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

lot_lizard · 14 Jul 2016, 20:40

Rimrul wrote: But you'd have bot_lizard to cheer you up.

I have one of those already... I married her a few years ago. Except this one reduces my net worth. I WASSUP to find out how much lower my balance is

Just curious... has anyone ever considered kickstarter for any projects? I didn't bother searching before posting this, but making the molds for future F barrels and flippers certainly seems like a worthy kickstarter project. I mean if they'll back launching a replica Tardis into space, why not us..

elecplus wrote: Well, this is a first! UPS did not come get the pkgs for lot_lizard. I called, and they apologized, said they would connect me to international shipping. Why international, I asked. Because they are going to MN, said the operator

That's hysterical... I think I see why they also missed the pickup

lot_lizard · 15 Jul 2016, 15:35

It dawned on me that I never posted renders of the final version that I am currently having produced. Again, the improvements were:

Added tension screws
Added cutouts for the extrusions on the underside of the existing case's top shell. When we move to a thicker top plate, this would have added more pressure on the top shell than I think is desirable
Increased the size of the "eyelets" in the top corners of the backplate. The fit was good before, but I would prefer it was floating since we are now using the front of the assembly to hold it in the case. This is a much more sound approach than the previous model M assembly (why those tabs that pass through the eyelets would easily break in shipping). We are actually a considerably better fit in the case than the legacy assembly.
Rounded all of the corners off
Slight correction of the function row alignment
Rotated the barrel lock for the AT/F122 barrel
Left spacebar stabilizer barrel moved in (to the right) one key so that someone could take advantage of a split spacebar design (if that ever becomes available)

If pictures of the model M case parts I am referring to would make things more clear, let me know

: Top plate final; topPlateTensionTop.png (21.63 KiB) Viewed 6624 times

: Top plate with barrels removed so it is easier to see the changes; topPlateTensionScrews.png (8.46 KiB) Viewed 6624 times

: Bottom Plate with optional tension screws added (labels for giggles); bottomPlateTensionScrews.png (112.87 KiB) Viewed 6608 times

EDIT: I should have noted. These changes have been made to the full size (FEXT / 101-key), but I still need to assemble the previous one I had made to test everything before having the improved version produced.

lot_lizard · 15 Jul 2016, 17:45

Just chatted with the electroplater about our yellow zinc hot dips of the backplates (Chromate). The prices are very reasonable, and the turn around would be quick. I do need to know what kind of surface we would like to achieve. Options are:

Raw metal like this Unsaver
Polished like the Kishsaver
Sandblasted (think matte with less iridescence)
Brushed (bunch of tiny horizontal lines)
Engine turned (what I might actually do to mine regardless)

8_2014121518562619080_71422262100099_1422115878.jpg (137.28 KiB) Viewed 6604 times

As you can see... lots of options. We will go with one approach. Engine turning is about the only one that is cost prohibitive, but I'll check around. The others are cheap (polished being the most of cheap options). Would it make more sense to create a poll for this?

EDIT: I should also say, this is only if you chose the Chromate option later. The other colors would be available for the backplate as powdercoating

Hypersphere · 16 Jul 2016, 01:14

I like raw metal. The Spirit of St. Louis look is also great, but probably cost-prohibitive, as you say.

fohat · 16 Jul 2016, 01:48

The Spirit texture is awesome but I would rather save it for the metal top case shell that can actually be seen.

I would request a 2nd row of screw holes through the middle of the alphas or I may end up drilling them myself, somehow, if possible.

Otherwise I am indifferent to the insides. An M doesn't make a statement when "naked" while an F certainly does.

lot_lizard · 16 Jul 2016, 02:38

Hypersphere wrote: I like raw metal. The Spirit of St. Louis look is also great, but probably cost-prohibitive, as you say.

