• Please review our updated Terms and Rules here

Any interest in a SOL-20 keyboard replacement?

dfnr2

Experienced Member
Joined
Dec 6, 2011
Messages
465
Location
Dallas, TX
Would there be any interest in a SOL-20 keyboard replacement? I recently picked up a non-working SOL-20 which is complete except for the keyboard, and the chances of picking up a keyboard at a reasonable price are pretty slim. I already have an ASCII encoder PCB, so making a new keyboard is just a matter of laying out the keys.

I am already placing an order for keycaps that will covermost of the keys, but the SOL-20 has some less common key widths, including several 1.25x and a couple of 1.75x keys, and the keyboard has light gray, dark gray, and off-white keycap colors. The missing keys would be difficult to source from commodity suppliers, and would have to be custom-made. The SOL-20 keycap kit would add 36 keys on top of the "classic" keycap kit. I'm guessing that if I order 20, that would be in the range of $30-$40 on top of the "classic" keyset (which would be in the $40-$50 range). The keycap colors are not truly matched to the original, since the base keycap set is darker to broaden the usefulness in other systems.

If there's not any interest, then I'll find a way to approximate the one-off keyboard with off the shelf parts.

The keyboard is not a replica of the original. It doesn't use the same capacitive circuit or keys. But it does have exactly the same layout with the same functionality, including the 3 LEDs (UPPER CASE, SHIFT LOCK, LOCAL) and 3 outputs (RESET, BREAK, LOCAL). It uses cherry MX or Futaba MD-4PCs keys. Neither has quite the same feel as the capacitive keys, but the Futaba keys approach the travel of the keytronic switches, and do have a nice feel.

Also, the original keyboard angles the keycaps at about 10-12 degrees. A 3D printable adapter could be made (as has been done for Apple II ALPS keys). But that's not part of the original plan.

Dave
 
Hi Dave,
Welcome to the SOL-20 club.
Don't forget the OSI though ;-)

Phil

Definitely will not forget OSI :) Lately I've been turning my attention to the fallout of the great shelf failure of 2019, and directing a little TLC to those systems before putting them back up on the (properly rebuilt) shelf.

Dave
 
Oh..that doesn't sound good.
Hope you can get the Sol-20 keyboard project off the ground.
 
I already have the layout partly done. I will finish in the scraps of time I have for hobby stuff in the evenings. The keyboard will use the ASCII encoder from the unified retro keyboard project (https://github.com/osiweb/unified_retro_keyboard). The main issue is all the special size/color keycaps used by the SOL-20 keyboard, which will require a special order.

Dave
 
I think it would be really wonderful if you could make a replacement keyboard for the SOL-20. It would be a great achievement. My support 100%.
 
You might try talking to these people: https://pimpmykeyboard.com/

That is the retail site for Signature Plastics, which is indeed where I'm ordering the keys from. I just got a quote back for the SOL-20 add-on set for the keycaps I'm using for the "ADM-3A" style keyboard, and it's not cheap. If I order a quantity of 100 of the ADM-style kit along with quantity 20 of the SOL-20 add-on keys, the price will work out to about $120 for a full set of SOL-20 keycaps. If I order only 10 SOL-20 add-on kits, it will be closer to $170 for a SOL-20 kit, but the up-front cash outlay will be smaller, and the potential for storing a grand or more worth of unused keycaps may be smaller. That's why I want to gauge interest before placing an order.
 
That's why I want to gauge interest before placing an order.

It would depend on the final keyboard you are planning to create. It is was a near exact hardware replica, with the same geometry, which would drop directly into a SOL-20, and work with the same key format and key labels, I don't think anyone would have an issue paying $200 to $300 for it or more.

The aim I think would be to make a near perfect replica, even if it didn't have the unique capacitive key detector. As long as they keys looked right and you could drop it in directly and the keyboard perimeter around the keys matched the original so it looked normal when the SOL-20 metalwork surrounded it. In other words, if you could turn out a keyboard that looked and worked just like the original part, that would be the attraction. If you can do that, then you could worry less about the cost, because SOL-20 enthusiasts would all want one.
 
The keyboard will be an exact mechanical fit to the case, and also compatible with all the signals, but will not be capacitive. It will leverage the already existing ASCII encoder hardware and firmware.

