which is known to run ok on 390E

Don't ask how to hack password. (BIOS Passwords)
Sharedoc
Notebook Genius
Posts: 679
Joined: Mon Aug 18, 2003 8:46 pm
Location: Finland

Celeron 300MHz processors are famous for their overclocking potential. If it runs stabile 450MHz then it is ok.
Sharedoc
Notebook Genius
Posts: 679
Joined: Mon Aug 18, 2003 8:46 pm
Location: Finland

Floppy/USB not working correctly properly is propably caused by clock generator providing erroneous 14.318MHz, 24MHz and/or 48 MHz reference frequency. I just can't figure it out why it goes so for PIII but not for Celeron...
borc2m
New visitors - please read the rules.
Posts: 10
Joined: Tue Mar 21, 2006 7:30 pm
Location: Slovenia

Hello,
now I have bought two 128mb Hynix rams. I inserted them in my TP 390e.
The problem is that they are read by BIOS as 64mb. CPU-Z repots that it goes for 128mb ram per module. But for the total of both it says 128mb.
And also if I put only one module in it is still 64mb. Max for TP 390e is 256mb.
Is that because I have after overclocking 100mhz(99,5) fsb,but this rams are 133mhz?
I have tried also PC100 ram and it works OK(it is known as 128mb).
Sharedoc
Notebook Genius
Posts: 679
Joined: Mon Aug 18, 2003 8:46 pm
Location: Finland

Propably PC133 RAM timings do not work for such old laptop.
borc2m
New visitors - please read the rules.
Posts: 10
Joined: Tue Mar 21, 2006 7:30 pm
Location: Slovenia

My gues is because this ram has chips on only one side,that means 128mb on one side, but 390e supports only 256mb on two slots, that means 64mb per side of a slot.
So max on one side off a slot is 64mb, also if there is more than 64mb.
Am I right? I am just guessing.
zying4788
New visitors - please read the rules.
Posts: 6
Joined: Sun Feb 05, 2006 10:58 am

Who have 390 X usages slice of main plank diagram,( front side+ reverse side), I want to contrast them to have what different.
DanceMan
BIOS Rookie
Posts: 56
Joined: Mon Sep 23, 2002 12:19 am
Location: Vancouver, BC, Canada

borc2m wrote:Hello,
now I have bought two 128mb Hynix rams. I inserted them in my TP 390e.
The problem is that they are read by BIOS as 64mb. CPU-Z repots that it goes for 128mb ram per module. But for the total of both it says 128mb.
And also if I put only one module in it is still 64mb. Max for TP 390e is 256mb.
Is that because I have after overclocking 100mhz(99,5) fsb,but this rams are 133mhz?
I have tried also PC100 ram and it works OK(it is known as 128mb).
The problem results from using higher density ram than the chipset can handle. The reason PC100 is working is simply that older sticks are almost always lower density. You can usually guess the density by counting the total number of chips on the stick. Sticks with half the number of chips for the same total Mb will be higher density. You can often get some info on this from looking at the photos on eBay for the size of ram you're trying to research.

In your case, you need 128Mb sticks with 8 chips total, 4 on each side. Doesn't matter whether they're PC100 or PC133.
so many parts, so little time.....
rottenmutt
New visitors - please read the rules.
Posts: 5
Joined: Thu Jul 27, 2006 5:58 pm

http://smartdata.usbid.com/datasheets/u ... cy2285.pdf

i got a 390e with the cypress CY2285PVC-2 pll chip, see above pdf. i lifted pin 14 from the circuit board and straped it high with a 10Kohm resistor to pin 1. i boot and cpuz says it is 66mhz :cry:

help please.

i did see there are 4 dip switches on the MB, don't know what they are for.
rottenmutt
New visitors - please read the rules.
Posts: 5
Joined: Thu Jul 27, 2006 5:58 pm

shit this is what happens when u do this kind of stuff at 3am. i strapped it to pin 1 which is ref0 (3.3V 14.318-MHz reference clock output). i will retry with pin 8, or 13. :idea:
rottenmutt
New visitors - please read the rules.
Posts: 5
Joined: Thu Jul 27, 2006 5:58 pm

wOOt
100-MHz of fsb love'n:)

No usb or floppy drive. Seems strapping pin 14 high drowns out the 48-MHz usb clock signal. I will play with different resistors to see if I can get it to work.
I do not measure any conductivity from pin b34 (see ftp://download.intel.com/design/mobile/ ... 366801.pdf) to pin 14 on the cypress chip.
A work around would be to obtain another cypress chip and run it piggy back to power the usb clock signals, would need to trim off all other outputs.
rottenmutt
New visitors - please read the rules.
Posts: 5
Joined: Thu Jul 27, 2006 5:58 pm

usb and floppy are now working using a 10Kohm resistor. didn't work when put together on the table, but when assembled into the case it worked.

starting to get a little graphics corruption around the curser after it has been on for a while, maybe heat or the agp bus isn't running at the correct speed.
trillium
BIOS Newbie
Posts: 16
Joined: Wed Nov 29, 2006 12:07 am

For some time now I have been following the ThinkPad 600E CPU upgrade thread at http://www.wimsbios.com/phpBB2/viewtopi ... sc&start=0, and have been fascinated by the longevity of the thread, and the amount of great information from sharedoc and many others - thanks to you all! When I came across this tread on upgrading the CPU in a ThinkPad 390E I decided to give it a try, since I had an old unit (2626-DOU, P-II at 300MHz) lying around doing nothing. These are my findings after modifying the FSB to 100MHz - please excuse the long post (which is my first in any forum!) but I wanted to give detail in the hope that someone might be able to suggest fixes for the problems that I have encountered.