Can I ask what would be "worth it" for that look? I realize that the answer "depends on the total cost", but less say hypothetically we are $150 for the assembly (minus switches), but likely less. Is it worth 5 bucks?

fohat wrote: I would request a 2nd row of screw holes through the middle of the alphas or I may end up drilling them myself, somehow, if possible.

Unfortunately the barrels of the F are touching throughout the middle of the board. So unless you drill through the corners of 4 barrels, we really don't have an option. I have put together a design for phase 2 (when we make our own stuff from scratch), that has quarter circles in all four corners of every barrel. That really gives us loads of options, but that is for a while from now (not phase 1), I will go back and triple check though. I would like to leave the PCB untouched as much as possible so it works well for the legacy M bolt-mod style, but we MIGHT be able to slip something else in around the perimeter (spacebar area, etc)

I will say though, we are bending the top plate more than the originals. This means we have more down pressure in the middle of the boards when we assemble than any other F. We can get away with this approach because we can apply lots of pressure without having to "slide" the plates and lock those legacy tabs. I appreciate their original approach, but human beings weren't meant to do that by hand. They (no question) had an assembly process with presses

fohat · 16 Jul 2016, 02:52

lot_lizard wrote:
Can I ask what would be "worth it" for that look? I realize that the answer "depends on the total cost", but less say hypothetically we are $150 for the assembly (minus switches), but likely less. Is it worth 5 bucks?

fohat wrote: I would request a 2nd row of screw holes through the middle of the alphas
I will say though, we are bending the top plate more than the originals. This means we have more down pressure in the middle of the boards when we assemble than any other F. We can get away with this approach because we can apply lots of pressure without having to "slide" the plates and lock those legacy tabs. I appreciate their original approach, but human beings were meant to do that by hand.

Personally, I would not pay even $5 for a fancy surface on something that is never seen, but would gladly pay that for the outer case.

I had not picked up on the extra bend and pressure. That is very heartening and encouraging.

I'm sure that you meant: "human beings were not meant to do that by hand" didn't you?

lot_lizard · 16 Jul 2016, 02:58

fohat wrote: I'm sure that you meant: "human beings were not meant to do that by hand" didn't you?

Indeed... Thank you for catching. That would have been confusing for sure (I blame Siri and her autocorrect fluff

). Corrected the previous post

Techno Trousers · 16 Jul 2016, 05:57

Any of the finishing options on the bottom plate is fine with me. As Fohat said, we won't be looking at it too much. After I put them together, it'll most likely be something only the person who ends up inheriting my keyboards will see!

Now, for a metal case in phase 2, that engine turned finish would be 100% wicked! How durable would that be?

lot_lizard · 16 Jul 2016, 12:48

Techno Trousers wrote: Now, for a metal case in phase 2, that engine turned finish would be 100% wicked! How durable would that be?

Aren't you a mind reader

... It would be as durable as any. Just like with different grain densities of sandpaper (this uses a rouge instead), you have options with this finishing approach. Think of it more as a polishing vs a cutting technique though. It is very smooth to the touch.

Nice avatar btw

Techno Trousers · 16 Jul 2016, 16:33

lot_lizard wrote: Nice avatar btw

Thanks! I was about time I chose one.

lot_lizard · 16 Jul 2016, 18:02

Back to business...

I have been giving producing the barrels with the integrated stabilizers a lot of thought, and it just isn't cost and time effective until we are ready to make molds for several parts at once (we got lucky for phase 1 with the Cindy F switch haul). I could make them as a hobby style project, but I am going to have enough to tackle, and dont want that impacting our schedule.

But... we still can implement a high quality alternative for cheap that has flexibility (easy to paint or replace if somehow later damaged without wrecking the surface of the top plate). Plus, you should be able to add these to existing other model F's with minimal effort if that is desirable. We will have metal ones produced (see the renders below) that will be tapped, and will be attached from the bottom of the top plate with countersunk M2 screws (like the rest of the board). I am not a fan of metal on metal contact, so I will have them coated with friction-resistant surface (like ceramic) for cheap. We can also produce a version with minimal additional cost that would properly secure the thinner F wire if we like (especially interesting if you are using to repair an old F. We will use this identical part for both the spacebar and vertical stabilizers on the full-size keypad.

