[ Home ]
 
PA-2007 How-to's Page
 

Index

  • [ How-to-guide for PA-2007 core voltages ]

    • Learn how to get every core voltage from 2.1V to 3.3V in steps of 0.1V without modifying your motherboard.  Now your PA-2007 has core voltage support for all current Socket 7 processors, including the new 2.2V K6 266/300 & K6-2 266/300/333 processors.
  • [ The K6 266 overclocking report ]

    • This report covers the overclocking experiences of one successful overclocker, namely Jonathan Sasse, with his PA-2007 motherboard and K6 266 cpu combination. 
  
How-to-guide for PA-2007 core voltages 
[ CPU Core Voltages | CPU Voltage Jumper Layout | Voltage Circuit Schematic | Jumper Settings ]
 

 

CPU Core Voltages

The core voltage can be set between 2.1V and 3.3V in 0.1V increments without soldering on your  motherboard. 

The reason we need core voltages as low as 2.2V is to support the new AMD K6-266 & AMD K6-300 MHz processors manufactured using the 0.25 micron process.  Previous AMD K6 processors manufactured using the 0.35 micron process used higher voltages, e.g. AMD K6-166/200 MHz require 2.9V & the AMD K6-233 MHz requires 3.2V. 
 

CPU Voltage Jumper Layout 

This the FIC PA-2007 rev 1.2 motherboard component layout for the VR jumper block.  PA-2007 motherboards that are missing the 2.1V jumper position (11-12) cannot achieve 2.2V. 
 

 
                   VR 

3.5V       1 O---O 2 
3.3V       3 O---O 4 
3.2V       5 O---O 6 
2.9V       7 O---O 8 
2.8V       9 O---O 10 
2.1V     11 O---O 12 
 
 
 
 

Voltage Circuit Schematic Diagram

I did not have a schematic diagram for my motherboard so I began my investigations into core voltages by reading the manual and then visually inspecting my PA-2007 rev 1.2 motherboard.  I found 6 pairs of jumper positions labeled VR used to set these voltages and also found 7 tiny surface mount resistors beside the VR jumpers.  Close to the resistors I found a screen printed label identifying the resistors as R267, R268, R269, R272, R274, R277, & R701.  There is only 6 jumper positions but 7 resistors - one resistor extra.  Why?  See below. 

I used a digital multimeter to measure the circuit and  I chose to draw a schematic diagram rather than explain with words my findings. 
 

 
 
                            VR 
                          jumpers         o----[R267]----o         < core voltage 
                                               |                                        output 
       1-2     O   O-----[R268]----o 
                  |                            | 
       3-4     O   O-----[R269]----o 
                  |                            | 
       5-6     O   O-----[R272]----o 
                  |                            | 
       7-8     O   O-----[R274]----o 
                  |                            | 
      9-10    O   O-----[R277]----o 
                  |                            | 
     11-12   O   O-----[R701]----o 
                  | 
                  | 
              GND 
 
 

I measured the core voltage settings at the R267 resistor on the side closest to the CPU socket.  The core voltages can only be measured with a split voltage CPU in the CPU socket and in my case I used a Pentium MMX processor.  First, I made sure that when I shorted the VR jumper position 9-10 the voltage measured 2.8V and then I shorted the VR jumper position 11-12 and confirmed the 2.1V setting.  I verified all voltage settings with a digital multimeter.  Then I began by doing multiple jumper shorts and I tried various combinations while measuring the core voltage output at each point and documenting it all.  Following are the results: 
 

Jumper Settings 
To achieve some of the voltages it is necessary to solder a resistor onto a little connector ( I found that the turbo switch connector in my PC case would fit on the VR jumpers so I cut it off and left about 1.5cm of the wire exposed so I could solder the resistor to it.)  A 1/4 or 1/2 watt resistor will do in all circumstances as there is very little current flow through the circuit.  I admit there is some work involved in soldering the resistor onto the connector but it will be worth it to get the core voltages of 2.1V to 3.3V in steps of 0.1V. 
 
 
 
 

Core voltage* Actual voltage Jumper Settings
2.1V  2.123V 11-12 short
2.2V 2.196V 11-12 shorted with a 47K ohm resistor
2.2V 2.251V 11-12 shorted with a 100K ohm resistor
2.3V 2.340V 11-12 short  &  9-10 short
2.4V 2.414V 11-12 short  &  7-8 short
2.5V 2.501V 11-12 shorted with a 100K ohm resistor  &  9-10 short
2.6V 2.625V 11-12 short  &  5-6 short
2.7V 2.735V 11-12 short  &  3-4 short
2.8V 2.791V 9-10 short
2.9V  2.899V 7-8 short
3.0V 3.072V 11-12 shorted with 100K ohm resistor  &  9-10 short & 5-6 short
3.1V 3.134V 11-12 short  &  9-10 short  &  5-6 short
3.2V 3.205V 5-6 short
3.3V 3.301V 3-4 short
 
 
* Note: As you can see the nominated core voltage setting is not always the exact voltage i.e. 2.1V is not exactly 2.100V but is actually measured to be 2.123V. 
 