Another issue is that the keycaps had a slight angle to them (10-12 degrees) level out the keycaps on a sloped keyboard mounting angle. These keycaps cannot be made with an angle. I don't think the difference is a problem in practice, but it would be nice. I think I have a solution. The good news is that the solution would require a change to the PCB's, which are cheap, so the keycaps could be used either way.

Dave
 
The keyboard will be an exact mechanical fit to the case, and also compatible with all the signals, but will not be capacitive. It will leverage the already existing ASCII encoder hardware and firmware.

Another issue is that the keycaps had a slight angle to them (10-12 degrees) level out the keycaps on a sloped keyboard mounting angle. These keycaps cannot be made with an angle. I don't think the difference is a problem in practice, but it would be nice. I think I have a solution. The good news is that the solution would require a change to the PCB's, which are cheap, so the keycaps could be used either way.

Dave

Sounds great.

What about the Red LED's in the shift lock , upper case and local buttons, is it possible to replicate those too ?
 
Yes, those are in there. For those keys, the red LED window is the second shot in the double-shot molded key, so the legends have to be pad-printed. I checked the SOL-20, and they are done the same way, except instead of pad-printed, those keys seem to have the legend carved into the key, then painted black.
 
Yes, those are in there. For those keys, the red LED window is the second shot in the double-shot molded key, so the legends have to be pad-printed. I checked the SOL-20, and they are done the same way, except instead of pad-printed, those keys seem to have the legend carved into the key, then painted black.

I really want to see photos of what you have created. When will the first prototype be ready ?
 
I'm still laying out the PCB, a half hour at a time in the evenings, but should be done this week.

The chicken-and-egg problem is that without the keycaps, it is only possible to imagine what the keyboard will look like, and I'm trying to figure out how many sets to order. Based on the interest in the OSI keycaps, of which I had 20 sets and still have about 5-10 left over, there may be interest for 3-4 SOL keyboards, so it looks like an order of 10, even at the greater price, might be the most prudent. The keycaps take about 2-3 months to produce, once ordered.

If you want to see what the ADM-style keyboard looks like, I have pictures on the OSIWeb forum.

The alphanumeric color is a little different from the SOL. From the Signature Plastics color selection, The alphanumerics are color GD, the dark keys are GX, and the white ones are WFO:

gray.jpgwhite.jpg

The alphanumeric gray is a little darker than the SOL, but was selected to also work well with a wider variety of other machines as part of the ADM-style keyboard.

Dave
 
OK, Still a bit of work on layout, but here's what it will look like:

sol-kbd.jpg

The diode per key is for n-key rollover, which the capacitive keyboard gets for free, since a capacitive keyboard cannot have "ghosting".

The keys will all be the correct size. The alphanumerics will be a little darker (to leverage bulk pricing), and the dark and light keys will be very close to original.

The extended ADM-style keyboard looks like it will be in the $75-$100 range for a new keyboard. The SOL PCB will be about the same as the ADM-style, but the SOL keycap kit costs an extra $140-$150 if I order 10, which decreases to $90-100 if I order 20. The question is, are there more than 9 people besides me who might want a SOL-style keyboard.

The SOL keycap kit extends the ADM-style kit with 35 special keys:


Color GD:
1u: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9
2u: SHIFT

Color GD with red LED lens on right side and pad-printed legend:
1.5u: "SHIFT LOCK" one word per line

Color GX:
1u CTRL (qty of 2),
1u: CLEAR
1u: LINE FEED (1 word per line)
1u: - (for keypad, centered in the keycap)
1u: + (for keypad, centered in the keycap)
1u: (division symbol for keypad, not /, centered in the keycap)
1u: . (period for keypad, centered in the keycap)
1u: * (for keypad, centered in keycap)
1.25u: up arrow
1.25u: down arrow
1.25u: right arrow
1.25u: left arrow
1.25u: HOME CURSOR (1 word per line)
1.25u: TAB
1.25u: REPEAT
1.5u: RETURN
1.75u: ESCAPE

Color WFO with black legend
1u: LOAD
1u BREAK
1.75u MODE SELECT (1 word per line)

Color: WFO with red translucent LED lens on right side, and pad-printed white legend.
1.5u: "UPPER CASE" one word per line
1.5u: "LOCAL"

BTW, I encourage anyone to double check those sizes and look for any errors.

Dave
 
Last edited:
I just placed a keycap order. The turn time is estimated to be around 12 weeks. Stay tuned.

Dave
 
Very cool !

