(Model MF) Remodeling the Model M (aka.. the Mara)

User avatar
lot_lizard

11 Aug 2016, 18:20

Techno Trousers wrote: I do think, however, that phase 2 will generally have more appeal to the Cherry/Topre/Alps enthusiasts more so than the IBM enthusiasts, especially if you end up creating 60%, Ergodox, White Fox, or other small/non-standard layouts.
That's another driving reason actually. All of the arguments about dated layouts, form factors, and cap choices would become a moot argument. They would be more willing to try and understand why we think the buckling spring is special. Plus Phase 1 will cover MOST of our desires as an IBM purist (especially if we wrap it up with a metal case mini GB later).
Techno Trousers wrote:
lot_lizard wrote:
Techno Trousers wrote: Not sure the barrel lock mechanism just yet (XT or AT style), but it will be one of them.

We want people to have the opportunity to add switches back to these harvested donor boards if they choose to in the future.
For point 1, let's make a mental note to have that choice finalized before phase 1 ordering opens, so we can know what type of phase 1 plate to order to make it compatible with the upcoming barrels. And for point 2, won't point 1 kind of make that impossible for the "losing" lock option? I suppose what would make the most sense is to choose the barrel type based on the distribution you're getting from your donor boards. My gut says that the AT style will be most plentiful, but maybe not.
For Phase 1, the top plate will support both barrel types (XT and AT). So even if you went with AT barrels out of the gate, XT barrels could be exchanged for it at any point in the future. Hopefully that will make it more appealing to those that aren't sure yet what they want to do about procuring switches since the option of either working will be there.


For Phase 2 (when we actually produce a new barrel from scratch), we will pick one to keep the mold costs down. Either way you would be able to order an extra set of barrels and return the original donor board back (though you would need to pull the barrels from the Phase 1 assembly to return to the donor if it was the alternative lock).

andrewjoy

11 Aug 2016, 18:23

I would say go for both the AT and XT style on one barrel , not only will it hold it in place better but people can buy them for ether type of board when restoring them and just snip off the part you don't need.

Failing that XT stile as the little pin is easy to remove and as long as its between 2 others you can use it in an AT /122 no problem as i have done on the very keyboard i type this on :)

User avatar
lot_lizard

11 Aug 2016, 18:28

andrewjoy wrote: I would say go for both the AT and XT style on one barrel , not only will it hold it in place better but people can buy them for ether type of board when restoring them and just snip off the part you don't need.

Failing that XT stile as the little pin is easy to remove and as long as its between 2 others you can use it in an AT /122 no problem as i have done on the very keyboard i type this on :)
Well done... I like it. I would even think we could make the AT style more like the XT "peg" so it is easier to snip off and still not impact the integrity of the "lock" (versus having it fused to the side of the barrel like it is today).

User avatar
alh84001
v.001

11 Aug 2016, 19:14

Did anyone ever speculate why IBM did the change from XT version to AT? It seems like such a small thing to get bothered about, so I can only assume there were some non-random reasons for doing it.

User avatar
fohat
Elder Messenger

11 Aug 2016, 19:46

alh84001 wrote: Did anyone ever speculate why IBM did the change from XT version to AT? It seems like such a small thing to get bothered about, so I can only assume there were some non-random reasons for doing it.
A "side buttress" is much more forgiving, probably both in manufacturing and installation.
You are shooting for a larger and easier-to-hit target, and it won't bend or break.

User avatar
lot_lizard

11 Aug 2016, 21:36

alh84001 wrote: Did anyone ever speculate why IBM did the change from XT version to AT? It seems like such a small thing to get bothered about, so I can only assume there were some non-random reasons for doing it.
I have wondered the same thing. The best conclusion I have ever come up with is the tie to the XT layout. The XT, Bigfoot, and the 3178 (Blue Switch) all use the peg lock, and have an XT layout. The 4704s, F107/122, and the AT use the "side buttress" <quoting Fohat> style, and more "ANSI like" layouts.


As far as timing, there is a substantial amount of time overlap between all of these (basically a full decade with similar start times), so benefit is questionable for any one over the other. I think it boiled down to more of the layout the new board would use (XT or not), and then they just augmented existing designs. Complete guess though, but the only rhyme or reason that has made sense to me thus far. I will say this, the XT spacebar is garbage by comparison, but the barrel material and dimensions aside from the "locking" mechanism are identical.

The real question would be the flipper paddle differences. No clue how to even guess there. I just opened up 33 F122's and the full versus "clipped" flipper was basically a 50/50 ratio with date having ZERO impact. XT's share similar flipper discrepancies.

User avatar
seebart
Offtopicthority Instigator

11 Aug 2016, 21:39

Was it really a change or were both not offered seperate from each other?

