Converting old Xerox 820 keyboard to USB
- Redmaus
- Gotta start somewhere
- Location: Near Dallas, Texas
- Main keyboard: Unsaver | 3276 | Kingsaver
- Main mouse: Kensington Slimblade
- Favorite switch: Capacitative Buckling Spring
- DT Pro Member: -
- Contact:
I recently acquired a xerox 820 keyboard from a vintage electronics store. I have looked online and cannot find this type of connector. All I found was that it is from a xerox 820 system from the 70's. I have no idea if it is even possible to convert this board at the moment.
Pics: Connector Detail: Really looking to figure out which switch type, since I can't find the kb details on the wiki. I also would love to convert it to USB, if possible. I have a leonardo pro micro and a Teensy LC on hand.
Pics: Connector Detail: Really looking to figure out which switch type, since I can't find the kb details on the wiki. I also would love to convert it to USB, if possible. I have a leonardo pro micro and a Teensy LC on hand.
- XMIT
- [ XMIT ]
- Location: Austin, TX area
- Main keyboard: XMIT Hall Effect
- Main mouse: CST L-Trac Trackball
- Favorite switch: XMIT 60g Tactile Hall Effect
- DT Pro Member: 0093
Ooh, ITW/Cortron magnetic valve. Very nice. Can you post a photo of the compete keyboard for us to see? It looks similar to a couple of the ones that I will be selling.
This has a November 1983 date code.
The IC in the second photo is an Intel 8048 microcontroller. It is an extremely common keyboard controller for keyboards of this era.
Are you familiar with magnetic valve switches? Look up the four patents on the label for more information. The idea is that two loops of wire are inductively coupled with a ferrite loop. When you send a pulse down one wire, you can see it on the other wire. That is, unless there is a magnet in the way (which happens when a key is pressed), in which case you don't see the pulse and you know the key is being pressed. These should support a really high scan rate but exactly how high depends on the strength of the magnet and some other factors. It's a really neat, clever contactless design that is like Hall Effect but cheaper and simpler, and like reed switches but more durable.
The takeaway is that these switches are active low.
If you *really* want to know exactly what it is doing you could attempt to dump the 8048's internal ROM and disassemble it. That is some work but understanding that would be authoritative. It would tell you the polling sequence, the timings, and the protocol. It may or may not be easier than using something like Soarer's Analyzer.
As for a USB conversion I think your best bet is to pull out the 8048, wire in a Teensy++ and work on replacing the controller.
The first step is to figure out how polling works. Presumably the switches are on a matrix. The question is whether the 8048 sees raw matrix pins, or sees pin muxes first. A lab notebook and a multimeter with continuity tester would help to figure that out. Each switch has four pins. Pins 1-2 are the first loop (pulse) and pins 3-4 are the second loop (detect).
Good luck...
This has a November 1983 date code.
The IC in the second photo is an Intel 8048 microcontroller. It is an extremely common keyboard controller for keyboards of this era.
Are you familiar with magnetic valve switches? Look up the four patents on the label for more information. The idea is that two loops of wire are inductively coupled with a ferrite loop. When you send a pulse down one wire, you can see it on the other wire. That is, unless there is a magnet in the way (which happens when a key is pressed), in which case you don't see the pulse and you know the key is being pressed. These should support a really high scan rate but exactly how high depends on the strength of the magnet and some other factors. It's a really neat, clever contactless design that is like Hall Effect but cheaper and simpler, and like reed switches but more durable.
The takeaway is that these switches are active low.
If you *really* want to know exactly what it is doing you could attempt to dump the 8048's internal ROM and disassemble it. That is some work but understanding that would be authoritative. It would tell you the polling sequence, the timings, and the protocol. It may or may not be easier than using something like Soarer's Analyzer.
As for a USB conversion I think your best bet is to pull out the 8048, wire in a Teensy++ and work on replacing the controller.
The first step is to figure out how polling works. Presumably the switches are on a matrix. The question is whether the 8048 sees raw matrix pins, or sees pin muxes first. A lab notebook and a multimeter with continuity tester would help to figure that out. Each switch has four pins. Pins 1-2 are the first loop (pulse) and pins 3-4 are the second loop (detect).
Good luck...
- Redmaus
- Gotta start somewhere
- Location: Near Dallas, Texas
- Main keyboard: Unsaver | 3276 | Kingsaver
- Main mouse: Kensington Slimblade
- Favorite switch: Capacitative Buckling Spring
- DT Pro Member: -
- Contact:
So solder out just the controller and use a mulitmeter to test it? I don't even know what a multimeter is to be honest. Or solder out a switch and test with the tester? I am a total noob at this stuff, completely new to me so I might need some spoonfeeding. Neat switch design BTW, I had no idea about the switches.
