CPU all Pentium and P-MMX , all AMD K5, all AMD K6, AMD K6 II, AMD K6 III and some other CPUs
Core Voltages any Voltage from 2.0 Volt up to 3.3 Volt (and below and above!)
CPU Multiplier from x1.5 up to x6.0 (x4.0 - x5.5 with modification)
BUS frequency from 50 MHz over 66 MHz and 75 MHz up to 83 MHz
L2 CACHE Boards with 256 KB should be upgraded to 512 KB
TAG RAM TAG RAM can be inserted to fully support 512 MB of RAM
RAM fully support two banks of Page Mode / EDO / Parity / ECC RAM up to 512 MB
BIOS I have not much to say about BIOS yet.
Hard Disks No UDMA support, but UDMA Drives will work fine on T2P4
ZIP My good experience with ATAPI ZIP
USB USB
Keyboard, PS/2 connectors and replacing the fuse
COM / LPT Serial and parallel port connectors and pin outs
Suspend Switch My bad experiences with the suspend switch.
Media Bus What is the Media Bus?
Since the internet is a fluid thing, pages often change their url. So
some of my url's may directly link to the very popular (according to web
hit counters :-) 'not found on this server' page. But often the given link
comes near the target, so with a bit of 'url hacking' you may
reach the target.
What do I mean with 'url hacking'? Just cut of the last part of the
url and try again.
Example :
If ftp://ftp.heise.de/pub/ct/ctsi/setk6v2.zip
is broken
try ftp://ftp.heise.de/pub/ct/ctsi/
You may get a ftp directory or a html page where you see, that there
is a new version.
You also may get the 'not found on this server' page, or an 'you are
not allowed to access this directory' page, or whatever. If you don't succeed,
just cut more and try to find links that seem to point to the right direction.
rev 3.0
split voltage (MMX), USB, Tag Ram Socket, 75 and 83 MHz BUS speed.
rev 3.1
512 KB on board cache --> COAST socket removed. Better voltage regulator
design.
rev 3.1 with 3.2V O.K. Sticker reserved
voltage jumper becomes 3.2V
general:
You need a socket 7 or a super socket 7 CPU.
(Of course old socket 5 CPU's still work, too. :-)
Check the core voltage that is required. (AMD prints
it on the top of the CPU)
Check that you can reach the desired speed with a FSB
of 66 MHz (or 75 or 83 if you like to over clock) with a multiplier
supported by the CPU!
If it's a new kind of CPU (and NOT just more MHz)
it may not be initialized optimal by the BIOS. Find out if there a new
BIOS or a setup program or how many performance you loose without proper
setup?
Example : You replace your old P133 with a K6-2-400 and
you still use BIOS 205.
best : Use BIOS 207 beta 2
second best : Use set K6V2
last but not least : You gained a performance boost of
more than 300% with the upgrade, so you don't care about the 5% you loose
without proper initialization!
AMD:
All AMD K5 CPU's are fully supported.
K6-166 , K6-200, K6-233 are fully
supported.
Most K6-233 run fine at 2.9V without a problem.
If yours doesn't look here how to get 3.2
volt
You can use the K6-266 & K6-300 &K6-2-266 & K6-2-300 without modification if you use higher Bus speeds.
All the K6-266 & K6-300 & K6-II-266 & K6-II-300 & K6-II-333 are fully supported when you add the BF2 jumper.
All K6-II faster than 350 have the CXT core. (and some 350er and slower, too!)
The CXT (Chomper eXtended Technology) has little changes in the cache architecture so it needs a different initialization to work at full speed. BIOS 207 BETA 2 will set it up correctly. If you use an older BIOS you will lose few % performance.
But instead of the proper BIOS you also
can use setK6V2.exe or setK6V3.exe (freeware) from ftp://ftp.heise.de/pub/ct/ctsi/setk6v3.zip
to correctly initialisize your K6-II with CXT core. (V3 has been reported
to cause problems :-(
Also CPU Idle (now shareware, I have still the freeware version
:-) (and another program from the same author) are mentioned to be able
to do this, too. http://cpuidle.home.pages.de
K6-II-366 need the BF2 jumper!
I expect the K6-II-380 to be the same as a 400er.
