• Please review our updated Terms and Rules here

BIOS setup help

Amigaz

Experienced Member
Joined
Jan 31, 2008
Messages
426
Location
Stockholm, Sweden
Have bult together a nice PC with a motherboard from around 1990 which has an Intel 386DX 33mhz + 8mb RAM
When I got the board it had no cmos battery and the BIOS settings were lost so my problem now is that I have no clue on how to set it up for max performance.
I've never used this BIOS version/type before but I've set up it so I at least have a working at solid system but I cannot figure out how to set the HIOS to run my 386 at 33mhz, the norton sysinfo results below show the poor performance :(

Here's some pic's of the BIOS setup etc (sorry for the crap quality):

p1010340xv0.jpg


p1010344al4.jpg


p1010347zu5.jpg


p1010337ia6.jpg
 
Does pressing the TURBO button on the front of the computer change the performance as shown by Norton SysInfo?
If not (turbo switch not connected to motherboard, or not connected properly), the motherboard may running in non-turbo mode by default.
 
Does pressing the TURBO button on the front of the computer change the performance as shown by Norton SysInfo?
If not (turbo switch not connected to motherboard, or not connected properly), the motherboard may running in non-turbo mode by default.

I have no turbo button on this case...it's too new..it just has a tutbo led indicator and it's on since I've put a jumper on the turbo pins on the motherboard so it's in turbo mode all of the time

Norton Sysinfo would have reported another mhz rating than 25mhz if it wouldn't have been in turbo mode...these 386 motherboard go down to 4-8mhz XT mode when they're in non turbo
 
I have no turbo button on this case...it's too new..it just has a tutbo led indicator
A case with a turbo LED but no turbo switch to go with it. Odd! Must have been produced just at the time when the 'turbo' functionality was being phased out. I guess the LED made users happy that their machine was running as fast as it could.

I've put a jumper on the turbo pins on the motherboard so it's in turbo mode all of the time
Note that on some motherboards, shorting the turbo pins takes the motherboard out of turbo mode.

... these 386 motherboard go down to 4-8mhz XT mode when they're in non turbo.
I try to avoid assumptions. Of course, if the motherboard is indeed in turbo mode, removing the turbo jumper will see the performance drop even further. It's a quick test to rule it out altogether.

... and the BIOS settings were lost so my problem now is that I have no clue on how to set it up for max performance.
SysInfo is showing your 386-33 as a "80386DX 25Mhz". The BIOS can determine the "80386DX" bit quite easily. The BIOS arrives at the "25Mhz" figure by timing how long certain 60386 instructions take to execute. So we can say with good certainty that your CPU is being provided with a 25MHz clock, instead of a 33Mhz clock.

The standard method of generating the CPU clock on 386 motherboard is to divide the oscillator by 2. So on my 386-40 boards, I see an 80Mhz oscillator ("tin can"). I have another 386 board that according to the manual will take either a 386-16 or 386-20 CPU. On that board are two oscillators, 32Mhz and 40Mhz, and you change the position of a jumper (selects which oscillator is used) according to which of the two CPUs is installed.

Do you have a 66Mhz oscillator on your board?

By the way. Regarding the three 'clock' related items shown in the middle of the second BIOS screen shot. The manual for one of my 386 boards shows that the board has the same BIOS (exactly the same screens) as yours. For those three 'clock' items, the manual states, "It is not user adjustable". The three values shown in the screen shot match what is shown in the manual.

The values in the third screen shot of yours (chipset registers) differs slightly from what is in my manual, but that could be because of different motherboard design.
In any case, the procedure shown in the manual to "clear the CMOS RAM" is to power on whilst pressing the INSERT key, continuing to hold down the key until the "screen display appears".
 
A case with a turbo LED but no turbo switch to go with it. Odd! Must have been produced just at the time when the 'turbo' functionality was being phased out. I guess the LED made users happy that their machine was running as fast as it could.


Note that on some motherboards, shorting the turbo pins takes the motherboard out of turbo mode.


I try to avoid assumptions. Of course, if the motherboard is indeed in turbo mode, removing the turbo jumper will see the performance drop even further. It's a quick test to rule it out altogether.