- XMIT
- [ XMIT ]
- Location: Austin, TX area
- Main keyboard: XMIT Hall Effect
- Main mouse: CST L-Trac Trackball
- Favorite switch: XMIT 60g Tactile Hall Effect
- DT Pro Member: 0093
(don't worry about being new to this, just Google anything you haven't seen before)
A multimeter is a piece of electronic testing equipment. It usually tells you resistance between two points, and can also check for continuity. Whenever the two test points on the multimeter are shorted it goes Beep. This is good for understanding how a circuit is laid out.
One easy thing to check for: what do the pins on the main connector actually go to inside the board? What pins are those on the 8048? Consult an 8048 data sheet for pin names and numbers, something like:
http://pdf1.alldatasheet.com/datasheet- ... 048AH.html
A multimeter is a piece of electronic testing equipment. It usually tells you resistance between two points, and can also check for continuity. Whenever the two test points on the multimeter are shorted it goes Beep. This is good for understanding how a circuit is laid out.
One easy thing to check for: what do the pins on the main connector actually go to inside the board? What pins are those on the 8048? Consult an 8048 data sheet for pin names and numbers, something like:
http://pdf1.alldatasheet.com/datasheet- ... 048AH.html
- Redmaus
- Gotta start somewhere
- Location: Near Dallas, Texas
- Main keyboard: Unsaver | 3276 | Kingsaver
- Main mouse: Kensington Slimblade
- Favorite switch: Capacitative Buckling Spring
- DT Pro Member: -
- Contact:
What exactly do you mean by that? Do you mean where do they go? Are you talking about the outer connector or the one on the PCB?XMIT wrote: ↑
One easy thing to check for: what do the pins on the main connector actually go to inside the board?
I found this: http://html.alldatasheet.com/html-pdf/6 ... 048AH.html
So could I use the info on this sheet for wiring up the teensy? Still a little confused on which exactly.
I have a full pic of the board below: Or do you want a pic of the whole PCB?
- XMIT
- [ XMIT ]
- Location: Austin, TX area
- Main keyboard: XMIT Hall Effect
- Main mouse: CST L-Trac Trackball
- Favorite switch: XMIT 60g Tactile Hall Effect
- DT Pro Member: 0093
Over PMs it sounds like you're having some trouble figuring out how to drive/sense these magnetic valve switches.
These switches have four pins, let's call them 1-2-3-4. You would connect pins 1 and 3 to a microcontroller and pins 2 and 4 to ground. Then, you send a pulse down pin 1, and see if you can detect a pulse on pin 3. If you can, the switch is not pressed. Otherwise, the switch is pressed.
It may be important to get the size of the pulse just right. For that you would ideally use an analog function generator and a two channel oscilloscope to get a "feel" for what the response of the key switch is. In this case, you would drive pins 1 and 2 with a square wave, and see what comes up on pins 3 and 4.
A cheap logic analyzer can also serve as an analog oscilloscope, such as:
http://www.ebay.com/itm/111540669744
, and use it with: http://sigrok.org/wiki/Supported_hardware
Anyone with a benchtop function generator and oscilloscope want to volunteer and see what the square wave response is for these switches across a range of frequencies at a 50% duty cycle? A 0-5V square wave should do it. I'd start at 10Hz or so and see what it looks like, getting a feel for: how far the switch needs to be depressed to change the output waveform, the latency/max scan rate, and how "rounded" the output waveform is.
These switches have four pins, let's call them 1-2-3-4. You would connect pins 1 and 3 to a microcontroller and pins 2 and 4 to ground. Then, you send a pulse down pin 1, and see if you can detect a pulse on pin 3. If you can, the switch is not pressed. Otherwise, the switch is pressed.
It may be important to get the size of the pulse just right. For that you would ideally use an analog function generator and a two channel oscilloscope to get a "feel" for what the response of the key switch is. In this case, you would drive pins 1 and 2 with a square wave, and see what comes up on pins 3 and 4.
A cheap logic analyzer can also serve as an analog oscilloscope, such as:
http://www.ebay.com/itm/111540669744
, and use it with: http://sigrok.org/wiki/Supported_hardware
Anyone with a benchtop function generator and oscilloscope want to volunteer and see what the square wave response is for these switches across a range of frequencies at a 50% duty cycle? A 0-5V square wave should do it. I'd start at 10Hz or so and see what it looks like, getting a feel for: how far the switch needs to be depressed to change the output waveform, the latency/max scan rate, and how "rounded" the output waveform is.