K6-II-400 also 'likes' proper
setup by BIOS or SET K6, but don't need the BF2 trick!
K6-II-450 and 500 also 'like'
proper setup by BIOS or SET K6, don't need the BF2 trick but NEED overcloking
of the FSB!
Since the CXT core was invented to make the onchip L2 cache of K6-III possible I really recommend to use BIOS #207 BETA 2 (or setK6), or a lot of it's speed benefit may fade away!
K6-III-400 'needs' proper setup by BIOS or SET
K6, but don't need the BF2 trick!
K6-III-450 and 500 'need' proper setup by BIOS
or SET K6, don't need the BF2 trick but need overcloking of the
FSB!
The K6-III now comes in two versions : 2.4V and 2.2V.
The 2.2V type consumes round about 30% less power, or to say it
in a just more dramatic way: The 2.4V type consumes 50% more power
than the 2.2V type.
So if you are able to get the cool 2.2V type, get
it!
BTW: Just because I don't mention the K6 2+ it don't mean it wouldn't run on T2P4. It's just, that they aren't out now, and I haven't searched for information about them. (So read my page and read AMD's info about K6 2+ requirements and see if T2P4 can fulfill them :-)
I seriously recommend to use extra cooling of the voltage regulators when using something faster than 400 and for 2.4V CPU's !
Remember: Do yourself and your AMD a favor and spend $10 more and buy a big quality heat sink!
BTW: The K6-2-400 and the K6-III-400 were intended by
AMD to run at 4 x 100MHz or at 6 x 66MHz!
Since 100MHz is faster they perform better at 100MHz of course, but
the difference is not thaaaat big!
Especially with the K6-III the difference is very small!
So anybody telling you K6-2-400 or K6-III-400 'need' 100MHz, or SDRAM
or super socket 7 etc doesn't know what he's talking about!
If you you use any K6 faster than 333MHz with WIN95
you will get one of the following errors on most startups:
- Device IOS failed to initialize. Windows Protection Error. You must
reboot your computer.
- Windows Protection Error. You must reboot your computer.
There is a patch available for WIN95 OSR2 also called WIN95b
(4.00.950b)
There is no hope for old WIN95 (4.00.950) even not with service pack
(4.00.950a).
One of the few good things about WIN98 is, that it don't need
a patch for fast K6! :-)
Look at the control panel which version you have. (4.00.950?)
Follow http://www.amd.com/products/cpg/k623d/win95_update_k6.html to the Mickysoft page with the update.
BTW: This is not a compatibility problem of AMD, but a problem of careless programing by Mickysoft. It is caused by a timer overflow in a timing loop and can also happen in faster P III or whatever faster CPU!
All classic Pentium and all Pentium MMX are supported.
Remember: Intel chips have no multiplier higher
than x3.5! And most of the multipliers are crippled in slower CPU's to
protect overclocking!
IBM / Cyrix:
see Roy Lanes page http://www.concentric.net/~tukester/cyrix.shtml if you want to know more about Cyrix CPU's. I guess all will work on this board, but I'm not interested in Cyrix, because of their bad marketing (confusing the customer with different CPU's with different Clock rates, but same names!) so I haven't dealt with them and can't say for sure. But I don't expect problems!
Winchip Centaur:
The Winchip C6 is a cheap solution to put some life back to a board that has no split voltage. Good CPU for rev 2 board, bad CPU for rev 3 board.
The board automatically detects if your CPU needs single voltage supply (like K5 , C6, Pentium classic etc.) or if your CPU needs dual voltage supply. (K6, Pentium MMX etc.)
If you have a single voltage CPU, you can select between standard voltage (JP17 = STD) of ~3.3 Volt or higher voltage (JP17=VRE) of ~ 3.5 Volt. The setting of the dual voltage block (JP20) doesn't make any difference then!
If you have a dual voltage CPU you can get every
Voltage between 2.0V and 3.3V without soldering on your Mother Board!
It makes NO difference
if jumper 17 is set to STD or VRE with a dual voltage
CPU!
This is the wiring schematic of my P55T2P4 Rev 3.0 Voltage
jumper block.
