Mysterious XT keyboard issue with extended keyset \ VOGONS

Mysterious XT keyboard issue with extended keyset

Topic actions

Post a reply

First post, by Der_Richter

Posted on 2020-10-09, 10:53

Der_Richter Offline

Rank Newbie

Rank: Newbie
Posts: 45
Joined: 2017-10-17, 20:21

Hi.

After recently acquiring a “10MHz Turbo Board”, which is a Taiwanese XT clone board with an 8088 and a jumper to statically overclock it to 10MHz, basically, I have fixed and fiddled to the point it now boots fine and states it has a Phoenix BIOS dated 1985 and 640K of on board RAM so a fully populated board. That’s well and good. However it refuses any AT keyboard and requires XT protocol. That is ok. What is weird is that the keyboard controller seems to mess up as soon as an extended keyset is loaded. Then it does only one key press before hanging the input completely. I discovered this when loading PC DOS 3 from floppy. Being a Swedish user I have “keyb sv” in the startup and that works on an IBM machine as well as other clones. In this case however it allows me to press a single key before locking the input. When I removed that line, it works as intended and keyboard stays operational. The exact keyboard operation is not detailed in the motherboard manual, it just says “DIN connector for keyboard” and nothing of any config or the controller used. I’ve never seen this behavior before. So I tried a different extended keyset, and sure enough... Same issue. Remove that and use BIOS default, everything works.

What gives? How can the loading of an extended keyset driver in PC/MS DOS mess with the keyboard controller/motherboard so that it stops responding to keys? Regardless of the key pressed? AFAIK this is not a proprietary board but a general purpose clone board. Is it really not compatible with DOS extended keyset? Why? Never seen that before.

I would like my key mappings back. Any ideas?

Preserver, refurbisher, collector. In that order.

Reply 1 of 4, by kdr

Posted on 2020-10-09, 12:11

kdr Offline

Rank Member

Rank: Member
Posts: 215
Joined: 2020-09-20, 12:40
Location: New Zealand

First off, an XT system doesn't have a keyboard controller as such. It's basically just a shift register hooked up to Port A of the 8255 PPI chip.

The BIOS has a handler for IRQ1 (keyboard interrupt) that reads the scan code from the shift register, clears it (to let the keyboard know it can send the next scan code), and then interprets the scan code... stuff like keeping track of whether the ctrl/alt/shift keys are pressed, whether caps lock is on, etc. This is where the scancode-to-ASCII translation happens. Then the key is stuffed into the keyboard buffer, from whence it is made available to DOS.

I have no idea what happens once DOS reads the keys from the buffer, alas. And I don't know at what point KEYB hooks into this process.

Reply 2 of 4, by Cyberdyne

Posted on 2020-10-09, 12:18

Cyberdyne Offline

Rank Oldbie

Rank: Oldbie
Posts: 949
Joined: 2011-09-18, 18:45
Location: Estonia, Tallinn

It smells like a buggy or incompatible keyboard bios routines.

I am aroused about any X86 motherboard that has full functional ISA slot. I think i have problem. Not really into that original (Turbo) XT,286,386 and CGA/EGA stuff. So just a DOS nut.
PS. If I upload RAR, it is a 16-bit DOS RAR Version 2.50.

Reply 3 of 4, by Der_Richter

Posted on 2020-10-18, 17:34

Der_Richter Offline

Rank Newbie

Rank: Newbie
Posts: 45
Joined: 2017-10-17, 20:21

Cyberdyne wrote on 2020-10-09, 12:18:

It smells like a buggy or incompatible keyboard bios routines.

Late reply, other projects getting in the way, but his might be it. If I use a BIOS ROM that I flashed with a later date Phoenix BIOS, it works just fine to load the charset via KEYB. Weird thing is the machine was sold in Sweden back in the day. I wonder what keyboard it shipped with then... Perhaps a US charset which would have been confusing to all the business typists 😀

Preserver, refurbisher, collector. In that order.

Reply 4 of 4, by Horun

Posted on 2020-10-18, 20:38

Horun Offline

Rank l33t++

Rank: l33t++
Posts: 6024
Joined: 2019-09-19, 02:19
Location: USA

Der_Richter wrote on 2020-10-18, 17:34:

Cyberdyne wrote on 2020-10-09, 12:18:

It smells like a buggy or incompatible keyboard bios routines.

Late reply, other projects getting in the way, but his might be it. If I use a BIOS ROM that I flashed with a later date Phoenix BIOS, it works just fine to load the charset via KEYB. Weird thing is the machine was sold in Sweden back in the day. I wonder what keyboard it shipped with then... Perhaps a US charset which would have been confusing to all the business typists 😀

Odd ! Years ago got a old 386SX-16 with a German 101 AT Keyboard which worked well but if using a English 101 AT Kb it did odd things. I was able to swap the original board KB controller with one off another board and then the English KB worked fine. It was like the original KBC was mated to that particular keyboard somehow.....never heard of it before or since until you posted this. Very interesting

Hate posting a reply and then have to edit it because it made no sense 😁 First computer was an IBM 3270 workstation with CGA monitor. Stuff: https://archive.org/details/@horun

Go to top of page Go to top of page

Back to General Old Hardware

Main menu

Common searches