 

Now read WindWalker's PA-2007 Voltage Adapter page. He provides practical explanation with photos to help you. 

 

[ Top ]
 

 
How-to-guide for updating system BIOS
 
[ Upgrading System BIOS | Flash EPROM Updating Procedure ]
 
 
 
Upgrading System BIOS 

FIC manufactured the PA-2007 motherboard with a programmable BIOS chip called a Flash EPROM.  The system BIOS is updated with a BIOS file obtained from FIC using Flash Memory Writer utility program.  This allows a convenient way to update the system BIOS. 

Please note that updating the system BIOS is serious and will not always result in performance gains therefore only update when necessary.  Remember updating an incorrect file to FLASH EPROM could cause the system to fail to boot up.  Carefully follow the procedures in the manual and double check always,  or ask a qualified technician to assist. 
 

Flash EPROM Updating Procedures 

Make sure the system is running in real mode.  This utility will not operate if the system is in protected mode or virtual mode.  This means that you cannot program the motherboard BIOS when your system is running Windows or running with a memory manager such as HIMEM.SYS. 

Note:  Make sure you have a copy of the Flash Memory Writer utility (FLASH5xx.EXE) and the BIOS file together on the boot floppy disk or on the hard disk drive. 

The following describes three easy ways to run the system in real mode: 
 

  1. Boot from a floppy disk formatted with the "FORMAT A: /S" command without creating CONFIG.SYS and AUTOEXEC.BAT files.
  2. If you are using MS-DOS 6.x, you can bypass the AUTOEXEC.BAT and CONFIG.SYS when booting by pressing <F5> when  "Starting MS-DOS... "  line is displayed on the screen.
  3. For Windows 95 users, press <F8> when  "Starting WINDOWS 95... "  line is displayed on the screen. Then from the Microsoft Windows 95 Startup Menu choose "Safe mode command prompt only".

    4.  
 
Note:  For DOS and Windows 95 users,  failing to boot without a memory manger will result in the "Memory Insufficient error message" when using the Flash program.
 
 
Once you have accomplished the above tasks, you can run the Flash Memory Writer utility.  To execute the utility, change to the directory containing both the FLASH534.EXE and the BIOS file, then at the DOS prompt, type eg: FLASH534 113CD133.AWD and press Enter. 
 
 
Note: Use the standard Flash BIOS utility (FLASH534.EXE,  unless FIC specify otherwise.
 
 
From the Flash Memory Write menu, select N to Do you want to save BIOS? if you do not want to save your current BIOS.  Select Y to Do you want to save BIOS? if you want to save your current BIOS,  type the current BIOS name and the extension after File Name To Save: eg. 109CD12.AWD.  I strongly recommend saving the current BIOS so you can always go back if there are problems.  From the Flash Memory Write menu, select Y to Are you sure to program?  The BIOS file will be downloaded to the Flash EPROM.  Wait until it displays Message: Power Off or Reset the system. 

WARNING:  If you encounter problems while downloading the new BIOS, DO NOT turn off the power, as this will corrupt the BIOS preventing the system from booting up.  Just repeat the process, and if the problem still persists, download the original BIOS file you created in the procedures above. 

After successfully downloading the new BIOS file, exit the Flash Memory Writer utility and then turn off your system.  When the system starts to boot up, press <Delete> to enter CMOS SETUP UTILITY Menu.  Select LOAD BIOS DEFAULTS to load the default BIOS values.  Then select LOAD SETUP DEFAULTS and execute it.  Save and exit CMOS SETUP UTILITY. 

Congratulations you have successfully flashed your motherboard BIOS . 
 
 
 
 

Flash Memory Writer utility (FLASH534.EXE) switch options. 

FLASH534 [(FLASH)[PATH][FILENAME]] [/Py,Pn] [(SAVE)[PATH][FILENAME]] [/Sy,Sn] [(CLEAR CMOS)[/CC]] [(CLEAR PnP)[/CP]] [(CLEAR DMI)[/CD]] 

[/Py,Pn] PROGRAM BIOS ANSWER Y or N 
[/Sy,Sn] SAVE BIOS ANSWER Y or N 
 
 
 
 [ Top ]
 

 
The K6 266 overclocking report
 [ Pushing the Limits ]
 

 I will start from the top as far as preparing you 2007 for the newer K6 CPUs.  As many people have read, and some of us have had the chance to experiment with, there is a beta BIOS available for this motherboard.  This BIOS apparently comes in two flavors, 113cd15 and 113cd133.  I have yet to determine much of a difference in the two, and am currently running the ‘133’.  At any rate, there are 3 situations to watch out for if you are about to address a BIOS flash for your system. 

