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

User avatar
lot_lizard

16 Jun 2016, 13:27

tentator wrote: About moving connector in between f row and numbers row I see that could be nice idea (you then assume to solder the controller on the back directly piggy backed right?) Because it would even allow hackers and keyboard choppers to maybe cut away the whole f row ( you know to create a 60% model f ;))
Good point, along with being more compact, it is more" flexible configuration" friendly. But more importantly, it provides a fancy way to hide the controller (assuming we aren't integrated into the PCB) and connection cables into this floating base/stand. Even if the Cherry/flat plate stuff doesn't work out, I still like the idea of the artisan base with the curved plate.

I am a couple of days away from having something to show, but I like it. It's very low profile (no higher than the M), but will appear like the plates are hovering. Working out the stability issues at the moment. There is a massive upside if we like it because it eliminates the cost of a bloated case

User avatar
Hypersphere

18 Jun 2016, 16:45

I just now discovered your project (actually, I was pointed to it recently by Muirium). I did see the title of the thread earlier, but I passed it by, thinking it was only about some rearrangement of the classic full-size Model M.

Now I see that you are talking about converting the SSK to a capacitive buckling spring (Model F) version of the SSK. This is great! I am very interested.

Although over the years my preference in form factors has converged on 60%, I still like TKL, and the IBM SSK is perhaps the most beautiful rendering of the TKL.

Please add me to your list of interested folk.

Thanks for undertaking this project!

User avatar
tentator

18 Jun 2016, 22:16

You may never know if this will eventually evolve to a 60% pcb plus case and with model f flippers! In my dreams it will even have a pointing stick between g and h :))))

User avatar
Phenix
-p

18 Jun 2016, 22:29

And in my dream it has 2 pointing sticks and is split (and if possible made so that halfes can join together, like VE.A)

But it has F flippers, so it should be great anyway ;)

User avatar
DMA

19 Jun 2016, 09:22

Oh. I also dream about adding trackpoint to model F.
I even have about 8 thinkpad keyboards from T400 to T530 (long story of collecting dead thinkpad keyboards from our IT :) ) as a source of trackpoints.
It was pointless for unicomp - but I'm going to try it on XT frame sometimes and if successful - reproduce on 122key model F i recently have.
Dremel and poxipol will likely be my tools.

User avatar
tentator

19 Jun 2016, 16:52

Unicomp even sells the three scooped keycaps to leave space for the stick between them :)))

User avatar
lot_lizard

21 Jun 2016, 16:41

Status Update:

I stopped by the machine shop today to check on the plates, and I have good and bad news.
Good news - The bends (curve) turned out perfect on all four plates (2 top/2 bottom). The bottom plates fit both the SSK and full-size cases perfectly
Bad news - Somehow these yahoos managed to brake (90 degree offset bend) the top plates at the wrong location on the back. I explained the front offset brake had to be EXACTLY as I had it in the printed prototype (which they did perfectly), but then told them tolerance was available for the back offset. Somehow, they managed to fubar the easy part. I should have been more specific, but you would think a printed prototype and verbal announcement of "make it look like this" would have been enough :)

Upside, they are going to recut the top plates and process, which gives me the chance to sneak-in the winkeys and the universal cutout to support both XT and AT barrel locks (so a silver lining). So our prototypes will now be a reflection of the end result (as of now). I should know something within a week.

User avatar
Techno Trousers
100,000,000 actuations

21 Jun 2016, 17:26

Wow things are progressing fast! I need to get moving on my label assignment. At least I'm finally done with my side work project for the time being.

User avatar
lot_lizard

22 Jun 2016, 04:06

Techno Trousers wrote: I need to get moving on my label assignment. At least I'm finally done with my side work project for the time being.
Yes. In the words of Mary Poppins (4-year old daughter in the house)... "Spit-Spot". Aka... "Chop-chop"

User avatar
pyrelink

22 Jun 2016, 04:20

Forgive me if I have already read it, and forgotten it, but have you gotten back the analysis of the exact flipper material used? Very interested in that.

User avatar
lot_lizard

22 Jun 2016, 04:40

pyrelink wrote: Forgive me if I have already read it, and forgotten it, but have you gotten back the analysis of the exact flipper material used? Very interested in that.
I haven't... Thanks for reminding. I'll call the lab tomorrow to get status. I can tell there is a TON of ash in the mix, but I'm anxious as well to see how much. It is a very unique ABS mixture for sure

User avatar
lot_lizard

23 Jun 2016, 13:09

Phase 2 update... The Shapeways order for the switch silencing "improvements" has shipped, and should arrive Saturday. I ordered the parts with incremental updates so we could test the adjustments as independent variables to measure effectiveness.

andrewjoy

23 Jun 2016, 13:24

Will we be able to order a " classic" model F setup without all this silencing rubbish ?

