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

Khers · 20 Aug 2016, 21:57

LEDs? Eeeew!

ramnes · 20 Aug 2016, 22:44

lot_lizard wrote: ↑Okay... I'm FINALLY back in town from across the pond. I tried uploading these a couple of days ago, but my reception was terrible.

You are about to be bombarded by pictures. This would be the latest stainless steel version of the SSK. I have a couple of things to make sure I am in LOVE with, but it really is just about there (would be passable as-is).

This has become my favorite board all over again (mostly a 1395682 with OG APL caps mixed in choice spots). I didn't break out the buffing wheels and rouge, but this will easily look like mirrored chrome with 10-15 minutes of extra effort. I think in the end, it is a mirrored top plate with "engine turned" back.

[Lots of awesome pics]

Does it ping?

Anyway, ping or not, this is absolutely marvelous. How do I get one?

Techno Trousers · 21 Aug 2016, 01:50

ramnes wrote: Anyway, ping or not, this is absolutely marvelous. How do I get one?

First, you'll need a Model M for the keys and case (101 or SSK), or buy an old school case and keys from Unicomp. If you want to guarantee that you can build one you should also get yourself one or more Model Fs to pull the barrels and flippers from. lot_lizard had salvaged a bunch of F keyboards thanks to Cindy, but I suspect that demand will end up outstripping that supply. All the rest of the parts you'll need will be purchasable in the group buy. I think that'll kick off in around 4-6 weeks. That's just my best best guess, and could be optimistic.

Techno Trousers · 21 Aug 2016, 02:20

I should mention that one other option is to do it "Memento" style. Buy all the parts in phase 1, then wait for new barrels, flippers, and metal case that are going to be produced in phase 2. I'm going to do that for one TKL MF.

lot_lizard · 21 Aug 2016, 04:16

ramnes wrote: ↑Does it ping?

This is the MILLION dollar question in my eyes, and what I would have been asking as well from the beginning. To me, and I own/experimented with multiples of each, the order of F switch "pedigree" is the following from worst to best (comparing only switch tone, responsiveness, tactile feedback, and reverberation... not the boards themselves). Keep in mind they all use identical switches, so other factors in the assembly are impacting

3178 (Blue Switch)
F107 (Unsaver)/F122
AT
XT (ignore the spacebar)
4704 (Kishsaver and Battleship)
Bigfoot (ignore the spacebar)

Without any tensioning or foam considerations, we are between the F122 and XT. Everything is exceptionally consistent, but there is a slight "ring" when compared to the upper end of the spectrum. After adding the tension screws, we are better than the XT. Also... this is all subjective opinion... Others will feel differently.

I am convinced what hurts the F107/122 is the height (y-axis.. number of rows) of the board. We have significantly improved that impact, but we are essentially a 7 row board if you consider the "dead row " between the numerics and functions. We have improved the design, and are close to the 4704 before foam improvements (after tensioning), but not sure if we ever reach the Bigfoot. It is just a special board for some reason that eludes me. Why a Stradivarius can't be reproduced?!? I am working hard at it, and think we can best the 4704 tone in the end after foam experiments... But I am not sure we ever hit the Bigfoot (crisp feel and "warm").

Hopefully this helps. We are definitely an incredible F at moment, and we WILL improve with foam... but I doubt we reach Bigfoot even though that is the goal.

lot_lizard · 21 Aug 2016, 05:22

Khers wrote: ↑LEDs? Eeeew!

I'm not sure what to think myself without seeing it. The idea would be to make a cutout similar to the following on Capslock (for every switch), and let someone dream up a LED membrane in the future for every key. I'm not convinced how interesting this is given that PBT keys obviously wouldn't be transparent to the legends, but it has MINIMAL impact to add the capability to the plate. If someone constructs a LED membrane layer that lives between the top plate and foam (with integrated LEDs), then we would be off and running.

fohat · 21 Aug 2016, 15:38

lot_lizard wrote: ↑
I am convinced what hurts the F107/122 is the height (y-axis.. number of rows) of the board.

but we are essentially a 7 row board if you consider the "dead row " between the numerics and functions.

Personally, the noise of a raw Model F is too much for me, and that is why I always do a floss mod. It does not silence the keyboard, but it cuts the ringing ("singing" it has also been called) overtones while leaving the underlying sound and feel unchanged.

As for that wide expanse of "dead area", during my refurb I always do my "Model F bolt-mod" (which is not like the Model M version where the bolts take the place of something, but rather just a bit of added pull at just the right places) where I use all the original tabs as intended and add 3 bolts: between 5 and F5, between 0 and F10, and just right of Enter (diagonally above Left Arrow).