This will raise the height the stabilizer bar is held by 1mm, but I have tested, and it has zero impact on operation. We will have the front edge beveled to keep the wire from getting caught during cap installation.

: Ceramic coated stabilizer clip tapped for two M2 screws (original version that was improved upon... see next image); stabilizerClip.png (7.02 KiB) Viewed 6506 times

: Version 2 that has a deeper top clip, and slightly tapered on edge; stabilizerClip.png (3.73 KiB) Viewed 6461 times

: Clips installed in a location that would still support a split spacebar later if that becomes available; topPlateStabilizerClips.png (29.92 KiB) Viewed 6506 times

: Top plate with 4 additional holes to secure the clips; topPlateStabilizerClipsTop.png (14.34 KiB) Viewed 6500 times

: Countersunk M2 screws from underneath; topPlateStabilizerClipsBottom.png (14.89 KiB) Viewed 6500 times

fohat · 16 Jul 2016, 18:12

lot_lizard wrote:
we still can implement a high quality alternative for cheap that has flexibility (easy to paint or replace if somehow later damaged without wrecking the surface of the top plate). Plus, you should be able to add these to existing other model F's with minimal effort if that is desirable.

just assume there are 4 additional tiny holes that are countersunk on the bottom

This is just great.

I have been bending sheet metal (hammered out from bean cans) into "Z" brackets and epoxying them down. That actually works great, and they can be "tuned" by bending them, but looks butt-ugly.

My recommendation is that the top bar be as long as possible, and beveled to guide the wire into place.

pyrelink · 16 Jul 2016, 18:34

Wow, fantastic. The ability to implement on other Model F's is great. I will need a bunch of extras on those!

Techno Trousers · 16 Jul 2016, 19:43

Wow, that's an awesome idea for the stabilizers! I'm thinking these could totally replace the specialized barrels, since then we wouldn't have to produce a small quantity of those at a higher cost, and perhaps even two versions for M and F sized wires. But I'm a little torn, since those stabilizer barrels are such a super cool invention.

I was cleaning off and inventorying key caps yesterday, and I noticed that there is a fifth horizontal stabilizer point on the 101 Ms: The 0 key on the keypad. My plan is to use the number pad keys from an F-122 for my FEXT, since they look a lot cleaner (numbers only, no secondary legends--isn't this how the number pad always should have been, since there are dedicated arrow and other keys? But I digress). To fully accomplish this, I'd have to use a model F wire stabilizer for the 0 (or resign myself to using a 0 key with a secondary legend). I love that these screw-in metal stabilizers could be made for both M and F sized wires!

lot_lizard · 16 Jul 2016, 19:45

fohat wrote: I have been bending sheet metal (hammered out from bean cans) into "Z" brackets and epoxying them down. That actually works great, and they can be "tuned" by bending them, but looks butt-ugly.

My recommendation is that the top bar be as long as possible, and beveled to guide the wire into place.

Actually, it's good to know someone has some experience in this space. I have been lucky, and never had a stabilizer clip issue on any of my M's or F's. So I just made this have the stock M dimensions for clip (aside from raising the deck 1mm to support the tapped screw). I have often wondered if it would be more ideal to have the clips be a bit deeper, and assumed they kept them as short as functionally possible to make them less likely to crack.

The stock depth is ~4mm. Looks like we have ~2mm we could increase before you have installation issue with barrel. Splitting the difference with a 1mm increase probably makes sense. I verified, and it would not present issues when pulling the key out either (don't want it to get caught). The following render includes the improvements. I'll edit the previous post with a before and after so this dialogue is not confusing later.

: stabilizerClip.png (3.73 KiB) Viewed 6491 times

pyrelink wrote: The ability to implement on other Model F's is great. I will need a bunch of extras on those!