(Thanks to Nicola Fabiano for the 29 K Ohm part)
+----->
reference for output circuit
|
+-----29KOhm--
<-- stabilized 3.3V source
|
2,5V O O---85KOhm---Ground
|
2,7V O O--130KOhm---Ground
|
2,8V O O--180KOhm---Ground
|
2,9V O O--210KOhm---Ground
|
res. O O---85KOhm---Ground
(1 MOhm for the latest 3.1 boards)
Vout= R/(R+29KOhm)*3.3V
ASUS used a lot of different resistors for the last jumper (called res. on my board) so it depends on your Board what voltage this gives.
measure the core Voltage
You can measure the Core Voltage, without a CPU inside, but you must know which of the hundred pins you must measure and you have to short the Vcore detection pin. Without a cpu and without the shortcut you will always measure the single voltage for an old Pentium without MMX and not the correct core Voltage for MMX CPU's!
But there is an easy way!
Measure the Core Voltage at the middle pin of the Single
Power Plane jumper (JP 17 VRE/STD) against Ground with Pentium
MMX or K6 in socket. Please test this first with the standard
jumper settings (2,5 Volt, 2,7 Volt etc..) so you can be sure your measurements
are correct!
higher Voltages
To get higher voltages you simply solder a resistor (¼
or ½ Watt) on a little connector and put it on the 2.9V jumper position.
(Don't change resistors on the MB like ASUS says!
It's easy to destroy a MB!)
To get 3.0 Volt you need a 100K Ohm resistor.
To get 3.1 Volt you need a 270K Ohm resistor.
To get 3.2 Volt you need a 820K Ohm resistor.
To get 3.3 Volt you leave all jumper positions open.
lower Voltages
To get lower Voltages than 2.5V you had to set more than
one jumper.
(Thanks to Fabian Robok the low voltage measure massacre
master.)
for 2.0V you had to set jumpers 2.5 & 2.7 &
2.8 & 2.9
for 2.1V you had to set jumpers 2.5 & 2.7 &
2.8
for 2.2V you had to set jumpers 2.5 & 2.7
for 2.3V you had to set jumpers 2.5
& 2.8
for 2.4V you had to set jumpers
2.7 & 2.8
for 2.5V you had to set jumpers 2.5
I found a table for T2P4-S (onboard SCSI) on
http://matrix.crosswinds.net/~flozo/t2p4sup/index.html
(german only!)
for 2.0V you had to set jumpers VID1 & VID3
for 2.1V you had to set jumpers VID1 & VID2
for 2.2V you had to set jumpers VID2 & VID4
for 2.3V you had to set jumpers VID0 & VID4
for 2.4V you had to set jumpers VID0 & 2000
Ohm resistor over VID3
for 2.5V you had to set jumpers VID3
Of course all Values aren't exact the named voltages,
but they are close to them.
( And don't expect with 2.5 jumper an exact voltage of
2.500000 Volt at the CPU ! )
There are some more combinations and you can use the
5th jumper too and you can adjust the voltage exact if you use a resistor
instead of one jumper, but I see no reason for that.
There is absolutely no need for an exact voltage!
The CPU don't care if it's 2.2000V or 2.2343V or 2.2132V or 2.1926V.
Higher voltage causes more heat, but higher crashing temperature and
less voltage causes less heat and lower crashing temperature.
Do you care, if your snickers weights exactly 60g or 59.27g or 60.32g?
NO! It was enough, or you are still hungry!
The same with the CPU. The Zero's after the 1/10v are just to 'pleasure'
your eyes, the CPU don't care!
If it's not stable, just raise the voltage a 0.1V step.
extreme Voltages EXPERTS ONLY!
Sure you can add a resistor from voltage reference direct
to ground (power LED Connector) to get lower than 2 Volt or the resistor
goes to 5V (power LED Connector) , that gives higher voltages than 3.3V.
But this is strictly reserved for experienced overclockers
with electronic knowledge (and newbies who want to destroy their Board
and CPU)!