What sort of electronics will be used in the encoder, will it be based on a keyscan system with TTL logic C's and a dedicated OTP ROM, like the original, or will you do it with a PIC micro , or an existing ASCII encoder IC system for that general type of keyboard design ?

I found with the original keyboard, the capacitive detector they have there still works with a conductive rubber disc, or a conductive metal disc (like the type found in calculator keypads) it stretches the output pulse a little, but it doesn't cause any issues or bounce problems. As long as the resistance is less than about 68k. Because I struggle more with software than hardware, if I was doing a replica, I'd attempt to replicate the original electronics with the TTL's too, which would work with the new key switches even though not likely capacitance types, but it is probably a lot quicker nowadays just to program a custom micro to do it. I have programmed a few Microschip IC's for various tasks, but its a slower process for me than doing things with hardware.
 
Very cool !

What sort of electronics will be used in the encoder, will it be based on a keyscan system with TTL logic C's and a dedicated OTP ROM, like the original, or will you do it with a PIC micro , or an existing ASCII encoder IC system for that general type of keyboard design ?

I found with the original keyboard, the capacitive detector they have there still works with a conductive rubber disc, or a conductive metal disc (like the type found in calculator keypads) it stretches the output pulse a little, but it doesn't cause any issues or bounce problems. As long as the resistance is less than about 68k. Because I struggle more with software than hardware, if I was doing a replica, I'd attempt to replicate the original electronics with the TTL's too, which would work with the new key switches even though not likely capacitance types, but it is probably a lot quicker nowadays just to program a custom micro to do it. I have programmed a few Microschip IC's for various tasks, but its a slower process for me than doing things with hardware.

This is part of a project to provide a unified replacement keyboard platform for various old micros. So far, I have a keyboard that can be configured for OSI, Apple II, or a generic ASCII keyboard; and another keyboard for the SOL-20. A single ASCII encoder board can fit any of the key matrices, so I have chosen to use regular key switches in a matrix (with optional diodes for anti-ghosting) as the most general solution. The keyboards can take Cherry MX-compatible switches, or Futaba key switches (which are a bit more of a vintage feel and a bit closer to the SOL).

The encoder uses a micro. The firmware is portable C with a unit-test suite and a thin layer of code for the hardware, so it can be easily ported to various processors. The code itself is designed to be straightforward to back-port to assembly on vintage 8-bit CPUs. The current version of the encoder board uses an Atmega328P only because it is so familiar to many hobbyists who may have been exposed to the chip via arduino, and may be comfortable with the DIP form factor. Of course the limited I/O pins of the 28-pin package require adding an additional 4 chips to the design, where a simple ARM Cortex m0 with a high pin count would only take a single chip.

The hardware and firmware are on github. The boards and code are relatively beta. I'm currently adding in the capability for supporting multiple keymaps via DIP-switch settings. I plan to also add support for selecting parallel or serial output via a DIP switch.

The output routine is a single function that can be replaced, for example, to bit-bang odd protocols such as DEC VT100 or HP 98x6 keyboards. Since key press and release events can call functions, implementing a rotary encoder such as on the HP98x6 keyboards is trivial. The DIP switches are also implemented as keys in the matrix, calling functions when flipped, keeping the core code simple and agnostic to the emulations.

I am also working on an OSI interface for the OSI-compatible key matrix, and a "universal switch matrix" interface board, that will attach to any of the keyboards, and emulate an arbitrary switch matrix for machines that do their own scanning. The universal matrix will be based on the ASCII encoder, but instead of writing out a keycode, the micro will set bits in a 2-port RAM, which can then be scanned by the host. Perhaps with a fast enough microcontroller, emulation matrix can be kept in RAM on chip, if the ISR can place the "columns" on an I/O port between the host CPU row-write and column-read operations.

Everything is pretty close. Hopefully by the time the keycaps arrive, the boards and firmware will be solid.
 
Not late at all. I've placed an order for the keycaps and am putting finishing touches on the PCB design. The firmware works just fine, but I'm making some changes so the LEDs and output lines (LOCAL, /RESET, BREAK) can be specified in the keymap definition, so that multiple keymaps can be supported without having to hard-code a bunch of special cases. I should have the keycaps in a couple of months, and then you'll be able to see how it looks. I expect to have more keycap sets than I can ever get rid of, but I'm hoping some folks might just like the look of the keyboard.

Dave
 
Back
Top