PESA K-4722 Subtitler/Videotext Keyboard

[DevilZmods]

Hi, First post, go easy on me.

I got this Board on eBay for its colourful Keycaps. Its referenced once in this forum and "sort of cherry compatible" seemed worth taking the shot. Heres the link to the old post: viewtopic.php?f=2&t=18616

I want to write down everything i could dig up about this board:

s-l1600.jpg (297.14 KiB) Viewed 1371 times

This particular one has a german layout with umlauts and ISO style num-row, however the "," and "." keys are ANSI.
The case is constructed from two welded and painted sheetmetall halves and has two small ribs on the underside to angle it slightly.

s-l1601.jpg (285.43 KiB) Viewed 1371 times

s-l1602.jpg (260.01 KiB) Viewed 1371 times

It has a fairly minimal alphanumeric block and a bunch of Function keys specific to its purpose as a Subtitle/Videotext device.
The Keyboard has a 6P6C Modular Connector socket as its only output.

IMG_20200422_190438_676.jpg (3.89 MiB) Viewed 1371 times

The serial number doesn't really tell us anything, a sticker on one of the microcontrollers has a handwritten date on it: 25.11.87
From a logistics point of view i can only imagine someone covering the ICs labeling during assembly, not before.

PESA Electronics is an american company that's still in business, but the closest thing to keyboards they do nowadays, are button arrays for server rack mounnting to go along with their networking equipment.
The only other mention of this board is a French media company listing it along other equipment for two of their post-production rooms: https://www.sonygraf.com/instala_in.html

On to the switches:

s-l1604.jpg (270.47 KiB) Viewed 1371 times

IMG_20200422_185912_503.jpg (2.58 MiB) Viewed 1371 times

The board has an interesting spacebar setup, where the switch sits on one side, while the stablilizer itself also has a spring inside.

IMG_20200422_192629_472.jpg (2.06 MiB) Viewed 1371 times

IMG_20200422_193231_484.jpg (3.4 MiB) Viewed 1371 times

As assumed in the other thread, the switches are Clare/Pendar, in fact they are SFL5101, as described in this wiki article: wiki/Clare/Pendar_low_profile_reed
There is also two illuminated SIL5101.

Here's a teardown of that switch:

IMG_20200422_192838_350.jpg (2.32 MiB) Viewed 1371 times

IMG_20200422_193015_516.jpg (2.2 MiB) Viewed 1371 times

One interesting thing to note is the free floating plastic appendix on the switch stem, whose function is presumably to dampen the switch bottoming out. It is undoubtedly the reason for the switches mushy feeling at the end of the keypress.

The fact that these are reed switches makes me want to keep the board intact, but i will probably never get the pcbs logic components to work with modern equipment. If anyone knows what to do with this, or wants it for collecting, let me know.

The keycaps are thick doubleshot abs (assuming from the yellowing) with the legend colour inner cup extending all the way to the bottom of the caps on most caps. On some that inner piece doesnt go s far down, which lets me assume that the molds got a redesign at some point. The distribution of older and newer style caps is random. The red keycaps stick out, in the way that the plastic looks and feels cheaper, more like modern toy abs plastic.
These Caps are in a somewhat sorry state, with some of the stems shattered. I can only repeat what has been said in the Previous post, that they are "somewhat cherry compatible". The stems sit very tight and without trimming the internal ribbing around the stem, most caps wont fully seat on a cherry style switch. That said i transferred the alphas to my planck and they look and sound swell, but i still have to fix up some of the stems and decide whether i want to take a dremel to the ribbing.

I also plan on doing a quick spotlight on how i glue these shattered keystems.

Well for now that's it for my essay, hope you get a kick out of it.

[kps]

The serial number doesn't really tell us anything, a sticker on one of the microcontrollers has a handwritten date on it: 25.11.87
From a logistics point of view i can only imagine someone covering the ICs labeling during assembly, not before.

On the contrary, that's probably the date the EPROM was burned, so, before assembly, but probably not terribly long before. (You can tell it's EPROM by the circular window outline.)