  1. If you have a revision 1.1 Motherboard and plan to run the 266 (or above, for further reference I will refer to all K6 266 and above as 266 to save space), make sure you have jumpers 11-12 for voltage regulation.  This is essential for being able to do this via jumpers (we are all now aware of the resistor option, however well that may or may not work without an 11-12 pin set).  Some 1.1 rev boards only go up to 9-10 and that does not offer much in the voltage range for these CPUs.
  2. You may have a 1.2 rev motherboard, but you have jumpers for flash EPROM type (EP1 and EP2 located just to the left of the BIOS chip).  These jumpers are set by default to a non standard setting to prevent the boot block.  The necessary position of these jumpers for a flash must be in the Intel 28F001 position, or EP1 1-2 EP2 2-3 .  If not, on attempted flashes you will get either ‘Program Chip Fail’ or ‘Erase Chip Fail’ messages.
  3. You have rev 1.2 and your EPROM settings are hardwired to a flashable position, lucky you flash away :)
 
*Note:  FIC sends Flash utility 534 with the ‘133’ BIOS and Flash 540 for the ‘15’ BIOS, if anyone knows why this is or basically what is the deal with these two BIOS files feel free to set me straight, I have tried both, I am using the ‘133’ now because that is what I am at 375 with now, and well, haven't had much time to tinker with BIOS options, had some correspondence with some people that tried both and they thought with the higher speeds the ‘133’ tested a little better, not sure.
 

 After you have flashed you will need to get your board jumpered correctly.  Again, with the many revisions of this board out there, make sure you have the settings necessary to get the most out of a ‘266’.  One being the 75 MHz bus, some boards apparently either do not have this option, or when tried is VERY unstable, I have never had a problem with the 75 MHz bus setting since I was running my 166->225.  If you know that your 75MHz bus works fine, great, I wouldn't expect any problems.  I wish I had all 5,6 or 7 revisions of this board to play with, but I don't so you will have to make your call based on your own experiences with your system. 

 For starters, I am using the ‘standard K6 fan’, meaning, the big black heatsink and fan that usually accompanies the K6.  I am also using a thermal compound, for overclocking, I would recommend nothing less, really a bare minimum here.  I suppose I was greedy, but hopeful when I first put my 266 (really a 266) in my system as I set it for 75x4 right off the bat using the 2.1V setting.  I wouldn't start any higher than this at first, get a feel for what you have going here. 
 

Pushing the Limits: 
 I had high expectations for the entry level 266 as I had such great luck with the entry level 166 about 9 months ago (or whatever).  I have come to believe that AMD has a VERY high yield on their early chips, which is why we see the higher rated CPUs out so early.  The 75 MHz bus speed became my standard for this CPU; I never bothered with anything lower as I wanted to see how high it would go.  At 300 MHz (75x4), I was impressed, trying this with an old BIOS BTW, is a huge waste of time don't bother.  I ran some benchmarks, did some testing, tried Quake II, etc., seemed stable enough, okay, power down and off to 337.5 (75x4.5).  Very nice.  It worked flawlessly (still at 2.1V here), did the same as before with regards to benchmarks, seemed fine, ran some Quake demos for some time to see if there were any stability problems, no sweat.  Okay fine.  I originally though as far as this board would go would be 337.5 and I had reached my limit.... or had I?  Turns out with some playing around, there is a 5x setting.  Would this chip run at 375, well of course I had to find out. 
 
 

*Note:  What I can only guess is FIC's way of officially not supporting the 75MHz bus, the BIOS ID only identifies the CPU to ‘the nearest 66’, meaning POST and it reads 337.5 as 333 and 375 as 366, windows apps and benchmarks verify the real speed.
 

 So I clocked it up to 375, and noticed some problems.  There were some instability issues here at 2.1V.  I shut down and went to the 7-8 11-12 ~2.4V setting and fired back up.  This time, it booted fine.  I went about the usual methods of testing and playing to see if the system would take this new found 375 MHz setting, and it went quite smoothly.  I have had it this way for about 2 weeks or so, running really hard for the majority of the time.  I have all sorts of benchmarks, I will post those in a separate message if the demand is there, and I have screen shots of the results from various programs.  Quake II for those that are interested did well, Yielding an average of about 49FPS in time demos and about 67FPS in timed play.  Forsaken yielded an amazing 215 FPS during gameplay, a ridiculous number and a true testament to the Monster II I was using, that value is hardly realistic and can best be viewed as a novelty as anything.  I am curious to see everyone's results who tries any of this as I am sure we could all learn a thing or two about what people are trying. 

 Some of you may be surprised at my system components; I will list them here: 

64 MB EDO 60ns RAM (TexIns) 
Diamond Stealth 3D 2000 PCI Video Card 
AMD K6 266 (.25 micron, no 3DNow! ) 
PA 2007, rev 1.2, 586A, ICW/Phaselink Clock Generator Hard Drives:  1.6 GB NEC (mode 4), 1.2 GB Quantum FB (mode 4) Some tests were done using a 12MB Monster II VooDoo2 accelerator. 
BIOS 113cd133 
IDE ZIP Insider (ATAPI) 
POS Sound card, USR 33.6 Modem. 
 
 

As for jumper settings the only undocumented (or unsupported settings) are as follows: 

5x Multiplier      -     1-2   2-3   2-3 
75 MHz            -     1-2   2-3   1-2 
2.4V                 -     7-8       11-12 
2.1V                 -         11-12
 
 
 


 
 
[ Top ]
 
 
 Copyright © 1998 Antonio Cataldo
All Names and Trademarks are the rights of their respective owners.
 Disclaimer
 
updated 27th June 1998