This tightens everything up nicely and is quite easy to do. I use 4-40 x1/2" or M3 x8mm machine screws and nuts with washers, sometimes rubber washers although I don't know if that really matters.

Hypersphere · 21 Aug 2016, 15:44

Regarding "ping", my XT sounds and feels so great I almost don't mind the ping. Nevertheless, I went ahead and did the "floss mod", and I found that it really works to remove the ping without sacrificing the other good aspects of the XT's feel and sound. So, if the FSSK has ping, people could always try the floss mod, which is completely reversible.

About LEDs, I am not sure I like this idea. How certain is it that the cutouts would have minimal impact on the sound and feel of the board? OTOH, if there is to be no lock-light panel in the FSSK as there is in the full-size M, it might be nice to have LEDs just for Caps, Num, and Scroll.

E TwentyNine · 21 Aug 2016, 16:21

fohat wrote: ↑ As for that wide expanse of "dead area", during my refurb I always do my "Model F bolt-mod" (which is not like the Model M version where the bolts take the place of something, but rather just a bit of added pull at just the right places) where I use all the original tabs as intended and add 3 bolts: between 5 and F5, between 0 and F10, and just right of Enter (diagonally above Left Arrow).

This tightens everything up nicely and is quite easy to do. I use 4-40 x1/2" or M3 x8mm machine screws and nuts with washers, sometimes rubber washers although I don't know if that really matters.

I wish someone would do an analysis of the XT vs AT keyboard assembly "sandwich", measuring the pieces disassembled (including curve of the plates) and assembled.

The XT assembly is noticeably tighter and I wonder what exactly causes that given that their construction on the surface is virtually identical. Different curves in front/back plates? Less of a gap, more compression? What?

Muirium · 21 Aug 2016, 16:28

Did I miss all this? Anyone got a three level deep TL;DR? My SSKs need this badly. But everything is up at top priority in the time/expense list this year, so I can't help with any development / unfinished product duties, alas.

chzel · 21 Aug 2016, 16:36

No, you didn't miss it, but it's not too far away!
Stainless sandwich with XT/AT barrels and flippers, i$'s PCB and redesigned form factor for xwhatsit's controller by wcass.
Both SSK and full size, Cindy hooked us up with a ton of F122's for barrels/flippers.
Open your existing M, remove guts, replace with this, move caps and voila! IBM Model MotherFucker!

Spoiler:

Price projected at about 150$ or less.

lot_lizard · 21 Aug 2016, 16:53

Inspired by Wodan and his research in foam cutting techniques: workshop-f7/ibm-beamspring-model-f-foam ... 14403.html

I am going to pick up a 100W laser etcher/cutter to produce our "foam" sheets myself. One with a larger working area would be ideal so we can produce a full sheet at a time without intervention. I want to experiment quite a bit with various materials and thicknesses, and our volumes aren't really high enough to necessitate a die. For the price of the die, we can go a long way towards the laser cutter. I would appreciate having one anyway, and would absorb the cost difference.

There has already been loads of research about this topic, but I think we can extend it if we had the right equipment in-house. Hypersphere had a nice reference post attached below to see the previous experiments of others.

Hypersphere wrote: ↑As you perhaps know, there have been a number of threads on these topics. In addition to whatever comes up here, you might also wish to consult the other threads. Here are a few examples. Note that the last one is about beam springs.

https://geekhack.org/index.php?topic=58814.0

workshop-f7/replacement-foam-in-ibm-model-f-t11657.html

keyboards-f2/replacement-foam-mat-for-i ... t8033.html

photos-f62/ibm-beamspring-3278-t8818.html

I still think true sound deadening materials might be our best bet, and want to give them a go. With high enough power laser cutter, we could even cut sheets from substances like Dynamat (butyl rubber) equivalents that are backed without foil, or even etch the metal ourselves (Limited Edition logo, etc) with products like Cermark

: image.jpeg (36.38 KiB) Viewed 5136 times

lot_lizard · 21 Aug 2016, 17:12

Muirium wrote: ↑Did I miss all this? Anyone got a three level deep TL;DR? My SSKs need this badly. But everything is up at top priority in the time/expense list this year, so I can't help with any development / unfinished product duties, alas.

Welcome back my friend... You had no worries. I would have saved you a couple knowing you would have eventually returned

Hypersphere wrote: ↑About LEDs, I am not sure I like this idea. How certain is it that the cutouts would have minimal impact on the sound and feel of the board? OTOH, if there is to be no lock-light panel in the FSSK as there is in the full-size M, it might be nice to have LEDs just for Caps, Num, and Scroll.