User avatar
lot_lizard

23 Jun 2016, 13:35

andrewjoy wrote: Will we be able to order a " classic" model F setup without all this silencing rubbish ?
No idea yet what any of the ordering holds for phase 2, or even if this silencing bit is worth persuing (I appreciate the loudness myself). Eventually I suspect we make a poll if the silencing is even interesting in practice. For sure phase 1 (the drop-in) is classic only with the reuse of your existing M caps

User avatar
tentator

23 Jun 2016, 14:07

I vote for silencing!! :P

andrewjoy

23 Jun 2016, 14:16

I am a round 1 boy , so you round 2 boys can do whatever you like :).

Drop in replacement, one 101 and one SSK please , as much as i want one of them new 4704s this is just as cool

User avatar
vsev

25 Jun 2016, 16:56

In my opinion this, in many ways, is even better than news 4704s !

User avatar
Hypersphere

25 Jun 2016, 21:57

I just wanted to add my comments to those of others regarding Phase I and Phase II of the Model MF project(s).

Phase I and Phase II seem like distinct projects, although it might have been difficult to achieve Phase II without first accomplishing Phase I.

I should think there would be many who would like to see the drop-in using an original SSK case. I for one admire the design and styling of the SSK, yet I have wished that the switches could be Model F.

As for capacitive buckling spring switches (Model F or Model F-type) in a novel setting -- this is an interesting challenge and not without risk.

Despite the fact that Ellipse has promised to deliver a reincarnated Kishsaver along with a new variant in a compact case, my own preference for Phase II is for a 60% form factor with a HHKB-like layout, and switches with the weight and feel of IBM Model F.

If the native layout were not HHKB, I would want full programmability, preferably with a dead-easy gui configurator. Beyond these preferences, there would be scope for some imagination in design and styling, such as a "floating look", polished steel or aluminum instead of black, optional LEDs, and perhaps even a new switch design, such as a switch that had the force-displacement curve of the IBM Model F but without a loud click --if such a thing were possible.

Regarding the issue of silencing, I am a bit schizophrenic when it comes to keyboard preferences. When I am using a Topre-switch keyboard, I am sensitive to the slightest extraneous noise that the keyboard makes. I particularly don't like the return-stroke sound, and so I have installed Silencing Rings in all my Topre-switch boards, and I have lubed all the stabilizers and switch rails. Now just about all I hear is the downstroke "thock", which I find satisfying.

However, when I am using a keyboard whose switch mechanism is inherently noisy, such as IBM buckling springs, I expect them to be noisy. It is an honest type of noise that is intrinsic to the working of the switch. This is in stark contrast to something like Cherry mx blue switches, where the "tactile bump" has literally been grafted to the stem and the click has been artificially added as well. I object to these non-purposeful noises but not to the sounds that the switch makes in order to carry out its function.

Therefore, I would not suggest investing time, effort, and funds on trying to suppress the sound of buckling spring switches. Instead, embrace the noise! However, it could be okay to tune the sound, but this might be more a function of the case design. This is an aspect of the XT that gives it what I consider a pleasant sound -- the tight-as-a-drum case of the XT, whose tension is maintained by just two screws on the back. To my ear, the XT sounds much nicer than the AT, and this has much to do with the case design.

Tuning the sound of a keyboard shares some features with musical instruments. The sound of a violin depends mightily on the strings being held under tension and vibrating against the bridge, which transmits vibrations to the hollow wooden body. It is still a mystery why a Stradivarius sounds better than a current mass-produced violin, each in the able hands of someone like Itzhak Perlman. It is also not altogether clear how the position of the internal sound post modulates the sound, but musicians will work with a violin maker to adjust the position of the sound post to fine-tune the instrument. The point is, acoustics is still an art as well as a science.

