What's wrong with 7bit's Avatar (keyboard)?

User avatar
RC-1140

19 Mar 2012, 18:44

Hi everybody,
recently I have bought a keyboard which probably everybody has seen at least once. The Terminal Model F from 7bit's Avatar. Yup, that's right, exactly the one from 7bit's avatar. It looks absolutely beautiful, and the capacative buckling springs go all "Kaploink", but when I tried to connect it to my PC using Soarer's Teensy converter the red LED on my Teensy lit up. I panicked and pulled the cable, and rechecked all the solder joints. When I found none I tried it again. Same result. Then I overcame my fears and opened the Board. Once again the build quality of IBM keyboards amazed me, but when I looked at the cable, I was shocked. The red (data) lead was missing most of its insulation. I held the cable up, and connected it again, this time the keyboard worked. I was able to type. Then suddenly the keyboard sent a whole lot of strange keycodes, and looking on the cable I realized I must have let my hand down. Upon lifting the lead again it worked again. I tried insulating the lead, and tried it again, but once again, only the red LED lit up. I checked all cables again, and measured the cable with a multimeter, but everything seemed alright. Then I tried connecting it with a cable of an Terminal Model M which had the same pinout, but once again I only saw the red light. But another thing which is confusing me, is that the LED lights up after some unspecific time. Sometimes right after connecting, sometimes after the lock LEDs flashed up (which I added to the teensy). Right now I'm really sad, to look on a poor old Model F which doesn't work. Does anybody know what it might be?

:cry: :cry: :cry:

RC-1140

User avatar
Maarten

19 Mar 2012, 20:32

If i were you i'd check the board a bit better before using it... sound like it may have shorted out a bit in the past. Could be some loose stuff inside....

User avatar
Ascaii
The Beard

19 Mar 2012, 21:24

I would recommend reflowing the other leads as well...sounds like one might be partially cold, causing a "Wackelkontakt".

User avatar
RC-1140

19 Mar 2012, 21:27

Well, I made the mistake of believing that 7bit tried it already. Anyhow I don't really think that it's completely dead. As I said, I got it to work once, and the painting on the metal plate seems to be isolating. Is anything known about the capacitors on the model Fs? I could imagine, that one of the capacitors is defective/dry. But before I start tinkering around I really want to know the opinion of other Model F fans, as I'm pretty unexperienced with them and would like to preserve its original state as good as possible.

Edit: with lead I didn't mean a soldering joint, I meant the wire. The Instructions on the inside said lead, so I used this word.

User avatar
Ascaii
The Beard

19 Mar 2012, 22:26

I know. Still sounds like a wiggly joint to me^^.

User avatar
RC-1140

30 Mar 2012, 18:23

Hey guys, here's a update:

It lives! The Model F is working using the Teensy Converter! In case somebody else has a similar problem with a Model F:

The problem were the screws connecting the metal plate and the controller board. Apparently IBM used this point as a GND connection for the controller, so if there is no connection between the controller PCB and the shiny backplate, the keyboard doesn't work.

But now I'm typing on a fantastic pinging Model F 122-Key Terminal Board. But I won't use it as a daily driver. It is just too loud. Even though I like the Model M ping, this board is just too much ping. I wonder whether it's possible to play some melodies on it... Anyhow I love the feeling of the switches. They are a lot lighter than ordinary buckling springs, and have an even sharper Actuation point. Kinda like Ergo Clears with a sharp actuation and lots of ping.

Great Board if it works!

User avatar
Ascaii
The Beard

30 Mar 2012, 20:07

Awesome! Sounds great!

User avatar
Soarer

30 Mar 2012, 21:48

Aha! So it's fixed, and as raucous as it should be :-D

I didn't think of that grounding wire when I saw this thread before, but yeah, the capactive switches need that shielding it seems.

User avatar
RC-1140

30 Mar 2012, 23:18

Yup, as an Model M fan I didn't know this, so I didn't inspect it at first. I'm happy that it's fixed now, it wasn't too cheap for me. But it was my dream keyboard after all, so it was definitely worth it. The capacitive Buckling Springs are by far the best switches I've tried yet. If they weren't so loud that my mother complained about it from outside my room with the door closed...

