Success! Displaywriter Beamspring using a Model F Xwhatsit Controller!

[orihalcon]

First off, reason for doing this is that there aren't any Displaywriter Xwhatsit Beamspring controller group buys on the horizon since they are unique to Displaywriters and can't be used with other Beamspring boards. This makes it hard to find enough people to go in on a buy and have the price be reasonable especially with how few unconverted Beamsprings remain given previous group buys.

The theory was that since Displaywriters use a 12x8 matrix, the Model F Xwhatsit Controller could be used as Model F’s use 16x8. Problem is that Beamsprings have the capsense pad down at rest whereas model F's have the capsense pad up at rest, so the state would need to be inverted from the default in firmware. Joc was kind enough to modify the firmware for me and tried it out and I can confirm it works!

8.jpg (830.21 KiB) Viewed 11024 times

Steps were:

1. Remove original PCB connector from original displaywriter controller - I did desolder it, but eventually decided to bend off the small pins that went into the PCB as I was planning to solder to the larger part of the connector that came out of the housing.

1.jpg (1.94 MiB) Viewed 11024 times

2. Mark out Which pins were grounds and rows vs columns on the Displaywriter. The outer 4 pins on either sides are the rows. The center 12 pins are columns. The others are unused.

2.jpg (1.54 MiB) Viewed 11024 times

3.jpg (549 KiB) Viewed 11024 times

3. Run small solid core wire from these positions to those on the Xwhatsit Model F controller’s corresponding locations. On the Model F Xwhatsit, there are 2 ground pins on either end, columns are the next 8 on the left front of the controller, 2 more ground pins, then the 16 columns. Orange wire was used for the rows and blue for the columns. Green wires are ground.

4.jpg (1.5 MiB) Viewed 11024 times

5.jpg (1.44 MiB) Viewed 11024 times

6.jpg (1.89 MiB) Viewed 11024 times

4. You may connect a solenoid and driver if desired.

7.jpg (1.29 MiB) Viewed 11024 times

5. Run a ground wire from any of the ground pins to the case.

8.jpg (830.21 KiB) Viewed 11024 times

Video demo here:
https://youtu.be/kto87pWCmmo

If anyone wants to convert their beam spring this way, let me know. I have Model F Xwhatsit controllers that I can pre-flash with this firmware for the same price as normal. I also have solenoid drivers and genuine IBM solenoids for sale as well if anyone is interested.

Hopefully this will bring some more Displaywriters back to life!

Firmware is also attached if anyone wants to try this with their existing Xwhatsit controller.

[joc]

Great! I'm glad you got it to work!

[mike52787]

I may do this with my 5251 due to the relative low supply/rarity/price of beamspring xwhatsits at the current time.

[Wodan]

Damn thanks for putting this together, you're a hero!

[Cattus_D]

Wow! Thank you for sharing!

[gazza341]

nice one. glad you got it to work, and thanks for sharing. neat wiring and soldering job.

[orihalcon]

I may do this with my 5251 due to the relative low supply/rarity/price of beamspring xwhatsits at the current time.

Would be great if it would work with the 5251's and other beamsprings, but all other beamsprings use a 24 x 4 matrix unfortunately, so the Model F Xwhatsit does not have enough columns to support them. The Displaywriter is the only oddball with a 12x8 matrix. The Model F Xwhatsit can support up to a 16x8 matrix only unfortunately.

[seebart]

Great job thanks for posting this orihalcon. I did know it was possible to use the Model F Xwhatsit Controller for the Displaywriter Beamspring but this really helps.

[Vizir]

Do you still have the controller, drivers and solenoids?

Sent from my Nexus 6P using Tapatalk

[JP!]

Looking good!

[rsbseb]

Neat old keyboard with an interesting layout. What are your plans for a case?

[orihalcon]

Do you still have the controller, drivers and solenoids?

Sent from my Nexus 6P using Tapatalk

I'm about to post some pre-made Model F Displaywriter controllers like what is shown in this post in the for sale section. You can also just buy the Model F Xwhatsit and I can put the displaywriter firmware on it at no charge, though note that this only will work for the displaywriter given the number of rows and columns available to use. The soldering is kind of time consuming, so would recommend buying pre-wired by me unless this is something that you enjoy doing.