I don't like this Idea, so I give no further advice!
max. current
The rev 3 boards use a switching voltage regulator. This regulator type don't convert the 'drop voltage' into heat, like linear regulators. There is only a small (compared to linear) heat production, because of the resistance of the used parts. When you use CPU's like K6-III that require high currents, these losses in the parts become a problem. To reduce this problem you can (I really recommend that!) put extra cooling on the regulators. There are two ways to do so:
1.) You can mount an extra fan (CPU fan for example) blowing over the heatsinks of the regulators.
2.) You can enlarge the heatsinks. But it's some work and you need some experience for this, so I don't recommend this method for the 'normal user'. If you want to do so, be informed, that every heatsink carries it's own voltage! You have to mount your extra heatsinks isolated, or make sure they can't hit each other!
One of the main differences between rev 3.0 and 3.1 boards is the enhanced
regulation circuit of the rev 3.1 board.
The rev 3.0 board will cause some electrical noise on the 5V and 12
power lines. This may influence soundcards and TV cards.
I had to add a coil and capacitor network to the 5V input of the voltage
regulator of my rev 3.0 board, to make my TV card produce a good picture.
When I upgraded my CPU from 333 to 500 I had to make the filter stronger
to get my picture nice again.
To give a number for the max. current is very difficult, because it
depends on many things.
The regulator design, the max. current of the transistor, the max.
current of the diode, the max. current of the coil, how big and good the
capacitors are, how 'smooth' the CPU requests the current, how much noise
can be tolerated on the core voltage, how much noise can be tolerated on
the 5V power line, how much heat is produced and taken away . . .
Because everything is twice as good with a fancy name the Bus is now called Front Side Bus. :-)
With the 4 official BUS frequencies (50, 55, 60 and 66
MHz) all parts of the MB including PCI cards and ISA cards are running
at their specifications. (or slower)
50 MHz JP8 2-3
JP9 2-3 JP10 2-3
55 MHz JP8 1-2
JP9 2-3 JP10 2-3
60 MHz JP8 2-3
JP9 2-3 JP10 1-2
66 MHz JP8 2-3
JP9 1-2 JP10 2-3
The remaining 4 frequencies exceed specifications for
MB, PCI, ISA, Cache, CPU and Disk. So it depends on your whole system if
they work or not.
68 MHz JP8 1-2
JP9 1-2 JP10 1-2 Turbo Mode nearly
no problems and nearly no speed increase.
75 MHz JP8 1-2
JP9 2-3 JP10 1-2
Works on most machines. Probably you have to slow down RAM and Disk in
BIOS.
83 MHz JP8 1-2
JP9 1-2 JP10 2-3
Works on some machines with good components.
108 MHz JP8 2-3
JP9 1-2 JP10 1-2 nobody ever claimed
to have a running system with 108 MHz (just a running PLL with frequency
counter attached) (For the dumb : 108MHz won't work!)
I warn everybody who don't know enough about overclocking to use faster bus than 68 MHz!
The clock multiplier sits in the CPU, not on the Motherboard!
The MB just hands over two signals to the CPU. This pins can be grounded with a jumper or not grounded. With this 2 inputs the CPU can select one of four multipliers. Usually the resulting multipliers are: (But there are chips with other multipliers!)
(BF1) (BF0)
X 1.5 JP11 1-2 JP12 1-2
Pentium classic / K5 only!
X 2.0 JP11 2-3 JP12 1-2
X 2.5 JP11 2-3 JP12 2-3
X 3.0 JP11 1-2 JP12 2-3
X 3.5 JP11 1-2 JP12 1-2
Pentium MMX / K6 only!
X 6.0 JP11 2-3 JP12 1-2
AMD K6-II with CTX core and AMD K6-III with CTX core only!
YES you see right 1.5 = 3.5 & 2.0 = 6.0
. It depends on the CPU what you get!
And Intel likes to disable higher multipliers in P166MMX
and P200MMX. (officially to prevent renaming, but unofficially to prevent
o'c)
From the K6-266 and up AMD has used a third input that is called BF2.
This input is only on AMD chips and will have no effect on any Pentium
Cyrix etc.! If this BF2 pin is grounded K6 266 MHz and faster will
add 2 to the multiplier set by the other two jumpers. This allows multipliers
up to X 5.5.
To get X 6.0 (only available with CTX core!!!)