Recently, I have been experimenting with different cases for my V60MTS-C. I thought I would dress up the board by transplanting it into a low-profile TEX CNC aluminum case. It looked quite sleek and more modern than before, but in the aluminum case, the overall sound of the keystrokes was raised in pitch (or perhaps the low notes were attenuated). The sound of the click mechanism was accentuated at the expense of the thocky sound of bottoming out the keystrokes. The new case made my Matias Click switches sound like Cherry blues. Not good to my ear! Next, I tried a Royal Glam black walnut case. Again, it looked nice, but now the sound was deadened, and because the Matias boards will only accept 3 of the 6 possible mounting screws, the keyboard wobbles when hitting keys in the upper two rows. It turns out that that the cheap stock plastic case sounds the best of the three. The Matias plastic case appears to have a larger hollow volume than some other stock cases I have seen, and I think this acts rather like the body of a violin or classical guitar to deepen and amplify the sound.

I really ought to take the time to make some recordings, plot some amplitude vs frequency graphs, and then post my findings. One point I'd like to make, which is sure to be controversial, is that the aftermarket keyboard case racket is not improving the sound of keyboards (to my ear), although some would argue that a well-made aluminum case makes the board look more modern and sophisticated. A metal case also increases the weight of the keyboard, but this doesn't matter until you pick it up to appreciate the heft.

"In any case", I am looking forward to both, the drop-in Phase I solution along with the innovations that you and others devise for Phase II.

User avatar
pyrelink

26 Jun 2016, 07:48

Specifically the Phase 1 is what I am primarily interested in, as I do love my SSK layouts, but I really do not like the feel of Model M at all. I had never really ever considered silencing a buckling spring keyboard. Never even tried a floss mod, as like you say, these have never been and were never intended to be quiet. I am seeing "Phase 2" as being sort of how far can the Capactive Buckling Spring design be taken, and I think it would be interesting to see if silencing was possible without altering feel. So far the aspects of Phase 2 that I want to focus on are, layout, Cherry keycap adapters, and case design. If we couldn't get a silent buckling spring, I wouldn't be heart broken, but still think its worth the try. The Ellipse project is fantastic, but there are a lot of reasons (including price) that has turned me off to it so I am glad this project is here.

Also I completely agree with your findings on aftermarket cases. If sound is your primary concern, they certainly suck. I still love Blue Alps for their feel, and while they don't sound *bad* Blue Alps sound significantly better in their Leading Edge case and plate than they do transplanted into my low profile acrylic TEX case. I have tried aluminum cases and stock V60 and Poker cases as well, and I have sort of given up on trying to replicate their sound when left in their stock chassis.

User avatar
lot_lizard

28 Jun 2016, 16:40

Hypersphere wrote: Tuning the sound of a keyboard shares some features with musical instruments. The sound of a violin depends mightily on the strings being held under tension and vibrating against the bridge, which transmits vibrations to the hollow wooden body. It is still a mystery why a Stradivarius sounds better than a current mass-produced violin, each in the able hands of someone like Itzhak Perlman. It is also not altogether clear how the position of the internal sound post modulates the sound, but musicians will work with a violin maker to adjust the position of the sound post to fine-tune the instrument. The point is, acoustics is still an art as well as a science.
This is beautifully worded, and I agree this is really an exercise in "tuning"... not silencing. If we can really eliminate any of the side effect noises like reverberation (not only of the spring, but surrounding parts), I think we have the opportunity to raise the "perceived" quality of the switch. I wouldn't even mind it being louder, as long as it sounded more crisp ;)

I recently got back a bolt-modded airlines M from Phosphorglow. He tried something new with my board where he die-cut neoprene foam washers for each of the bolts. I am blown away by how much better he was able to tune not only the sound, but key-feel throughout the board as compared to his earlier versions. There is definitely something there that we need to be mindful of going forward.

UPDATE: I am traveling this week, but did get a call from the machine shop that the new top plates are ready. So we should have some progress this long 4th of July weekend. Again, these will have cutouts for the winkeys, and additional 4704/AT barrel fitting. Assuming there weren't mistakes, it would be the end deliverable to date.

And before I left, I opened up my Shapeways order, and they COMPLETELY shit the bed on these prints. The tolerances were off by as much as .4mm on the keystems (making them useless), and I attached a picture of the deformed mess the barrel turned out to be. Very disappointing, but they are currently reprinting and will send out "ASAP". I have always been very pleased with their accuracy, but there was obviously something wrong with the printing and quality control
Attachments
FullSizeRender.jpg
FullSizeRender.jpg (459.68 KiB) Viewed 4234 times

User avatar
shreebles
Finally 60%

29 Jun 2016, 11:50

This thread has taught me not to stay off DT for too long, else I might miss something essential.
I need some time to process all this, but either way looks amazing so far :)
Too bad I already killed an M for a second SSK assembly :twisted:

User avatar
lot_lizard

29 Jun 2016, 20:50

The metal shop forwarded me this picture (notice the barrel fitting on the bottom row where the tolerance is tightest). I am pleased with results, and thought others might find interesting
Attachments
image.jpeg
image.jpeg (2.94 MiB) Viewed 4227 times

User avatar
alh84001
v.001

29 Jun 2016, 21:16

Looking mighty fine.