I will be getting my shipment of "Non-Displaywriter" aka "Regular" beamspring controllers here in the next week or two. Have to test them all with an oscilloscope to confirm they are working before putting them up for sale.

I do have a limited supply of drivers and solenoids, but will list everything at once so that shipping can be combined to keep prices low, especially for international orders as these will be shipping from the USA.

I actually also have a couple of solenoid drivers that tested bad from the factory that I'd sell for cheap if anyone wants to troubleshoot. Most have all of the components on them, but I don't have a good way of chasing down which component is bad. PM me about that if interested.

[orihalcon]

Neat old keyboard with an interesting layout. What are your plans for a case?

... Though may integrate this into a desk at some point kind of like the System 3 that JP! just found

[Vizir]

Do you still have the controller, drivers and solenoids?

Sent from my Nexus 6P using Tapatalk

I'm about to post some pre-made Model F Displaywriter controllers like what is shown in this post in the for sale section. You can also just buy the Model F Xwhatsit and I can put the displaywriter firmware on it at no charge, though note that this only will work for the displaywriter given the number of rows and columns available to use. The soldering is kind of time consuming, so would recommend buying pre-wired by me unless this is something that you enjoy doing.

I will be getting my shipment of "Non-Displaywriter" aka "Regular" beamspring controllers here in the next week or two. Have to test them all with an oscilloscope to confirm they are working before putting them up for sale.

I do have a limited supply of drivers and solenoids, but will list everything at once so that shipping can be combined to keep prices low, especially for international orders as these will be shipping from the USA.

I actually also have a couple of solenoid drivers that tested bad from the factory that I'd sell for cheap if anyone wants to troubleshoot. Most have all of the components on them, but I don't have a good way of chasing down which component is bad. PM me about that if interested.

Can you put me down for the following:
1. The new controller for displaywriter
2. A solenoid driver (maybe 2, I'll check and let you know for sure next week)
3. A solenoid (perhaps 2 if you have enough)
4. A beamspring controller - non-displaywriter.

Thanks!

Sent from my Nexus 6P using Tapatalk

[consolation]

I may do this with my 5251 due to the relative low supply/rarity/price of beamspring xwhatsits at the current time.

Would be great if it would work with the 5251's and other beamsprings, but all other beamsprings use a 24 x 4 matrix unfortunately, so the Model F Xwhatsit does not have enough columns to support them. The Displaywriter is the only oddball with a 12x8 matrix. The Model F Xwhatsit can support up to a 16x8 matrix only unfortunately.

Why not just use two and have them show up as two separate keyboards doing half the keyboard each? There seems to plenty of room inside the cases for a small hub so you only have one cable coming out. My old CM Mech did this to get true NKRO over USB 2.0.

[PancakeMSTR]

What was the process for modifying the code to deal with the rest state being up vs down? Could I get more details on that?

Also could this be applied to a 24x4 grid?

[orihalcon]

What was the process for modifying the code to deal with the rest state being up vs down? Could I get more details on that?

Also could this be applied to a 24x4 grid?

You'd have to PM Joc as far as how he modified the firmware, but I assume it was a single line of code that just inverted the key state, though to be honest, I know nothing about programming

Wouldn't be possible with the Model F Xwhatsit as the number of rows and columns is determined by the hardware. However, Beamspring controllers use a 24x4 matrix, so if you have some other keyboard that you wanted to have a similar matrix for but with that keystate inverted from that, I would think that Joc could probably make that modification for you. I do have the beamspring version available if you need one.

Hope that helps!

[XMIT]

Though I've never gotten set up to do it, you can build xwhatsit from source.

[gnarlsagan]

@orihalcon I'm having an issue with my xwhatsit Displaywriter controller and was hoping you could offer some insight.
I had some chatter, and some keys not registering, but most registered fine, so I thought I would adjust the voltage. After setting the voltage to something like 80, none of the keys would actuate. Setting the voltage to 107 causes keys to sporadically actuate on their own, but keys will not actuate if I press them. It really is odd.
So it seems I can't find a voltage at this point that will allow any key to actuate when I press it, but the keys show as actuated without being pressed on the xwhatsit software when the voltage is raised.
Also, I tried the auto calibration feature for the voltage, but it recommended 38, which did not work. Should I perhaps be deactivating rows 13-16 in the software?
I also tried reseating the controller lining it up multiple ways to the left and right so the contacts meet, but maybe there is a particular way that I'm missing. Here are photos:
http://imgur.com/qXwTYg4
http://imgur.com/K7EPMCn
Do you think there is any way the controller could have been damaged in some way or misconfigured? My PC still recognizes it (although as an aside the software does not work on my Mac), and I am able to save changes to the eeprom without issue.
Thank you so much for any insight or help you could offer!