You are in fact right, good thing we have our wiki:
The Personal Computer Advanced Technology (PC AT), released in 1984, was the successor to the IBM PC and XT. Its revamped design was probably the final Model F keyboard design before the first Model M's appeared later that year. The layout was modified to address criticisms of the earlier PC keyboard, and somewhat resembled the layout of the IBM Displaywriter. The communications protocol between keyboard and computer was also redesigned, with bidirectional signals supporting lock lights. Despite its incompatibility with the older PC keyboard, the PC AT used the same DIN-5 connector. The AT keyboard protocol became standard, and is compatible with that of PS/2 keyboards, enabling it to be used with PS/2 connections via a simple DIN-to-PS/2 adapter. The US layout variant has the part number 6450200.

User avatar
Techno Trousers
100,000,000 actuations

11 Aug 2016, 22:11

seebart wrote:Was it really a change or were both not offered seperate from each other?

You are in fact right, good thing we have our wiki:
... Its revamped design was probably the final Model F keyboard design before the first Model M's appeared later that year.
I believe the AT keyboard was introduced in 1984, and the first 101 Ms came out in late 1985. I think a more likely progression was: AT (1984), F-122 (1984-85), which expanded upon the AT design by adding two top function key rows and a dedicated nav area, then the M 101 which tried to distill the F-122 layout into something more manageable and less expensive to produce, and which ended up setting the de facto keyboard layout going forward.

Note: I don't have any knowledge about the inner workings of IBM engineering. This is my speculation based on timing and layout changes.

When I build my FEXT, I already have a fake back story in mind: That it was a prototype lost to history; an initial version of the Model M layout, shrunk down from the F-122 and using internal components from that design. Management really liked it, so the mandate was to keep the layout but cut materials costs. Hence, the Model M was born. 8-)
Last edited by Techno Trousers on 11 Aug 2016, 22:16, edited 1 time in total.

User avatar
alh84001
v.001

11 Aug 2016, 22:15

lot_lizard wrote:
alh84001 wrote: The real question would be the flipper paddle differences. No clue how to even guess there. I just opened up 33 F122's and the full versus "clipped" flipper was basically a 50/50 ratio with date having ZERO impact. XT's share similar flipper discrepancies.
Interesting as well. All my XTs came with clipped flippers. Did those F122s have both kinds of unclipped flippers?

User avatar
seebart
Offtopicthority Instigator

11 Aug 2016, 22:16

You don't need to speculate:

wiki/IBM_Model_F

Check our data against wikipedia if you like.

User avatar
Techno Trousers
100,000,000 actuations

11 Aug 2016, 22:25

seebart wrote:You don't need to speculate:

wiki/IBM_Model_F

Check our data against wikipedia if you like.
Wikipedia says that the AT keyboard came out in 1983. Model M production started in 1985, so the F-122 bridged the gap between both of them.

User avatar
seebart
Offtopicthority Instigator

11 Aug 2016, 22:29

OK and the XT came out 81. Not sure if the F-122 was really intended to "bridge any gap" but I really do not know. But again wikipedia has other dates also on the XT:

https://en.wikipedia.org/wiki/IBM_Personal_Computer_XT
Release date March 8, 1983; 33 years ago
wiki/IBM_Model_F
The keyboard of the 1981 IBM PC 5150 is commonly referred to as the "PC XT" keyboard
I bet we could find this out though since IBM's online documentation is very good.

User avatar
Techno Trousers
100,000,000 actuations

11 Aug 2016, 22:37

seebart wrote:OK and the XT came out 81. Not sure if the F-122 was really intended to "bridge any gap" but I really do not know. But again wikipedia has other dates also on the XT:

https://en.wikipedia.org/wiki/IBM_Personal_Computer_XT
Release date March 8, 1983; 33 years ago
wiki/IBM_Model_F
The keyboard of the 1981 IBM PC 5150 is commonly referred to as the "PC XT" keyboard
I don't think it was planned or intended, per se, but the Model M layout definitely looks more like an evolution/scaling down of the F-122 than of the AT.

I think the problem with relying on information in the wikis for this is that we don't have definitive records of how and when they came up with the various layouts. We'll probably never know for sure on that, unless IBM has internal memos they'd like to share.

User avatar
fohat
Elder Messenger

11 Aug 2016, 22:44

seebart wrote:
Not sure if the F-122 was really intended to "bridge any gap"
The XT, AT, and M were "consumer" keyboards sold to individuals with personal computers.

The F 122-key terminal, and all the other terminals, were strictly commercial/industrial models and were never intended as keyboards for individual personal computers.

Therefore they were designed and sold to entirely different marketplaces.

User avatar
seebart
Offtopicthority Instigator

11 Aug 2016, 22:48

Techno Trousers wrote: I think the problem with relying on information in the wikis for this is that we don't have definitive records of how and when they came up with the various layouts. We'll probably never know for sure on that, unless IBM has internal memos they'd like to share.
That's one big problem with our wiki in general, not only getting the data but getting the correct data and being able to verify it. No, I don't think IBM will share development memos either they only post hard facts.

User avatar
lot_lizard

11 Aug 2016, 23:18

A very safe statement is that MOST ALL model F's (aside from the PC AT) initially launched in the early 80's and were in production for several years with overlapping dates. Both XT layouts with the "peg" and others with the "side buttress" where in production simultaneously... along with various flipper designs being used interchangeably throughout.