Sounds like having an F and M version for wire thickness is going to make a lot of sense, or at minimum a way to adapt the M clip to fit the F wire better

Techno Trousers · 16 Jul 2016, 19:55

Okay, I dismantled one of my keyboards with a non-detachable cable, and found something very interesting. At the point where the cable exits the case, IBM (or Lexmark) produced a specialized part that sits on the pegs that hold the controller PCB in the SDL cable versions. The opening on this part is an arch 3/8" wide and high (not an exact measurement, just measured with a tape). The integrated cables have a piece molded onto them that is restrained by this part. My thought was that perhaps Unicomp still sells this part, and we might be able to make use of it in the project. I opened a ticket with them to ask about that. I was picturing a combination of this part and a USB connector that fills the remaining opening just right, for a really clean and integrated look.

But as I'm typing this, I realized that we're probably shooting for the old "SDL" way of doing things, so the PCB will sit on those pegs? D'oh.

An album showing this part and how it works.

lot_lizard · 16 Jul 2016, 21:04

Techno Trousers wrote: My thought was that perhaps Unicomp still sells this part, and we might be able to make use of it in the project. I opened a ticket with them to ask about that. I was picturing a combination of this part and a USB connector that fills the remaining opening just right, for a really clean and integrated look.

But as I'm typing this, I realized that we're probably shooting for the old "SDL" way of doing things, so the PCB will sit on those pegs? D'oh.

Actually, that is exceptionally interesting. Not because I think it would work for us without a fairly significant amount of cutting/trimming, both for the USB outlet, but also to grind the bottom so that it could sit on the installed xWhatsit controller. I think we could optionally have some things printed modeled after it though (cheap low-def prints) that not only cover the hole, but hold the controller dramatically better. When we use something smaller than SDL, the controller will move vertically on those pegs when attaching/detaching the USB cable. Not a big deal by any stretch, but not ideal. This looks really easy to mock up, and should be < 5 bucks to print and paint in bulk. We could even paint in both standard and industrial case colors.

You were going to send me a few things. Do you mind throwing this in the box as well. I could measure for it, but would save time to just measure the completed part (vs. the inverse and account for tolerances). Thanks for posting the pics.

Techno Trousers · 16 Jul 2016, 22:03

lot_lizard wrote: You were going to send me a few things. Do you mind throwing this in the box as well. I could measure for it, but would save time to just measure the completed part (vs. the inverse and account for tolerances). Thanks for posting the pics.

Absolutely! It'd be great to have something really solid taking up that extra space. I think that might even help take some pressure off the USB jack when it's being connected/detached, or torqued when the keyboard or cable is moved around.

E TwentyNine · 16 Jul 2016, 22:05

I like the raw best also.

lot_lizard · 17 Jul 2016, 08:53

The "raws" have it... I'll take a couple of sample sheets to the dipper one day to see how they'll turn out.

For the tension screws that were added, I think I would like to experiment with o-rings. I mentioned before that Phosphorglow recently used in a bolt-modded airlines board of mine, and the feel was substantially better than without. Stands to reason it might have benefit here as well. The edge screws won't have a need since we will already be sandwiching rubber silicone (or the like) with the two layer foam process between the sheets. The tension screws however, are another animal out in the middle of the board.

: image.jpeg (480.8 KiB) Viewed 6475 times

lot_lizard · 17 Jul 2016, 20:40

Went back and double checked if any additional tension screws could be added, and the only spots would be above the function row. Originally, I didn't think these would be necessary, but since the tone of F5-8 was inconsistent with the rest of the board, I will add the ones at the top left corner of F5 and F9. Then we should be golden (remember, the top plate will be 33% thicker as well).

There were discussions early on about whether the full size 101-key PLASTIC barrel frame of a legacy M could be cut down to allow use of the FSSK PCB. This is proof that it would work (pics below of the FSSK PCB stacked on top of the FEXT PCB). It has no real impact on this project at this point since we are making our own metal plates (other than the tension screw locations are easier to map), but thought that it should be noted.