BF2 must be left as it is! (No wire trick! No soldering!)
read here how to get them :higher Multipliers
The K6-III has an onchip L2 that caches
the whole 4GB of Ram! :-)
If you use a K6-III, the motherboard cache then becomes
a L3 cache which is less critical for performance. If you
have only 256K on board or lack the Tag ram upgrade there is no real need
to get them. The performance gain might be in the 1-2% range so it's not
worth to spend much money or time to get them!
Motherboards with only 256 KB pipeline burst cache should
be upgraded with a COAST Module, if you are able to get one and if it isn't
to expensive. ($20) (They seem to have vanished! :-(
I heard some complaints about generic COAST Modules,
but also heard, that some work!
I have zero problems with my ASUS CM1 REV 3.0 COAST
Module. (even @83MHz zero problems) (very hard to find now! :-(
It has 2 big and 2 smaller chips. The second smaller
chip is the extended TAG RAM.
If your Module has only 2 big and 1
small chip you have no extended TAG RAM
on it.
If you have more than 64 MB of RAM in your Board
you need an extended TAG RAM (or your machine will be s l o w) Claims
for performance loss vary from 5% to 40%.
My Opinion is: Everybody who spends much money on RAM
and saves 2$ on the TAG RAM is dumb! or may have problems to get one :-(
or may have a K6-III :-)
If you have 256
KB L2 Cache and no COAST Module you need an extra TAG
RAM for caching more than 64 MB. And you should better try to get a COAST
module! (Good luck! :-)
If you have a COAST
Module check with a hardware detection Program the size of your TAG RAM.
- If
it says 8bit TAG RAM or 64 MB cacheable area you have to add an
extra TAG RAM for caching more than 64 MB.
- If
it says 11 bit TAG RAM or 512 MB cacheable area you don't need an extra
TAG RAM (its forbidden then)
If you already
have more than 64 MB or you can get more than 64 MB for a test, than ftp://ftp.heise.de/pub/ct/pcconfig/ctcm16r.zip
(The letter r in the filename gives the version and can differ!) will say
if all your currently installed ram is cached or not. (c't is the best
and outstanding computer magazine here in Germany! Look out, they have
some english articles, too!)
On some systems you must start version 'r' with
parameter '/long' to work correct!
If you have 512 KB L2 Cache on Board you need an extra TAG RAM for caching more than 64 MB.
Go to your local electronics shop and buy a standard 32K X 8 15ns Static RAM (expect 2-6 US$)
If you can't get it in a shop, you may have luck and find the desired chip on a 486er board.
The chip should be 15ns or better 12ns.
20ns won't work stable!
Be careful if it says -10 in the end of the printing,
if the chip is old, that might be 100ns, not 10ns!
Here is a list of part numbers that Jim took out of an ISSI cross reference. Of course this list isn't proved by me, but it should be correct.
Don't worry if you get Data sheets of different chips and the numbers at the address or data pins are different. That doesn't matter. As long as all address (data) pins on one chip are address (data) pins on the other everything is fine!
ASUS TAGRAM CHIP 5 volt, 28 pin DIP, 256K 32Kx8 15ns or faster.
ALLIANCE
A57C256-10,12,15
AT&T
ATT7C199P-12,15
CYPRESS
CY7C199-10,12,15PC
FUJITSU
MB8287-12,15PSK
MB8298-12,15PSK MB82B88-12,15PSK
HITACHI
HM62832UHP-15
HYUNDAI
HY63256LP-15
HY63256P-15
IDT
IDT71256SA-12,15TP
ISSI
IS61C256AH-10,12,15N
LOGIC DEVICES L7C199PC15
MICRON
MT5C2568-10,12,15
MITSUBISHI
M5M5278CP-12,15 M5M5278DP-12,15
NEC
UPD43258ACR-15
UPD43258CR-15
PARADIGM
PDM41256LA-10,12,15P PDM41256SA-10,12,15P
SAMSUNG
KM68257AP-12,15
SGS THOMSON M628032-12,15PSI
TOSHIBA
TC55B328P-10,12,15
UMC
UM61256AK-12,15 UM61256CK-12,15
UM61256DK-12,15
WINBOND
W24257AK-12,15 W24129AK-15
RAM
Bios Auto settings for 70ns and 60ns RAM are lying! It
always times memory correctly and not as it says, even at 83 MHz.