- XMIT
- [ XMIT ]
- Location: Austin, TX area
- Main keyboard: XMIT Hall Effect
- Main mouse: CST L-Trac Trackball
- Favorite switch: XMIT 60g Tactile Hall Effect
- DT Pro Member: 0093
Austin Goodwill is selling one of these:
http://www.ebay.com/itm/321862216781
I'll see if I can pick it up cheaply.
http://www.ebay.com/itm/321862216781
I'll see if I can pick it up cheaply.
- Redmaus
- Gotta start somewhere
- Location: Near Dallas, Texas
- Main keyboard: Unsaver | 3276 | Kingsaver
- Main mouse: Kensington Slimblade
- Favorite switch: Capacitative Buckling Spring
- DT Pro Member: -
- Contact:
Just FYI, these switches are very scratchy and hard to press. I am more focusing on my Fujitsu leaf spring board now.XMIT wrote: ↑Austin Goodwill is selling one of these:
http://www.ebay.com/itm/321862216781
I'll see if I can pick it up cheaply.
- seebart
- Offtopicthority Instigator
- Location: Germany
- Main keyboard: Rotation
- Main mouse: Steelseries Sensei
- Favorite switch: IBM capacitive buckling spring
- DT Pro Member: 0061
- Contact:
Very nice what you got there Redmaus. Good to see you're getting into real old keyboards now. Sorry I can't give you any tipps for the conversion.What does the switch feel like?
- Muirium
- µ
- Location: Edinburgh, Scotland
- Main keyboard: HHKB Type-S with Bluetooth by Hasu
- Main mouse: Apple Magic Mouse
- Favorite switch: Gotta Try 'Em All
- DT Pro Member: µ
Looks like he just said it! A lot of the really vintage stuff falls into that category. Random ass linear, with a shot of scratchy on the side.
- XMIT
- [ XMIT ]
- Location: Austin, TX area
- Main keyboard: XMIT Hall Effect
- Main mouse: CST L-Trac Trackball
- Favorite switch: XMIT 60g Tactile Hall Effect
- DT Pro Member: 0093
Those are ITW/Cortron Magnetic Valve switches. They are a kind of switch that I understand quite well and of which I have several other examples here in the house. I am already working on converting a similar keyboard to USB. The switches are only scratchy because they are quite dirty. Much like Micro Switch Hall effect key switches, these have the potential to be completely smooth with some love.
- seebart
- Offtopicthority Instigator
- Location: Germany
- Main keyboard: Rotation
- Main mouse: Steelseries Sensei
- Favorite switch: IBM capacitive buckling spring
- DT Pro Member: 0061
- Contact:
Absolutely XMIT, after I cleaned my Micro Switch hall effect switches they are by far the smoothest linears I own. Brilliantly simple in design too.
- Redmaus
- Gotta start somewhere
- Location: Near Dallas, Texas
- Main keyboard: Unsaver | 3276 | Kingsaver
- Main mouse: Kensington Slimblade
- Favorite switch: Capacitative Buckling Spring
- DT Pro Member: -
- Contact:
Maxi switch as in hall effect? Or is that micro switch? Hall effect is 5x better than the cortron one in my opinion
- XMIT
- [ XMIT ]
- Location: Austin, TX area
- Main keyboard: XMIT Hall Effect
- Main mouse: CST L-Trac Trackball
- Favorite switch: XMIT 60g Tactile Hall Effect
- DT Pro Member: 0093
Maxi Switch, see: http://deskthority.net/keyboards-f2/xer ... t9233.html
...which links to...
http://deskthority.net/wiki/Xerox_928-900451
http://deskthority.net/wiki/Maxi_Switch_vintage_linear
What is the model number of your Xerox keyboard Redmaus? It would be printed on a sticker on the back.
...which links to...
http://deskthority.net/wiki/Xerox_928-900451
http://deskthority.net/wiki/Maxi_Switch_vintage_linear
What is the model number of your Xerox keyboard Redmaus? It would be printed on a sticker on the back.
- XMIT
- [ XMIT ]
- Location: Austin, TX area
- Main keyboard: XMIT Hall Effect
- Main mouse: CST L-Trac Trackball
- Favorite switch: XMIT 60g Tactile Hall Effect
- DT Pro Member: 0093
It's here! And, good news Redmaus, the switches appear to be ITW/Cortron at first glance (and definitely not Maxi Switch discrete switches as some flavors of this keyboard offered).
So, if I ever find the time to work on it, I can help you get this keyboard working over USB, since I have an identical one here in the house. Not a bad deal for $16 or so from the local Goodwill chapter. (Plug: support your local thrift store whenever possible.)
So, if I ever find the time to work on it, I can help you get this keyboard working over USB, since I have an identical one here in the house. Not a bad deal for $16 or so from the local Goodwill chapter. (Plug: support your local thrift store whenever possible.)