View Single Post
Old 23rd Mar 2021, 11:29 am   #108
ortek_service
Octode
 
ortek_service's Avatar
 
Join Date: May 2018
Location: Northampton, Northamptonshire, UK.
Posts: 1,392
Default Re: The Transam Triton Personal Computer

Quote:
Originally Posted by GeraldSommariva View Post
Done a bit more research on my Keyboard the Model number is 84SD12-2 did not find any reference to that.

Its not a matrix scan keyboard, https://telcontar.net/KBK/tech/Hall_effect its Two-of-N encoding each switch has two open open collector outputs, some pulse like the letter and number keys but the Shift and Caps Lock stay low as long as you hold the key down.

So it part or the Micro Switch SD range https://telcontar.net/KBK/Micro_Switch/SD_keyboards date code is 77/39.

I think the 40pin chip on it is a custom rom so that not getting fixed anytime soon.

That's useful further info on it - I had assumed it might be an SD-16132, that was on the PCB photo you'd posted.

I'd not looked into Hall-effect key switches and that because they used single-ended outputs, you couldn't use a standard matrix-scanning IC.
But with these, each key effectively gives it's position by ground the relevant Row & Column key 'scan' matrix lines. So it actually simplifies the controller IC slightly and mainly just needs to be a ROM code converter of what codes are being presented to it from the keyswitches (plus some debouncing? & shift key handling).

After reading your links, I did a quick Google for Hall-effect Keyboard encoders and co-incidentally it found another page on that site: https://telcontar.net/KBK/tech/encoder_ICs

Where it summarises many of the standard scanned-matrix ones and does mention substitution of custom-ROM types (giving link to Acorn Stardot forum discussion), plus lists two 2-of-N for Dual-Output Hall effect switches:

Model Manufacturer Keys Modes Bits/key Output Coding Notes

MM5745 National 78 4 10 Parallel Custom
MM5746 Semiconductor 78 4 9 Parallell Custom

2-of-13 input (non-matrix); listed as 2KRO/NKRO but no explanation is given regarding how NKRO works without matrix scanning


So their might be some possibility of using one of those, if available.

I saw from another previous post "The recommended one was a George Risk keyboard model 756 - which used a custom version of the KR2376 (-012) so the Factory ROM on the controller is unique to that model KB"

So I presume your Keyboard must have used the same output codes (/ ASCII is standard) for most keys, in order for it to work on the Triton's firmware (Unless you had a special version of the Triton firmware for your keyboard's ROM-codes).
And so can probably find out what the ROM codes were, from people who have a working system / look at Triton's O/S code (A commented disassembly may be useful, to save some work).
An AVR etc. microcontroller could then be used to replicate the original IC, and shouldn't have to do too much other than outputting the right codes (although may need to do some debouncing, as well as shift-key handling)
ortek_service is offline