If you use 75MHz FSB and AUTO 60ns timing,
your RAM will perform worse than at 66MHz
and AUTO 60ns !
If you overclock you should always use
manual RAM setting and tweak them! Why take the risk of harddisk
corruption and PCI card failure and then make the RAM slower than normal
with AUTO settings???
Maximal possible RAM size : 4 x 128 MB = 512 MB (proved by Rick Lindsay)
Some people claim success with new 64MB EDO's with only 8 chips, others
claim no go with them.
So make sure you can return them, or use them in another system in
case they don't work!
BIOS
Because BIOS #207 BETA 2 often seem to change it's address,
I like to advise you how you may get it if my link is broken. Just go to
the location line in your browser and begin to delete the last part of
the link and try again. Repeat until you finally get something like a directory.
Then try to dig deeper following links like T2P4, HX chipset, Beta BIOS,
etc
ftp://ftp.asuscom.de/pub/ASUSCOM/BIOS/Socket_7/INTEL_Chipset/i430HX/P_I-xP55T2P4x/
The file to look for is named 207__2.zip or similar.
It has correct recognition of all K6-II & K6-III
It has support for IDE HD's up to 32GB
It has support for USB Keyboard (Dos, setup etc.).
It claims to be Y2K compatible.
BTW: The german ASUS side seems always to have the newest BIOS. http://www.asus.de/
And don't worry about BIOS reporting wrong CPU speeds as long as the correct CPU Family is detected! That's just a 'problem' of the three seconds the wrong number appears. (In the box at the next screen all is fine!) BIOS #207 beta 2 never shows more than 400 !
THE HARDWARE set's the speed! If you set FSB and
multiplier correct (6 x 66 for example) the CPU will RUN at that speed!
(or won't run at all!)
No way for the BIOS to alter
the speed on T2P4! (Except dooze mode in power management but that's something
totally different!)
If you jumper it 5.5x 83MHz it runs at 459 MHz,
no matter if the BIOS tells you it runs at 400MHz or if BIOS tells
it runs as fast as dozen pink elephants! It will run at 459 MHz!
If your K6-2-400 is displayed as K6-2-266, then better
read the page!
I have an old noname VX board with very old BIOS, that
recognized my K6-200 as a -S CPU at 66 MHz
and benchmarks showed me that it was only 3% slower than on my T2P4 with
BIOS recognition. And I doubt this is BIOS related! So I see the 'BIOS
won't recognize my CPU fear' with much less stress. (But that may differ
from CPU Type to CPU Type!)
(Update: this 'old' board now runs a K6-2-400 CPU 'named'
-S CPU at 66 MHz)
If you still worry about the speed of your CPU, you can download Ctwfm.exe (somewhere near the setK6 file on heise.de) and put it in your autostart folder and every time your win starts you can see the measured clock speed of the CPU and feel happy!
Flashing is risky. If it fails because of an error, black
out etc. your Computer won't boot again. You have to get a new BIOS Chip
or search for somebody with a flash programmer, or do hot swapping.
If you don't get the flashing success
message try again and again with new or old BIOS, don't try to reboot!!!!!
If you flashed new BIOS load setup
defaults and BIOS defaults! And enable Quick Power
on Test.
Enable VIRUS WARNING after you installed the Operating
System. It protects you from boot sector viruses.
If you encounter HD problems after you flashed a new bios, first try to use autodetect to fix it. Go to BIOS screen and choose AUTO and LBA for all HD's. (only for small ones don't choose LBA) Then boot, see if HD's work proper and then go back to BIOS and choose USER for the detected HD's and none where no HD is.
There has been rumor that ASUS changed the LBA address translation scheme somewhere between BIOS #107 and BIOS #205. I don't have proof that they did, nor that they don't! So if you encounter problems with the proper detection of a HD (up to 32GB) that contains partitions made with an older BIOS or made with another motherboard and the above mentioned autodetect don't fix it, new partitioning may fix the problem. Sadly to do that you have to backup everything, delete ALL partitions, reboot, make new partitions with the new BIOS, reboot and restore the backups. So better check twice if there isn't something else wrong.