User avatar
webwit
Wild Duck

30 Mar 2012, 23:23

Easily the best switches that can be used on a modern pc!

User avatar
RC-1140

30 Mar 2012, 23:48

I wonder how Beam Spring switches feel. But I feel like I could not afford them at all. They are only available in IBM Selectrics, right? I would love one of them, even though I wouldn't have any use for it.

Edit: Not IBM Selectric, I meant the IBM Displaywriter.

What switches do the Selectric use btw?

User avatar
Soarer

31 Mar 2012, 00:19

I think the DisplayWriter might be the most recent and most common, and imho the best beam spring layout, but there were also some terminal 'boards that had beams. Perhaps the 3278? Looks like it does: ex1 ex2.

(This WILL be the year I get my DisplayWriter 'board USB-ified!!)

ripster

31 Mar 2012, 01:44

Lol.

WackelKontact.

Funny German words crack me up.
Image

User avatar
webwit
Wild Duck

31 Mar 2012, 03:07

My beamspring keyboards:

Image

Image

Image

User avatar
bhtooefr

03 Apr 2012, 16:42

RC-1140 wrote:I wonder how Beam Spring switches feel. But I feel like I could not afford them at all. They are only available in IBM Selectrics, right? I would love one of them, even though I wouldn't have any use for it.

Edit: Not IBM Selectric, I meant the IBM Displaywriter.

What switches do the Selectric use btw?
The Selectric doesn't use switches.

Its keys physically engage the printing mechanism of the typewriter. The Selectric does take electrical power, but that is only to run an AC motor that drives the mechanism - you can actually hand-crank a Selectric and it still works with no electricity whatsoever.

Even the I/O mechanisms just added reed switches for getting key presses and solenoids for acting on the print mechanism to the Selectric design, but the actual key actuation was still the Selectric mechanism itself.

I highly recommend finding a Selectric (ideally a I or II, but the III will work, too) and typing on it, but it's not beam spring. It's a fairly light (feels like about 60 cN, IIRC) keypress, with nearly zero resistance until the tactile point, at which it feels like the key is pushing something out of the way (which it actually is, to fall into the interposer mechanism, IIRC), and then as soon as you make your way through the tactile point, the key nearly falls out from underneath your finger. I'd describe it as... almost like the smoothest Cherry blue you've ever used, but 100 times better.

Beam spring is an attempt to imitate the Selectric without actually having to have a mechanical typewriter mechanism. IIRC, we don't have a full list of the machines using beam spring, but it includes the 3276, 3278, and 3279 mainframe terminals, 5251 midrange terminal (the mechanical layout of this keyboard, FWIW, eventually was slightly tweaked (key sizes, but not positions) into the original Model F layout used on the System/23 Datamaster, which was reused as the original PC and XT keyboard), 5100 luggable, DisplayWriter word processor, and some others that I can't remember off the top of my head. I'd describe beam spring as a fairly light switch, in the 55 cN ballpark. Fairly low initial force, but it ramps up quickly, and then goes fairly linearly until the buckling point (note that you don't really push through the buckling point, you just get to it), at which point it has a sharp buckling event (but not a very forceful one, it's not as tactile as a Model F, I'd say).

Buckling spring, in turn, is an attempt to make a cheaper switch that's similar to beam spring. The Model F is in the 65-70 cN ballpark, and I fully believe that, but it feels like all the additional weight is coming up towards the tactile point - initial force doesn't really feel much higher than the beam spring, and force after the tactile event doesn't feel significantly higher. Edit: Initial force is actually lower, which can help get you momentum into the tactile point.

The Model M is an attempt to cheapen buckling spring again, and due to the tolerances necessary, everything got stiffer and tactility got lost (and dulled, too). Force ranges from 65 to 80 cN, but even a 65 cN Model M (later Lexmark or Unicomp boards) feels stiffer than a Model F due to higher initial force and higher post-tactile event force.

Post Reply

Return to “Keyboards”