IBM is and was a VERY siloed organization with different hardware and software being produced under different organizational tiers that RARELY work together. It wasn't until Power Hardware in the early 2000's that the p (AIX) and i (AS/400) collaborated in force. The z is still it's own animal. These trends would go back to the System 36/38, and everything else they had in place before that. The point is that there were substantial liberties the designers had to pick how that platform worked. As Seebart mentioned earlier... I would view these more as different "lines" than as a progression (other than the PC line). All of these terminal-based mainframes performed very different business needs (can't think of it like we do today, where a server (hardware and OS) rarely has a real dedicated purpose). These systems were just that... systems... you didn't load 3rd party software. Everything was integrated the second the customer flipped the switch.

I wouldn't call the wiki complete (it never will be), but the content seems very accurate (guys to an amazing job about not making up shit).

User avatar
seebart
Offtopicthority Instigator

11 Aug 2016, 23:26

Could not agree more, it certainly was not the same PC market with all the standards we have today where you can build your own rig any way you want it. The customer had his upgrade options that were quite limited and quite expensive.

http://www.oldcomputers.net/ibm5150.html

Arakula

12 Aug 2016, 07:33

Nice video. Takes me back to the days when I worked on these shiny, new and wonderful machines.

Oh, and if you want "dates of appearance":
http://www-03.ibm.com/ibm/history/exhib ... years.html

User avatar
Techno Trousers
100,000,000 actuations

12 Aug 2016, 08:05

That's a great info page, Arakula, thanks! Someone should update the wiki with this information. My chronology is definitely wrong, because the 3270 PC (and thus the F-122 keyboard) predates the PC-AT by almost a year. IBM was really on top of the computing world in the early '80s. They set the standards.

andrewjoy

12 Aug 2016, 10:21

I just love the rose they have placed on top of the case :P

User avatar
lot_lizard

12 Aug 2016, 12:20

Would anyone be DISSATISFIED with a White SSK PCB and a Green FEXT (full-size) PCB? I actually like the idea of them being different colors myself, but the poll hasn't been up long. We will move onto the xWhatsit and daughter board (USB connector) board color poll if we are.

Finally landed from my work boondoggle, and will process these PM's regarding parts this afternoon (thanks for the patience and interest). I will send out confirmation PM's just to make sure everyone is in synch as to when these will be shipped as some requested the parts go out earlier.

Also, the OshPark delivery of our USB connector boards appears to have arrived... So between FINALLY putting the stainless plates together, and the xWhatsit eventual daughter board, we should have some fun weekend pics.

User avatar
alh84001
v.001

12 Aug 2016, 12:52

I voted for white FSSK (although I have an original and I probably won't be getting one), and black FEXT, but green FEXT was my second option, so these are fine with me :). Just don't make them red ;)

User avatar
E TwentyNine

12 Aug 2016, 14:39

lot_lizard wrote: Would anyone be DISSATISFIED with a White SSK PCB and a Green FEXT (full-size) PCB?
Eh, kinda. I'm a traditionalist (and a little biased toward green anyway).

Come on, some of you people who voted for the Green full size, go vote for the Green SSK!

User avatar
fohat
Elder Messenger

12 Aug 2016, 14:54

E TwentyNine wrote:
I'm a traditionalist (and a little biased toward green anyway).
Yes, green is classic and neutral. But a different-colored controller might be OK.

User avatar
Techno Trousers
100,000,000 actuations

12 Aug 2016, 15:19

I'm fine with white and green for the PCBs.

User avatar
lot_lizard

12 Aug 2016, 16:36

I checked on pricing for having a couple of color options per each PCB, and we are looking at a ~1 dollar increase per. So let's do the following. Thanks for voting... I will update the poll shortly with the xWhatsit color choices if you want to check back.

SSK PCB Colors
  • White
  • Green
Full-Size PCB Colors
  • Black
  • Green
Attachments
Screen Shot 2016-08-12 at 8.58.04 AM.png
Screen Shot 2016-08-12 at 8.58.04 AM.png (142.3 KiB) Viewed 4411 times

User avatar
lot_lizard

13 Aug 2016, 01:04

If you voted Deep Purple (OshPark) or White, please re-vote... I would really only like one color for the controller to keep the costs down. I would change the poll to only two, but it would reset the counter. More votes in general (quickly)... would be appreciated

User avatar
dokyun

13 Aug 2016, 01:24

I voted purple...count me as black instead.

Can you edit the poll to allow changing your vote without resetting the results? I'd change my vote there if I could :)

User avatar
lot_lizard

13 Aug 2016, 01:34

You should been able to pick a different radio button, and hit Submit. If that option isn't there, let me know. It would be a bug I need to report.

User avatar
dokyun

13 Aug 2016, 02:21

Done. It's possible I'm merely blind.

Post Reply

Return to “Workshop”