[JP!]

@orihalcon I'm having an issue with my xwhatsit Displaywriter controller and was hoping you could offer some insight.
I had some chatter, and some keys not registering, but most registered fine, so I thought I would adjust the voltage. After setting the voltage to something like 80, none of the keys would actuate. Setting the voltage to 107 causes keys to sporadically actuate on their own, but keys will not actuate if I press them. It really is odd.
So it seems I can't find a voltage at this point that will allow any key to actuate when I press it, but the keys show as actuated without being pressed on the xwhatsit software when the voltage is raised.
Also, I tried the auto calibration feature for the voltage, but it recommended 38, which did not work. Should I perhaps be deactivating rows 13-16 in the software?
I also tried reseating the controller lining it up multiple ways to the left and right so the contacts meet, but maybe there is a particular way that I'm missing. Here are photos:
http://imgur.com/qXwTYg4
http://imgur.com/K7EPMCn
Do you think there is any way the controller could have been damaged in some way or misconfigured? My PC still recognizes it (although as an aside the software does not work on my Mac), and I am able to save changes to the eeprom without issue.
Thank you so much for any insight or help you could offer!

I ran into the same issues with mine, tried reseating, etc, etc but once I got the voltage dialed in just right it worked perfect.

I do know the the voltage threshold varies quite a bit from board to board, usually is in the 110 to 150 range. Mine works perfect at 120 and didn't even need to use the auto calibration.

Also if you plan to do any restoration work I would recommend cleaning the pcb as it is possible there could be contaminants such as dirt and dust either on the pcb or flippers.

[gnarlsagan]

@orihalcon I'm having an issue with my xwhatsit Displaywriter controller and was hoping you could offer some insight.
I had some chatter, and some keys not registering, but most registered fine, so I thought I would adjust the voltage. After setting the voltage to something like 80, none of the keys would actuate. Setting the voltage to 107 causes keys to sporadically actuate on their own, but keys will not actuate if I press them. It really is odd.
So it seems I can't find a voltage at this point that will allow any key to actuate when I press it, but the keys show as actuated without being pressed on the xwhatsit software when the voltage is raised.
Also, I tried the auto calibration feature for the voltage, but it recommended 38, which did not work. Should I perhaps be deactivating rows 13-16 in the software?
I also tried reseating the controller lining it up multiple ways to the left and right so the contacts meet, but maybe there is a particular way that I'm missing. Here are photos:
http://imgur.com/qXwTYg4
http://imgur.com/K7EPMCn
Do you think there is any way the controller could have been damaged in some way or misconfigured? My PC still recognizes it (although as an aside the software does not work on my Mac), and I am able to save changes to the eeprom without issue.
Thank you so much for any insight or help you could offer!

I do know the the voltage threshold varies quite a bit from board to board, usually is in the 110 to 150 range. Mine works perfect at 120 and didn't even need to use the auto calibration.

Thanks JP. But at this point depressing keys does not cause any keys to actually actuate in the xwhatsit software or on a key tester app. It unfortunately doesn't matter what voltage I select. Going to 107 voltage or above causes random keys to show as grey/actuated in the xwhatsit software, however, physically depressing those keys will not actuate them.

I've also tried different cables to no effect. I've also disabled rows 13-16, and settings save to the eeprom with no issue.

Before this keys did register upon being depressed, but some did not and others chattered a lot. So I tried modifying the voltage and ended up in this situation. The seller of the board insists to me that the board was in fully working condition before shipment.

So I'm thinking either the controller got loose or damaged or misconfigured somehow, and I was hoping there could be a fix for an issue like this. I have tried reseating the controller and realigning it to no effect.

[JP!]

I had the same symptoms as well with random keys showing up as grey or actuated. I literally started at 200 and worked my way down, one increment at a time until more and more keys started to register and the random key presses disappeared. It was like a fine line between madness and genius for me. Once I found the right value I then wrote it to eeprom.

Perhaps Orihalcon or someone here can comment further to this?