You may think some characters in this chapter are to big, but if you daily read the asus NG, you know, why I make this bigger every update. :-)
The T2P4 does not support UDMA. But all UDMA drives are backwards compatible and will work excellent on your T2P4.
I have tried a IBM DPTA 371350 7200rpm UDMA/66 drive on
my T2P4 and used H2bench (www.heise.de/ct ...) to see how fast it is.
At 66MHz Bus the onboard controller is limited to 13.5MB/s.
(16 MB/s at 83MHz)
So the drive can't run at it's full speed (~20MB/s
max.). But the real life performance loss is not as dramatic as it seems.
Of course you can use an UDMA 33 or 66 PCI card ( if
you like to spend the money and stick to 66MHz FSB)
You are using an old motherboard which can't play in the
high performance league.
So you have to make compromises everywhere in your system.
(no 100MHz, no SDRAM, no UDMA, no AGP no fancy other things)
So a fast Hd limited to 13.5MB/s (or 16MB/s) is a small
price to pay in my opinion.
But T2P4 is still a compeatable system in the computer 'middle-class' for an unbeatable price. :-)
I use BIOS version #205 (now #207 beta 2) with my ATAPI ZIP on my rev 3 and my rev 2 boards on channel #1 and channel 2# as master and as slave and never had any problems with it. (Even at 83 MHz it runs great!)
On my rev 3 board it was master on second IDE channel with CD ROM as slave. WIN 95 detects it automatically as drive F (my Disk is CDE). No problem with Drive letter B, because I have an old 5¼" Drive as B. If I say no second Floppy in BIOS, my ZIP Drive becomes Drive B in DOS without Drivers and stays Drive F in windows and no Drive B present. If I enable ZIP booting everything is the same, except I have a bootable Disk in my ZIP (normal format under WIN95 and sys b: under Dos!) then ZIP becomes A and Floppy B in Dos.
Later I used it as master on second channel with HD (with power of switch) as slave. - no problems!
Then I used it as slave of my Deskstar 5 on primary channel on my rev 2.1 board. - no problems!
Because I rarely used it since I got my CD burner and I got a good price for it and the 15 disks, I sold it!
my opinion on ATAPI ZIP Drive: quick, easy, small, widely used, cheap, but very expensive Media!
USB
BIOS #207 has added support for USB Keyboard (Dos, setup
etc.).
This is the copy of a message posted by Timo
______mainboard__________
| RAM
|
|
| _____/_____
| 12345 <-usb-connector| /usb-cable
/|
| 67890 <- 10-pins |
/__________/ | front-view
|
| | |
|
|
| | |
|
|
| | |
|
|
| |_1_2_3_4__| /
| PCI
| |__________|/
| PCI CPU
|
| PCI
| 1 black ground
| ISA
| 2 green + data
| ISA
| 3 white - data
_________________________ 4 red +
VCC
red 1 + VCC
white 2 - Data
connector 1
green 3 + Data
black 4 Ground
5 not connected
red 6 + VCC
white 7 - Data
connector 2
green 8 + Data
black 9 Ground VCC
10 not connected
Adding a 3A fuse into the 5 Volt lines of both connectors, to prevent damage of the board in case of a shortcut may be a good idea! And if you must make your own cable, getting ground and +5V from a Drive connector may be another good Idea!
The USB Adapter for T2P4 sometimes is hard to find (can't confirm that, saw them more than once!), but the Adapter for TX97 board is easy to get. So Joachim Klein found a way to use it instead of the 'original' ( generic :-) adapter on the T2P4. It has some advantage! It includes USB, PS/2 mouse and IRDA connector in one slot adapter.
Just connect the obvious!