SysInfo is showing your 386-33 as a "80386DX 25Mhz". The BIOS can determine the "80386DX" bit quite easily. The BIOS arrives at the "25Mhz" figure by timing how long certain 60386 instructions take to execute. So we can say with good certainty that your CPU is being provided with a 25MHz clock, instead of a 33Mhz clock.

The standard method of generating the CPU clock on 386 motherboard is to divide the oscillator by 2. So on my 386-40 boards, I see an 80Mhz oscillator ("tin can"). I have another 386 board that according to the manual will take either a 386-16 or 386-20 CPU. On that board are two oscillators, 32Mhz and 40Mhz, and you change the position of a jumper (selects which oscillator is used) according to which of the two CPUs is installed.

Do you have a 66Mhz oscillator on your board?

By the way. Regarding the three 'clock' related items shown in the middle of the second BIOS screen shot. The manual for one of my 386 boards shows that the board has the same BIOS (exactly the same screens) as yours. For those three 'clock' items, the manual states, "It is not user adjustable". The three values shown in the screen shot match what is shown in the manual.

The values in the third screen shot of yours (chipset registers) differs slightly from what is in my manual, but that could be because of different motherboard design.
In any case, the procedure shown in the manual to "clear the CMOS RAM" is to power on whilst pressing the INSERT key, continuing to hold down the key until the "screen display appears".

You mean you never seen an AT case with just a turbo led? :eek:

Have four such cases here, in the early socket 5/7 era these cases were very frequent....

Since my turbo led is on I assume it's in turbo mode? might be the other way around anyway...going to test without the jumper

One of the clock crystals is a 50mhz one so I'm starting to suspect it's someone who just threw in a 33mhz 386 on my mobo

p1010336kp2.jpg


p1010337ei2.jpg


So you mean the settings on the BIOS screen has no effect?
I can change the all.

My 386 mobo lacks CPU cache...that might effect the sysinfo score also maybe?
 
You mean you never seen an AT case with just a turbo led?
No, but there have been periods over the years where I didn't work on personal computers (other types, but not personal).

One of the clock crystals is a 50mhz one so I'm starting to suspect it's someone who just threw in a 33mhz 386 on my mobo
When I saw the 50MHz oscillator, I knew that was the cause. So you have a motherboard designed for either a 386-20 or 386-25. I expected to see a jumper close to the two oscillators, but maybe it is somewhere else on the board, or maybe oscillator selection is done via the BIOS.

Note the "25 Mhz" stamped onto one of the chips of the chipset. The board will operate at a maximum of 25 Mhz, and so replacing the 50 Mhz oscillator with a 66 Mhz will not work (reliably).

So you mean the settings on the BIOS screen has no effect?
No, I said that according to the manual I have, the 'clock' settings in the BIOS can't be changed. If you can change the 'clock' settings, it will be because you have a different motherboard than mine.

My 386 mobo lacks CPU cache...that might effect the sysinfo score also maybe?
It is too long ago for me to remember such stuff. Can anyone else answer that?

I do remember that for a while, some Taiwanese motherboards had fake cache chips and the timer speed was slightly altered so that speed utilities didn't reveal the fake chips. Later, people wrote software that could detect the presence of fake cache chips. I'm sure I still have that software.
 
Hi
It was quite common for the cheap PC sales houses to take
a faster CPU on a board that couldn't use the full speed.
They'd advertise the CPU as being 33MHz but not tell you the
board was only 25Mhz.
This was quite common so I suspect the processor was the original
on that board.
Dwight
 
Hi
It was quite common for the cheap PC sales houses to take
a faster CPU on a board that couldn't use the full speed.
They'd advertise the CPU as being 33MHz but not tell you the
board was only 25Mhz.
This was quite common so I suspect the processor was the original
on that board.
Dwight

Feels like most motherboard back then were cheap, they didn't even bother to name them ;)

Consumer PC hardware didn't become "working quality stuff" imho until the socket 7 boards arrived
Maybe the old XT's were better
 
Back
Top