[Sangdrax]

JP!, you pretty much nailed it. The only thing I'd add is that I had to increase my thresholds a few millivolts once I connected solenoid boards to kill chatter. Something about the current pulse through the controller I think.

Oh, and that the reference pads will show up as depressed as soon as you hit the nominal upper limit on the way down. That tells you you're in the right ballpark.

[gnarlsagan]

I had the same symptoms as well with random keys showing up as grey or actuated. I literally started at 200 and worked my way down, one increment at a time until more and more keys started to register and the random key presses disappeared. It was like a fine line between madness and genius for me. Once I found the right value I then wrote it to eeprom.

Perhaps Orihalcon or someone here can comment further to this?

So when you had random keys show up as grey/actuated, were you able to physically press other keys and see them also become grey/actuated? Now, no keys that I physically press show as grey/actuated in the xwhatsit software, although keys that are not physically depressed will randomly show up as grey/actuated (chatter). I can increase the voltage to the point where all keys show up as grey/actuated, and there is no point during the incremental increase of voltage that any physically depressed key will ever show as grey/actuated.

I hope that's clear. It seems like the keys and/or PCB is not communicating correctly with the controller. And to reiterate, before I started trying to adjust the voltage, most keys did work as expected, but a few keys did not register, and other keys would constantly chatter.

I really appreciate the help guys. I'd love to get this Displaywriter back into prime condition! I'm thinking it might behoove me to pick up another controller in case this one is broken for some odd reason. I have no idea what would cause it to stop working like this, unless it came loose or was unseated somehow, and I'm just too obtuse to reseat it correctly, but that seems unlikely since I've tried reseating it many times with no effect.

I should also mention that the green ground wire is correctly screwed to the case.

[Sangdrax]

Have you opened up the board module and cleaned between the PCB and the switches and all that? Cause it just takes a few little chunks of stuff to get down in there and cause chatter. 107 also seems a little low. I had to set mine around 138 I think.

You should have a couple reference pads show up as grey always too in the matrix. They're not actual keys.

[gnarlsagan]

Have you opened up the board module and cleaned between the PCB and the switches and all that? Cause it just takes a few little chunks of stuff to get down in there and cause chatter. 107 also seems a little low. I had to set mine around 138 I think.

You should have a couple reference pads show up as grey always too in the matrix. They're not actual keys.

Thanks I was definitely considering cleaning off the PCB. Regarding voltages, I have tried up to 170 iirc at which point all keys will show as grey/actuated, but at no point up to 170 did physically depressing any switch cause that switch to show as grey/actuated.

So, this is not a voltage setting issue, because at no setting between "no switches showing as grey" and "all switches showing as grey" does physically depressing a switch cause it to show as grey/actuated in the xwhatsit software.

So, for some reason switch presses aren't being read by the controller, but the controller is outputting chatter to the PC at around the 107 voltage mark. Going above 107 results in large swaths of the board showing up as grey/actuated.

Also, if I try to use the auto voltage function, it gives me the number 38, but this does not fix the above issue.

Do I need to set reference pads manually? Thanks again for the pointers.

[Sangdrax]

It might be that the controller firmware switch rest position wasn't reversed by mistake. Run the voltage up to like 115 where everything is actuated and try pressing some keys and see if any of the squares goes white.

Auto voltage just plain doesn't work that I've seen, so I wouldn't bother with it.

[orihalcon]

I tend to agree with everyone here that auto voltage calibration tends to not work. 120 usually works on the displaywriters but does vary from board to board. Kind of odd that it worked at first before modifying the voltage as there shouldn't be any damage done by picking a different value and going back. If the issue persists, I can exchange it for you if you got it from me.

Hope that helps!

[gnarlsagan]

I tend to agree with everyone here that auto voltage calibration tends to not work. 120 usually works on the displaywriters but does vary from board to board. Kind of odd that it worked at first before modifying the voltage as there shouldn't be any damage done by picking a different value and going back. If the issue persists, I can exchange it for you if you got it from me.

Hope that helps!

Hi orihalcon! Thanks for reaching out! I sent you a PM.

[JP!]

Hopefully things work out for you gnarlsagan. I think auto voltage is a bit misleading. Get it close with manual adjustment and if you end up with a couple for example that have trouble registering, then try the auto voltage. I recall the manual covers this feature in more detail.