User avatar
tentator

29 Jun 2016, 22:36

I tried myself a little bit on the sound dampening and so far not so encouraging things:
- trying to glue some vinilic dot on the same place you did on the barrel the sound does not really change at all.. so probably that's not the place where to work on..
- also trying to dampen somewhere else I noticed that the main point seems to be decoupling the barrel from the plate (probably that's also why there is always some strange foam-thing on ibm bucklesprings).. but if I take it to extremes and heavily decouple the barrel then I obtain actually a similar sound of that of a model m with a lot of usage and wear (!)
so to be honest I'd admit I think I'd actually then prefer to have the characteristic sound of model f and live with it since otherwise it seems it will kind of lose it's characteristics: what do you think?
What still is effective and I personally instead continue to do is reduce/attenuate the ring with dental floss (anybody tried or had ideas on alternatives?) because while the flipper noise is loud I find it nice but the spring ring is annoying to me because of the pitch more than because of the volume per se.
photo109275380194585909.jpg
photo109275380194585909.jpg (61.55 KiB) Viewed 4191 times
tent:wq

User avatar
lot_lizard

29 Jun 2016, 23:28

Excellent... Appreciate the help. When you say the following...
tentator wrote: trying to glue some vinilic dot on the same place you did on the barrel the sound does not really change at all.. so probably that's not the place where to work on..
Is this in reference to when the flipper returns to the barrel ("couples" using your terms), or on the decouple? I would have expected a slightly muted flipper sound when the key is released and returns from the PCB (though slight). It's good to know if you are not experiencing though.

Another method would be thicken the flipper paddle itself, but then we get into weight and spring actuation concerns.
tentator wrote: so to be honest I'd admit I think I'd actually then prefer to have the characteristic sound of model f and live with it since otherwise it seems it will kind of lose it's characteristics: what do you think?
Absolutely zero objections there... Just trying a few things to see if we bumble our way into an improvement. The more people taking a stab at it the better. Again... appreciate

User avatar
tentator

29 Jun 2016, 23:46

you're welcome and actually the test with the vinilic glue is that white dot you see in the picture indeed that should attenuate the return of the flipper.. so far i could spot only a really tiny difference in sound but it might also be that the material I used and the positioning of that dot is not the best..

but the above has nothing to do with what I mentioned as a second test later on about decoupling the whole barrel (rectangle section of it at the bottom) from the pcb actually below it (and this is actually very difficult to test for me) and/or from the plate it is inserted in (there usually ibm put already some black foam that degrates badly with age.. ;))
so this above operation seems to mute the whole "click" of the switch a lot such as kind of muting it.. could you maybe try as well with some thin layer of something between pcb and barrel and experiment?
I'll try something again tomorrow..

User avatar
Techno Trousers
100,000,000 actuations

30 Jun 2016, 04:28

Wow, that barrel on the right looks like it was left sitting on a hot stove!

User avatar
lot_lizard

30 Jun 2016, 05:00

Techno Trousers wrote: Wow, that barrel on the right looks like it was left sitting on a hot stove!
Nuts right?!? FAIL!!! Even a trained monkey wouldn't let that out the door

User avatar
DMA

30 Jun 2016, 05:56

I like the idea of putting something between the barrel and PCB. I am also in favor of lifting flippers' feet off the PCB with that "something" - basically leaving the window for the flipper itself. It may also make signal better (It's hard to make it worse - I have pictures in the other thread).

User avatar
lot_lizard

30 Jun 2016, 17:19

So I picked up the metal plates from the machine shop, and they look spot on for dimensions. They cut the replacement top plates out of 20 gauge steel (thinner than the original 18), but for the purposes of testing... They will suffice. Our run later will be back to 18.

I did go ahead and tap the m2 screws for the bottom seal, and test fit everything. Included a couple pics below to wet your whistle. I will go dark for a few days until I have everything finished (sometime this weekend), with several pics of the steps along the way. The moral... I can confirm it works
Attachments
image.jpeg
image.jpeg (909.2 KiB) Viewed 4103 times
image.jpeg
image.jpeg (1020.95 KiB) Viewed 4103 times

Post Reply

Return to “Workshop”