Gigabyte GA-P67-DS3-B3 combined PS/2 wiring - connect both KB+mouse

Hi,
is there some electronic guru who know how the combined PS/2 connector is wired?
I found there are 2 versions. One, like on MB GA-P85-D3 use pins 2 for MDAT and 6 for MCLK so it allows connect both KB and mouse via Y-cable splitter. It's simple and working. But it's not the case of GA-P67-DS3-B3 that seems to have pins 2 and 6 unconnected. I checked this with multimeter. So there must be some switching logic between SuperIO and PS/2 connector. I do a heavy google search but didn't found any motherboard schematic diagram that use this kind of combo PS/2 (I found only older schematics of MB with 2 connectors). Does anybody have idea how it is wired? Does anybody tried a HW modification to the motherboard to make possible both KB+mouse? I need it because of using various systems that not all have native USB support. I know there's USB legacy KB support but it's often coded so badly makes some program crash or not working.
My idea of HW mod is to connect second PS/2 connector for mouse directly to SuperIO (IT8728F have 2 separate PS/2 lines) while disconnect this pin from MB. Maybe it will require SuperIO registry programming to enable mouse PS/2.

Рекомендуется к прочтению по той же теме

RayeR писал(-а):
So there must be some switching logic between SuperIO and PS/2 connector.
No logic.
You can connect only one device - keyboard OR mouse.

RayeR писал(-а):
Does anybody tried a HW modification to the motherboard to make possible both KB+mouse?
I not tried, but IMHO your idea is real.
In your case MDAT and MCLK have pull-up resistor (8.2 kOhm, for example) to +3.3VDUAL. No other connections.
You must find this resistors, desolder it, connect MDAT and MCLK via series resistors (82 Ohm, for example) to pin 2 and 6 (for Y-cable splitter) or to new PS/2 connector, add pull-up resistors (8.2 kOhm, for example) to KB/MS power line and capacitors (180 pF, for example) to GND.

Перевод(вольный, извините за возможные неточности):

"Здравствуйте Уважаемые Специалисты! Знает ли кто-нибудь разводку для цепей совмещенного разъема PS/2 (mini-DIN, у которого половина окрашена сиреневым цветом, другая половина -- зеленая /прим. novice_dP)?
Мне удалось выяснить что встречаются два варианта исполнения. В одном случае (например, на GA-P85-D3) контакт 2 mini-Din используется для сигнала MDAT (линия данных манипулятора мышь PS/2 /прим. novice_dP), a 6-й контакт используется для сигнала MCLK (линия синхронизации мыши/прим. novice_dP) и, следовательно, возможно использование Y-разветвителя для одновременного подключения клавиатуры и манипулятора мышь с интерфейсами PS/2. Здесь все просто и работает должным образом. Во втором случае (GA-P67-DS3-B3) похоже на то, что 2-й и 6-й контакты совмещенного PS/2 не используются, это показала проверка (прозвонка) мультиметром. Следовательно должна быть какая-то переключающая логика между SuperIO и совмещенным разъемом PS/2.
Я искал в google схемы материнских плат с совмещенным PS/2, но безуспешно (находил схемы только с раздельными разъемами PS/2). Имеет ли кто-нибудь представление о том как именно разведены эти разъемы? Может кто-то пытался изменить(доработать/прим. novice_dP) схему материнской платы для обеспечения одновременной работы мыши PS/2 и клавиатуры PS/2?


Мне это нужно по следующей причине -- приходится использовать различные системы(вероятно, операционные системы/прим. novice_dP), часть из которых не имеет встроенной поддержки USB. Да, я знаю о существовании "USB legacy KB support", но, зачастую, этот модуль написан так криво, что программы его использующие либо "вылетают" с ошибками, либо не работают вовсе.
Моя идея аппаратной доработки -- подключить второй PS/2-разъем для мыши напрямую к чипу SuperIO (IT8728F имеет 2 раздельные линии PS/2), при отключении этого вывода от материнской платы. Возможно, потребуется программирование регистров SuperIO, для включения (разрешения использования) мыши PS/2
"
Hi RayeR. I have the similar problem You do with asus P8H87-Pro mobo. Y splitter does not work because there are no connection on combined PS/2 socket pin2 and pin6.

Hi, thanks for replies and translation for other users.

> maco
By the logic I mentioned a multiplexer or similar circuit that connect two PS/2 lines to the same pins of PS/2 connector. Multimeter probe indicate that superIO lines are not connected together. I can measute 82R between KCLK, KDAT and PS/2 connector but much higher resistance for MCLK, MDAT.