As you say, at minimum we do it for the lock keys. As far as structural and sound differences, in theory it would minimal. This LED membrane would live above the foam layer, and the foam wouldn't have the cutouts for it. Obviously lacking the metal would have impact, but these LED notches would be pretty small. Your hesitations are also mine though. We are going through all this effort to produce the best sound possible only to give up some of that for something that might never even get carried out. After Phosphorglow's feedback, we can vote

E TwentyNine · 21 Aug 2016, 17:18

Regarding MF, I'm kinda partial to calling the smaller version the MMF (Modern/Mini Motherfucking F) to parallel the SSK nomenclature.

wcass · 21 Aug 2016, 18:00

Re LEDs ... I've used light pipes to very nice effect on my custom model M. The light pipe takes the place of any bolt from a bolt mod. This build is not using most of the bolt holes, but we could add one one or more where space allows.

: 8under cap.JPG (97.43 KiB) Viewed 5096 times

: back.JPG (128.08 KiB) Viewed 5096 times

lot_lizard · 21 Aug 2016, 18:14

@WCass... What do you think about a membrane style circuit that sits between the foam and top plate that has integrated LEDs? We would have a cutout in front of each barrel that the LEDs rise up into. Then all the "wiring" is contained in the assembly, and could be installed as a single sheet across the entire board at once. Crazy talk, or reasonable? We would still vote for even wanting it in the end, but looking for plausibility and a guess at the manufacturing cost

fohat · 21 Aug 2016, 18:27

I would never use backlighting myself, but can you get it thin enough?

I have had trouble with "Z" shaped sheet metal brackets to convert Fs to use M spacebar wires - even the slightest bit too tall and they interfere with the downstroke of the key cap.

chzel · 21 Aug 2016, 18:34

Maybe the membrane with LEDs could sit on top of the barrel assembly? Like phosphorglow's coloured sheets? Easily removable and nothing to change regarding the actual sandwich.

lot_lizard · 21 Aug 2016, 19:17

chzel wrote: ↑Maybe the membrane with LEDs could sit on top of the barrel assembly? Like phosphorglow's coloured sheets? Easily removable and nothing to change regarding the actual sandwich.

That's actually what Phosphorglow is proposing (which I like also). I'm just exhausting the other approach because of the difficulty of wishing we had it integrated later. If we did like it, it would take research to make sure it had minimal negative effects (to fohat's point)

Khers · 21 Aug 2016, 21:10

lot_lizard wrote: ↑
chzel wrote: ↑Maybe the membrane with LEDs could sit on top of the barrel assembly? Like phosphorglow's coloured sheets? Easily removable and nothing to change regarding the actual sandwich.
That's actually what Phosphorglow is proposing (which I like also). I'm just exhausting the other approach because of the difficulty of wishing we had it integrated later. If we did like it, it would take research to make sure it had minimal negative effects (to fohat's point)

I like this approach as well. Those who want LEDs sacrifice very little using this approach and it risks nothing for those of us who have no interest in LEDs. (One of the reasons I like the SSK is its complete lack of ugly LEDs, but maybe that's just me?)

lot_lizard · 22 Aug 2016, 04:05

The laser cutter/engraver has been ordered... Fun stuff. The first experiment, which I think has the biggest promise, will be butyl rubber. It is the same compound used in Dynamat. I will be ordering some polymer backed version on both sides to allow it to be laser cut (aluminum basically had no chance with a hobby cutter), and backing will prevent it from being a sticky mess if needing to be removed later.

The downside to butyl rubber is that it suffers from "compression set". Meaning it doesn't "bounce back" very well once compressed, and remains permanently deformed. Most foams (even the neoprene we have preferred lately) suffer a similar issue if substantially compressed. Ideally, we will tension just enough for assembly, assemble, and tension while testing the results.

The MASSIVE upside to butyl rubber is the resilience is VERY low. So dampening and vibration resistance is the inverse (VERY high).

After butyl rubber testing is a long shelf life "latex" alternative. Ultimately, we are building a multilayer foam alternative that should surpass our previous attempts at replacing foam (at least on paper). I'm anxious to see the results

Techno Trousers · 22 Aug 2016, 05:25

Great news, lizard! I think you're right that it didn't make sense to have a cutting die made for a one-off project like this. Are you going to be willing and able to laser cut 200+ sheets though, assuming 2 per keyboard? How long does one sheet take to do?

Wodan · 22 Aug 2016, 07:34

Which laser are you getting?

