich Sias wrote:

"P2B" wrote in message


BTW, what *is* your motherboard?

P2B



Shuttle HOT-649a

Not found in the Powerleap compatibility database - odd, you'd think
someone would have tried it by now....

Google found this:

http://ftp.sandpile.org/post/msgs/20004914.htm

Doesn't sound too promising - similar symptoms to what I saw with
Powerleaps and 1.4s on the Asus P2B-DS at 140Mhz, occurring at 133Mhz on
the HOT-649a. The P2B-DS has a very similar physical layout, CPU slots
are 58.5mm apart. What's the slot separation on the Shuttle?

Dual slot - 1 with MSI slotkit so far. Good only for up to 600Mhz chips. I
have a pair of PowerLeaps waiting on sidelines but 1.4s are still expensive.
If I have to live a limit of 133 or most 140 I may rather have the 1.4s.
FSBs 66, 75,83, 100, 103, 105, 110,112, 115, 120, 124(1/3&1/4) 133, 140, &
150

Rich

Rich Sias wrote:
Whooo Hooo, I found the problem.
1st I reset to defaults and I could once again set for 75Mhz. I checked and
couldn't identify what was different. So not sure what was off.
2nd: I finally found an old pdf file of the motherboard manual. JP45 is
"Factory Reserved" in manual removing jumper forces 100Mhz FSB. I was using
the force 100Mhz jumpers on the slotket boards.

Doesn't matter where the '100 MHz FSB' is set. 100MHz is 100MHz

DUH ! They were automatic or
100. Leaving on auto and opening JP45 on MB and now I ran the experiments of
all OS's installed. Set vcore = 2.4, reading 2.36 to 2.38v. With big 6"
personal fan blowing on CPUs, starting temp 20c. Ran XP, Prime95 torture,
temp rose to 27c. Started outlook express and this message now at 25c,
Prime95 at Test 244.
66 100 103 105 110
112 115
33 33 34 35 36
37 38
DOS 7.0 boots boots boots boots Boots
Boots
Win95 boots boots boots btoos error
XP boots hangs boots boots blank"
OS/2 boots boots boots boots error

Win95 error was: Not enough memory in loading windows (all i can remember)
XP error was: After select XP from boot menu, screen goes black and stays
that way for 11 minutes. Rebooted.
OS/2 error was: The file OS2KRNL is not acceptable , Insert diskette &
restart.

I am speculating that the disk access is having trouble, this is the onboard
IDE with Intel chip. Should I be changing the RAS & CAS numbers from 3 to 2?
If so which item should I experimet with first ? Theres one with "delay" in
its name, "precharge?" in its name and one other. There is also 2 more
involving 16/8 bit to 32 bit delay? valuses 2 & 3. I doubt if changing these
will have any benefit for me. There are a whole lot of disabled and one
enabled on that page for things to be on or off. This is an Award BIOS 440BX
chipset.

I am able to run 75 Mhz just fine at 37Mhz PCI speeds. Cannot run 83Mhz at
41Mhz PCI speeds, periphials will not POST. I had some trouble at 112 with
POST and the on board Adaptec chip booting. Not sure if I can ever go with
115.

You're probably using an old UDMA33 drive. They won't overclock to 41Mhz
PCI. Most won't go past 37, or so, and some won't go past 34.


CPU is 366 Celery. 105 = 578 & 110 = 605 ( I may never get ALL componets to
work smoothly together at this level but desire to try in next 2 days, then
on to doing taxes and job searching.

Rich
seeking memory tweeks to push up to 110 if posible.

You should probably forget 110.

"David Maynard" wrote in message
Rich Sias wrote:

JP45 is
"Factory Reserved" in manual removing jumper forces 100Mhz FSB. I was
using
the force 100Mhz jumpers on the slotket boards.

Doesn't matter where the '100 MHz FSB' is set. 100MHz is 100MHz

Well, I went back and reconfirmed that the 100FSB jumpers on the MSI
slotkits had errors at 100 Mhz, but the 100FSB (JP45) works great, no
xxx.sys errors!!!


seeking memory tweeks to push up to 110 if posible.

You should probably forget 110.

You are right. Only Dos boots for a short while. Checkit fails pretty quick
every time.

Rich Sias wrote:

"David Maynard" wrote in message

Rich Sias wrote:


JP45 is
"Factory Reserved" in manual removing jumper forces 100Mhz FSB. I was

using

the force 100Mhz jumpers on the slotket boards.

Doesn't matter where the '100 MHz FSB' is set. 100MHz is 100MHz


Well, I went back and reconfirmed that the 100FSB jumpers on the MSI
slotkits had errors at 100 Mhz, but the 100FSB (JP45) works great, no
xxx.sys errors!!!

Well, something else is at play because the only thing the slotket 100 MHz
FSB jumper does is disconnect the BSel0 line: connected is low = 66.6 and
disconnected is high=100. And that's all the motherboard jumpers do too,
raise or lower those BSel lines (there are two if the mobo supports 133 MHz
FSB, BSel0 and BSel1). AND, that is all the processor does: a 100 Mhz FSB
processor simply has no connection to the BSel0 line, so it stays high, and
a 66 MHz FSB processor connects the pin to ground. If it took a 'mobo
jumper' to do it then processors wouldn't be "plug and play" per the Intel
specs.

I.E. The clock generator couldn't care less who, or what, does or doesn't
connect the pin to ground, nor where. It's either up, or it's down.

I'll bet you're inadvertently setting something you don't think you're
setting: As in there's more than one jumper, and more than one position, on
both the slotket and the mobo.




seeking memory tweeks to push up to 110 if posible.

You should probably forget 110.


You are right. Only Dos boots for a short while. Checkit fails pretty quick
every time.