I followed the directions given in rottenmutt's post, (plus the Cypress CY2285-2 clock generator chip's documentation he referenced), to get 100MHz FSB by pulling pin 14 high with a 10K-ohm resistor. After the modification the original CPU is reported by the BIOS as a Pentium II, 450 MHz! The system booted into Windows (98SE) OK, and WCPUID reports 100MHz FSB and the CPU to be a P-II at 450MHz! Sounds great, BUT ...
1) the floppy drive is not working - Windows Explorer shows a 5 1/4" drive, not 3 1/2" ??
2) USB is not working
3) serial port is not working
4) I haven't checked the parallel port, but the PCMCIA slots do appear to be working since a USB 2.0 card was detected OK.
The BIOS is at version R01_C9, the latest available for the 390E.

I have also tried a couple of other CPU's with this 100MHz FSB mod:
1) a P-III, 500 MHz (non-speedstep) - this is detected by the BIOS as a PENTIUM PRO, 500MHz, but the boot into Win98SE hangs! I tried doing a clean install of Win98SE with this CPU installed, and did manage to get into Windows with the default 640x480 video resolution. But after installing the NeoMagic AV256 driver from IBM all I got was a black screen!! I never did get it working.
2) a Celeron 466MHz - being designed for a 66MHz FSB I had hoped this might run at 700MHz, but no such luck!! In fact, NOTHING AT ALL other than the power-on light - no BIOS, no video!!

My lastest experiment has been to update the 390E's BIOS with the latest version for the 390X (ISB055WW), which takes the P-III CPU. The BIOS updated OK - after some fudging! (Initially the update program reported "This diskette does not support this system.", but I located where the model ID was being checked and changed the test value to be the same as in the 390E update program. Then, because the floppy wasn't working, I had to make a bootable CD of the update program in order to do the update!) With the 390X BIOS installed, the system runs the same as it did before with the P-II 300MHz CPU - i.e. the CPU is detected as a P-II at 450MHz, it boots into Win98SE and everything seems to work except for the floppy drive and the USB and serial ports. On startup the new BIOS does give a message saying "Re-work needed - using old RTC" but it continues to boot, or hitting F1 causes it to enter BIOS Setup OK. With the 390X BIOS the P-III CPU is now detected correctly as a "Pentium III, 500MHz", but the system still hangs on attempting to boot to Win98SE. And the Celeron 466 still gives NOTHING AT ALL!

I've learned a lot by reading this, and other, forum posts and would like to thank the many talented individuals who have shared their knowledge in these forums. But I'm strictly a 'follower' when it comes to these sorts of experiments with computers, and I've reached my limit! I'm hoping that perhaps some of the very talented individuals reading these forums may have suggestions on how to fix the problems I have identified with my 390E 100MHz FSB mod. Or might it be that the floppy, USB and serial ports on the 390E will not work at 100MHz FSB? (rottenmutt reported problems with his mod. 'on the table', but floppy and USB working OK when fully reassembled in the case; however that has not been my experience.)

One final piece of information. The 390E has no on-board memory, just two memory slots each of which can take up to 128MB of PC-66 memory according to the IBM website. I found that the original PC-66 memory worked just fine at 100MHz, but I also tried substituting PC-100 memory to see if that would make any difference; it did not! And I also found that - quite contrary to information on IBM's site - the 390E will take two sticks of PC-100 256MB memory for a total of 512MB, the same as the 390X!! In some forum I've seen a question about the maximum memory a 390E will take, and I can now say the answer is 512, not 256MB.
Sharedoc
Notebook Genius
Posts: 679
Joined: Mon Aug 18, 2003 8:46 pm
Location: Finland

Floppy, USB do not work because they are oveclocked 50% when FSB was clocked from 66 to 100MHz.

One possibility is to have a switch to select between 66MHz and 100MHz. Then you could use 66Mhz speed with the peripherals.
trillium
BIOS Newbie
Posts: 16
Joined: Wed Nov 29, 2006 12:07 am

Hello Sharedoc,

Thanks for your reply - I did wonder if the USB and floppy simply wouldn't work with the mod to 100MHz FSB, but the fact that 'rottenmutt' had reported (3 posts up) that his floppy and USB were working with the FSB mod (once he reassembled everthing in the case) made me think there might be a fix. In the data sheet for the Cypress CY2285-2, (rottenmutt gave this link: http://smartdata.usbid.com/datasheets/u ... cy2285.pdf), there is a 'Function table' that indicates the PCICLK, USBCLK and REFCLK frequencies do not change when the CPUCLK is changed from 66 to 100 by pulling pin 14 high. Does this not suggest that the USB/floppy should work OK? But then I don't really understand this technical stuff - I just try to follow instructions!

Do you have any thoughts on the other issues I noted, e.g. why the P-II/300 worked OK at 100MHz but I got absolutely nothing with the Celeron 466 (also a 66MHz FSB CPU)? (The Celeron CPU works fine at 66MHz.) Or why the video problem with the P-III/500 and the Neomagic 256AV driver? And what about using the 390X BIOS with the 390E? I haven't noticed any problems with it, and it does recognize the P-III correctly, so do you think it's OK to leave it installed? Any suggestions would be much appreciated.
Sharedoc
Notebook Genius
Posts: 679
Joined: Mon Aug 18, 2003 8:46 pm
Location: Finland

Hi,

Looks like pin 14 is bi-directional, so during startup it is read and the FSB determined (0V = 66MHz, 3,3V = 100MHz). During operation the pin feeds the 48MHz signal to USB devices.

Try 3,3kohm resistor between pin 14 and pin 28, that should work.
Post Reply