Let me know if you need some help getting started with a cheapo China laser

And should you order acrylic, make sure it's real acrylic and not polycarbonate!

lot_lizard · 22 Aug 2016, 07:56

Techno Trousers wrote: ↑Are you going to be willing and able to laser cut 200+ sheets though, assuming 2 per keyboard? How long does one sheet take to do?

SEVERAL minutes each!!!... I'll have several weeks to get them done though (can start before the GB even commences), and the cutting bed is big enough for multiple sheets. So "click- click", and vanish for a while. 3 sets a day would be very comfortable both from timing and effort. Just need to get the design hardened sooner rather than later. Have something in mind that I think will be very promising, but need to prove/tweak.

The only real problem would be that I think it works, start producing, and then I experience a problem from it late into real world testing while everything else is underway. The chances of that are slim, but always possible. Neoprene foam would be the fallback plan (which the machine could fly through if you crank up the watts and get the speed just right... In theory)

lot_lizard · 22 Aug 2016, 08:23

Wodan wrote: ↑Which laser are you getting?

Let me know if you need some help getting started with a cheapo China laser

And should you order acrylic, make sure it's real acrylic and not polycarbonate!

Something similar to the one below. I'm having one of my manufacturing friends order it through his supplier since he gets a bit of a cut on cost, and it would come with an ACTUAL warranty.

I appreciate the tips and the offer. I strongly suspect I will be reaching out to you. The work you have already done in this space looks excellent, and one of the reasons I'm attempting... My wife hates you btw

: image.jpeg (60.35 KiB) Viewed 5007 times

Wodan · 22 Aug 2016, 08:41

My wifes hates me as well! They should meet ...

Damn that thing looks serious. It's a laser on wheels ... I'm pb & jelly.

Well I'm also just a hobby laser guy so I can mostly just share all the stuff that went wrong for me. Cutting these foamed rubbers is a very forgiving job but cutting keyboard plates is a birch! 2mm across a 60% plate and you can scrap it ...

I didn't know what KERF is for the longest time ... I still don't know the right value for my laser but this is something I have to deal with before I laser any more plates lol.

The ultra-cheapo I got also has Chinas worst ventilation/filter system so I usually just attach a very old, crapps vacuum cleaner to the "outside" end of the exhaust pipe and just suck out the fumes from there ...

I think about 5 minutes into this game, I will start learning from you

Khers · 22 Aug 2016, 09:22

Wow! You guys are on a whole different level! I wish I had the space for stuff like that, but our house is too damn small, alas. I'll start bringing you up as examples of what could have been when my wife complains

lot_lizard · 22 Aug 2016, 13:13

Wodan wrote: ↑I think about 5 minutes into this game, I will start learning from you kh

Oh $HIT... This is suddenly feeling like a TERRIBLE idea!!!

Khers wrote: ↑I'll start bringing you up as examples of what could have been when my wife complains

You're welcome... It's our flavor of "trickle down economics"... One of us makes a horrid drunk decision, and all others benefit. Everyone should have their version of a "Wodan Wild Card" they can play when the odds aren't in their favor.

idollar · 22 Aug 2016, 14:41

Hello guys ....

You may have noticed that I have been very busy. I was not able to participate in this project since a while.
I keep typing with my FSSK in the office without any problem, and with each stroke, I remembered that I have to find some time for dt.

Anyhow ... I am back, hopefully with some additional time to help in this project.

I have been reading the posts since I last did, but it is complex to follow. Perhaps a summary of the status in the initial page would help. I am not asking, as it would not be fair to do so. It is just a side comment.

How can I contribute ? Is there anything that I can do ?

i$

lot_lizard · 22 Aug 2016, 16:15

idollar wrote: ↑How can I contribute ? Is there anything that I can do ?

Welcome back friend... I hope the break treated you well.

I did bring the laptop with, and will go clean up the OP's. I keep saying that, but mean it this time to help myself and everyone else regroup.

The plates and mechanical aspects are in a really good place. We still need to fully test the FEXT plates, but the initial rough fit looks great. WCass has done some great work for us on the mini xWhatsit front. His initial designs are completed, but need to be tested and receive final updates. WCass has also agreed to modify the PCB connections to allow our new xwhatsit to slip directly onto the PCB (like the beamspring, only inverted to lie behind the backplate). If you scroll back a couple of pages, you will see the "final" SSK plates in stainless steel where I attached the original xwhatsit directly to the PCB with header pins to simulate. Attached is a picture from that series of images

The only other changes we need to the PCB are 2/3 extra pads for an optional split spacebar, and an extra hole or how to help with tension. I would reach out directly to WCass to see where he is at with updates, and what help he needs.

Again... Welcome back to the party.