Yes, the wiring you have described I'm going to do now.
I already got a DDR3 module to run the MB. I checked that KB and mouse separately works in this PS/2 port. Even, whats's interesting, when I booted with mouse plugged in and then swap the keyboard in it was detected and superIO switched routing to keyboard PS/2 line so I can type. When I swapped back to mouse it didn't detected anymore, probably mouse driver would need reinitialize...

Here's important note about IT8728F: there's only preliminary datasheet ver 0.42 available on Internet that describes wrong pinout! on page 7. Fortunately I was able to find a Gigabyte GA-H61M-S2-B3 schematics circuit that use the same IT8728F and there's correct pinout on page 17.


So here we go:
MCLK - 114
MDAT - 113
KCLK - 112
KDAT - 111

Sure, I should add some ESD/EMi protector, capacitor and other "useless" components but first I'm curious if it will work. I'll let you know.

RayeR писал(-а):
By the logic I mentioned a multiplexer or similar circuit that connect two PS/2 lines to the same pins of PS/2 connector.
Yes, I know:). No any circuit that connect MDAT and MCLK to pins of PS/2 connector.

RayeR писал(-а):
that describes wrong pinout! on page 7
No. Pinout in official datasheet is right:).
Gigabyte orders special version that marking on 3-rd line:
IT8728F
xxxx-xxx
xxxxxxGB

GB - special version for Gigabyte only.
Rule "+31": add 31 to pin number from official datasheet - get pin number for special version.
Example: KDAT in official datasheet have pin number 80. For Gigabyte version - 8+31=111.

Aha, what a kind of obscure obfuscation! Why they requested this difference from ITE?
It's good to know in case of replacement, improper chip would release a magic smoke then, haha:)
My chip marking is exactly:
IT8728F
1118-DXA
AC07KDGB
Does also exist any other irregular variants like for Asus or so?
Do you have some newer version of datasheet than 0.42?

Whoaa, I made it! Time to do Steve Balmer's mokey dance

My first attempt had failed because I swapped all pins on PS/2 mirros side - I overlooked that I glued the connector upside-down instead that one is on board. Thankfully to current limiter protection didn't blew my mouse:) Shame on me.
So after I fixed the wiring and booted I found it works just fine! I was a little bit shocked that when I curiously swapped KB and mouse it still worked! So there must be some switching logic (protocol switching) embedded inside superIO so it can accomodate to KB/mouse on both lines. Also when I swapped KB and mouse at runtime the KB was immediatelly recognized and working. Mouse didn't because it required to call driver again to reinitialize it. After this it worked too. So big thank to BIOS writer engineers that did their job with technical thinking (instead of marketing thinking how to save a cent for 2-port connector) and allowed me to fix it thisd way. Now I can run native PS/2 like before without messing with messy USB legacy emulation code...
So anyone with decent electronic skills, micro-solder iron and steady hands can do it yourself:)
test setup


SIO-PS/2 wiring
ctmouse loaded successfully
I'll take some better and closer shots at daylight.

Update: I can see that KB/mouse automatic switching is configurable feature of SuperIO, see reg.
APC/PME Control Register 2 (PCR 2) (Index=F4h, Default=00h)
bit 7: Auto-swap of KCLK/KDAT and MCLK/MDAT
0: Enable (Default)
1: Disable

RayeR писал(-а):
Why they requested this difference from ITE?
Don't know.

RayeR писал(-а):
Does also exist any other irregular variants like for Asus or so?
W83667HG-A - some pin have specific function for Asus.

RayeR писал(-а):
Do you have some newer version of datasheet than 0.42?
No.

RayeR писал(-а):
Whoaa, I made it!

Just one more detailed photo of the tapped SuperIO:
rayer.g6.cz/hardware/gap67ds3/ps2mod1.jpg

I used a tip of sharp carpet-knife to lift the pin carefully from soldered pad while heating it by solder iron. It's clearly visible that those pins was routed to somewhere - they could left them unconnected when only one PS/2 used. I'm just wondering, really wanted to see the schematics...:)

Отправить комментарий

Содержание этого поля является приватным и не предназначено к показу.
  • Разрешённые HTML-теги: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <img>
  • You can use BBCode tags in the text. URLs will automatically be converted to links.

Подробнее о форматировании текста

Антибот - введите цифру.
Ленты новостей