This is the 2 x 9 pin TX-USB/MIR connector
_________
USB +5V |_10_|__1_|
USB +5V
USB-Port 1 - |_11_|__2_| USB-Port 0 -
USB-Port 1 + |_12_|__3_| USB-Port 0 +
Ground |_13_|__4_|
Ground
not connected |_14_|__5_| not connected
PS/2-Mouse data |_15_|__6_| PS/2 Mouse Clock
Ground |_16_|__7_|
Ground
IR-Receive |_17_|__8_| Fast IR Receive
IR-Transmit |_18_|__9_| +5V
This is the 2 x 5 pin USB connector of the T2P4
_________
USB +5V |__6_|__1_|
USB +5V
USB-Port 1 - |__7_|__2_| USB-Port 0 -
USB-Port 1 + |__8_|__3_| USB-Port 0 +
Ground |__9_|__4_|
Ground
not connected |_10_|__5_| not connected
This is the 1 x 5 pin IR-Connector of the T2P4
___
+5V |_1_|
Fast IR Receive|_2_|
IR Receive |_3_|
Ground |_4_|
IR-Transmit |_5_|
If your Keyboard don't work anymore, you probably blowed the fuse. (Maybe you didn't believed that there are some real reasons why you must switch of PC before connecting or disconnecting ANY cable!) The Fuse is the tiny little resistor like thing next to the keyboard connector marked FUSE on the MB. If you can't detect connection right through it, or you see its blown replace it with a 1A fuse. You can go to an electronics shop and try to get a replacement fuse of same style and desolder the old one, or you just can cut of the old one and solder any 1A fuse onto the rest of the pins. Make sure the new fuse can't make any short to anything!
I often heard the PS/2 mouse port called special, but
I can deny that! I just saw, that the PS/2 port I resoldered from an old
noname VX-board fitted pretty well in my T2P4 rev 2.1 and PS/2 mouse works
well with it. I also soldered PS/2 mouse ports directly to my rev 2.3 &
3.0 boards and a some other boards from other vendors and all work fine.
So it surly IS generic.
You can buy an adapter (have a look at USB part!), or
make one by your self like the following scheme.
This is the 2 x 4 pin PS/2-Connector of the T2P4
_______
+5V |_4_|_8_| not connected
not connected |_3_|___| missing pin
PS/2-Mouse Data |_2_|___| missing pin
Ground |_1_|_5_|
PS/2-Mouse Clock
This is a 6 pin PS/2 mouse connector from the mouse side!
PS/2-Mouse Data 2 3 not connected
Ground 1 4 +5V
PS/2-Mouse Clock 5|8 not connected
The COM connectors also have been called generic, but I can't confirm that (and can't deny that, too) But every connector I used on my T2P4 worked with it and is wired as follows:
MB connector SUB D 9 pol
SUB D 25 pol
(front view) (front view) (front
view)
2 4 6 8 10
1 3 5 7 9
1 DCD
1
8
2 RxD
2
3
3 TxD
3
2
4 DTR
4
-
5 GND
5
7
6 DSR
6
6
7 RTS
7
4
8 CTS
8
5
9 Test
9
22 (RI)
10 must
be cut of for some cables with protection against wrong direction!
remember : There are numbers on every SUB D connector, just look close at them!
I'm to lazy to open my PC to get the printer cable pin out, but I remember it is a bit different from COM ports. The first pin of MB is connected to the first pin of the connector. Then the next cable is connected to the pin in the other row, the third in the first row etc.
1 3 5 7 9 11 13 15 17
19 21 23 25
2 4 6 8 10 12 14 16
18 20 22 24
I don't have a serious use for the suspend mode, but as
you can imagine I just added a button to the SMI Lead because I like to
add such little things. A year later I bought a K6-200, new RAM, my ATAPI
ZIP and removed the SMI Button (because it was where the ZIP sits now).
And I got a totally stable System. Before that my system was somewhat unstable.
Sometimes it runs for hours, then it closes some programs every few minutes
because of protection errors. I thought It was the new RAM and CPU, but
I was wrong! A few weeks later I added a new Suspend Button. And guess
what --- an error every few minutes. Disconnected the Button from the MB
and totally stable system. I think the cable works as an antenna for some
electric noise waves that causes the CPU to random crashes.
So if you have a Suspend Button and unexplainable
crashes, try removing the cable from the MB!
To make it short: The media BUS is a normal PCI slot that can be used like every PCI slot and has an extra connector for ISA BUS. ASUS sells a board with a SCSI controller (that uses the PCI part) and an Sound card (that uses the ISA/Media BUS). That may be nice if you are running out of slots, otherwise just ignore it!
to be continued...