: image.jpeg (852.87 KiB) Viewed 6430 times

: image.jpeg (1.17 MiB) Viewed 6430 times

lot_lizard · 18 Jul 2016, 16:20

I'm stopping by the machine shop today before I head out of town for the week. They will be producing both the "final" FSSK and FEXT in stainless 18 gauge using our latest plans. The only real change to the FSSK is the addition of several extra tension screws where possible (without modifying the current PCB). I have added one extra tension screw that i$ will need to account for because it will be necessary. The problem with these tension screws, if they are used, other parts of the board that were tuned will now be out of tune (we are shifting down pressure)... so we need to make sure we can be as uniform throughout the board with there placement as possible. The new hole will not impact any of the existing PCB routes.

The following are also renders of the FEXT. All the improvements made to the FSSK have been accounted for the FEXT design as well. The only remaining piece will be the drilled taps for the vertical stabilizer clips (which I can add by hand later... no need to wait for). I will have to dig a board out of storage to re-evaluate where they should be placed.

: Full Size (FEXT) top plate (note: still need to add the vertical stabilizer clip tapped holes); topPlateTop.png (15.23 KiB) Viewed 6388 times

: Full Size (FEXT) backplate with tension screws added; bottomPlate.png (60.62 KiB) Viewed 6388 times

: Revised Space Saver (FSSK) with new tension screw pattern; bottomPlateTensionScrews.png (61.95 KiB) Viewed 6388 times

Spaceman1200 · 18 Jul 2016, 19:15

Nice work as usual lot_lizard

Techno Trousers · 18 Jul 2016, 19:56

For the stabilizers, are you going to add tapped holes for the other horizontal locations too? They would need to be there for:

(Both)
Left shift
Right shift
Backspace
ANSI enter
ISO enter?

FEXT
Numpad 0

I'll take a couple of quick pictures of my F122 wire stab locations shortly. If someone has pictures of a Model M plate for ANSI enter wire stab locations, that would be helpful.

Rimrul · 18 Jul 2016, 20:11

ISO enter is a vertical stab. The other two vertical stabs are Numpad + and Numpad enter. ISO enter is 1.25u wide, though.

Techno Trousers · 18 Jul 2016, 20:14

Okay, here are pictures of the horizontal and a couple of extra vertical stabilizer points I think we want, in order to be compatible with every type of cap IBM has produced for the AT keyboard and onward. I apologize for the potatoness of these photos,and the condition of my F-122 plate.

Here's the big cluster: Horizontal stabs for backspace, ANSI enter, and right shift. Note the vertical stab points, to allow for use of the F-122 style ISO enter. I don't need that key, but someone else might want it.

Spoiler:

Left shift

Spoiler:

Numpad 0

Spoiler:

Techno Trousers · 18 Jul 2016, 20:24

One other thing I should have asked much earlier, sorry: Is there any way for the screw-in stabilizers to attach from the top, rather than underneath? I suspect that there would be an issue with the size of the screw head and countersinking, but I thought I'd ask. It'd be a lot easier to reconfigure for F versus M wire stabilizers if they could be attached and detached from the top. Realistically, I think it'd be fairly rare where someone would want to do this, but I could foresee it happening, especially with folks that like to change between cap sets.

Rimrul · 18 Jul 2016, 20:30

Techno Trousers wrote: Here's the big cluster. Note the vertical AND horizontal stab points, to allow for use of the F-122 style ISO enter. I don't need that key, but someone else might want it.

Spoiler:
Stabs1.jpg

Arent those horizontals backspace, ANSI enter and right shift?

Techno Trousers · 18 Jul 2016, 20:32

Rimrul wrote: Arent those horizontals backspace, ANSI enter and right shift?

Yes. I edited the original post to make my description more clear