deskthority

As naive as the title says : )

I know there's the XTant and I don' particularly like that solution, this is more what I have in mind: While we all drool over the new kishies, and have a laugh at some fake sales, I wonder: isn't the original XT just FANTASTIC in terms of ergonomics and form factor?

No idea how much pcb+metal+foam would cost, I guess not that much for the PCBs, with the added cost of machining the metal barrel-holder and laser cutting the mat, it could probably be around the 100ish bucks with enough partecipants?

Would there be any interest? I can offer my service for the plate/mat cad files, unfortunately I suck at PCBing, maybe someone wants to help?

Also, call me an heretic, but once all the cad files would be done, I'd be curious to try out alternative materials, like acrylic barrel holder, or ink printed conductive sheet instead of PCB : D
https://www.instructables.com/id/Print- ... t-Printer/

Considering it would not require more barrels than what an XT already ships with, I think having these files around could be pretty cool!

Seems the pic comes from Orihalcon with a f107 board cut and mod, but its still a good idea.

I think there must be someone who could finish the PCB design if even the 3277 pcb has been completed. Xwhatsit could still take over the controller job also, but can't think of how to get the appropriate curved metal plate over the foam, even with the CAD file. Maybe need a machine tool to do the part, or use plastic plate for some substitution?

Very interested in this project if could go further, following.

Yes, that picture is from a geekhack thread on fitting an F122 chopped pcb in an XT case.

To make this into a guilt-free mod kit, we'd reuse the original bottom metal plate of the XT and just work on a new custom PCB + top barrel holder plate + mat.

I now see what you meant about the curved plate, yes that needs addressing, annoying : (

Cindy has some F107 internals for sale if somebody would like to try it.
I think I personally would rather prefer a revive of the FSSK project as my OCD doesn't like that there are no case parts at the empty area around the arrow keys. A custom PCB and (as already seen in FSSK/FEXT) is not an issue, same for the mat. Producing a plate with this curve seems to be rather difficult though.

+1 would mod my XT like that

The XTant PCB was, apparently, a bit of a nightmare in daily use. I gave mine (and a matching curved plate) to DMA to see what he could do with it with his much more suitable Common Sense controller. Don’t know what came of it.

This stuff is harder and more costly than it first appears.

looks like the easiest (yet not very easy) solution would be to have a printed paper to mark the holes for a bolt mod and rely on the bottom plate to do the flexing on the custom top plate.

Yup, this might turn not as neat as I thought : (

Muirium wrote: The XTant PCB was, apparently, a bit of a nightmare in daily use. I gave mine (and a matching curved plate) to DMA to see what he could do with it with his much more suitable Common Sense controller. Don’t know what came of it.

This stuff is harder and more costly than it first appears.

But for most people still the only viable option to get a small Model F with decent layout Sadly a Kishy is out of reach for most people and the F AT is already quite a bit larger.

andrea-i wrote: Yes, that picture is from a geekhack thread on fitting an F122 chopped pcb in an XT case.

To make this into a guilt-free mod kit, we'd reuse the original bottom metal plate of the XT and just work on a new custom PCB + top barrel holder plate + mat.

I now see what you meant about the curved plate, yes that needs addressing, annoying : (

For the barrel holder plate, if the existing holes could match where the barrel would place in, that would save a lot of time also. Sadly im afraid not. So to save money and time, maybe to drill some needed circles of 14 or 15mm at the precise needed places would be a better idea. Not sure if this would be feasible without wobbling other barrels nearby.

true again, the space between the barrels of an F doesn't allow for screws.
The only space available is along the line separating the left F rows and on the right the line separating the numpad.
Would that be enough to keep the center part pressed? probably not.

I'm happy to not only design the PCB but also stream it and teach people how to make Model F and beamspring PCBs for themselves.

Now, if someone could solve the F-barrel problem, that would be great!@

I almost think that using XT's for their barrels/hammers isn't a total crime especially since no parts are actually destroyed and you could aways rebuild the XT from the parts again. Definitely think it makes sense to reuse the back plate, so really all you'd need is the PCB, barrel plate, and Xwhatsit. The most difficult part about all this would be the curve in the barrel plate which I think requires a special type of shop tool that isn't super common. I think 6-8 bolt holes could be strategically placed that would hold it together well without much modification of the barrels if any.

Hard to believe it's been over 3 years since making it, but that's my F87 (as I call it) which is just a F107 cut off because I had some spare parts that otherwise wouldn't work in their original configuration. I really like the keys between ALT and CTRL and I think it would be really cool to create PCBs that would allow those keys to be added to F122's (as well as full columns around the arrow cluster like the F107 has) and a PCB to add those to AT's. That's what mainly sets the F87 apart from an Alt modded AT (and of course it lacks one extra column compared to the AT due to that divider being there on the F107 that isn't there on the AT.

See here for the full thread:
https://geekhack.org/index.php?topic=71408.0

oh hi! I was pretty excited when I first bumped into that thread, thanks for posting it and great job!

My thoughts on the barrel plate thing:
Curving it is not big deal, just a matter of making a clamp out of wood with the right profile.
What bothers me though, is that for the barrel plate to work like it does on model Fs, we'd need to machine some pretty strong metal, cost goes up and game is over.
More so if we ask whoever cuts the metal to clamp it to curvature or worst, if we ask people at home to clamp themselves.

I think the only way to know if this works is to actually try it with a bolt modded aluminum plate and see if it holds.

Also, I'd love to see __red__ designing the pcb on video!

andrea-i wrote: Also, I'd love to see __red__ designing the pcb on video!

Alright then - I'll arrange a google hangout for it next week. I'll recreate the PCB that I did for @JP! 's beamdesk.

Definitely interested in this project. I unfortunately do not possess any skills necessary to help, but I would definitely buy one kit, possibly more.

Curving the top plate is easy if your fabricator/machine shop has a slip roller.

Controlling the XTant was fiddly because the xwhatsit sets a single threshold for the entire board. DMA's CommonSense controller has per-key threshold and costs much less at just $10. This is the way to go.

The XTant layout needed 2 or 3 extra barrels more than what an XT has - this would add 5 to that (7 or 8 barrels, 5 or 6 flippers, 4 inserts).

I would be happy to help too.

I stand gladly corrected!

What I’d really love to see is kits. Bring your own XT, and we’ve got the PCB, controller, plate and even caps and stabs for you. Common Sense integrated into the PCB would be great. (DMA’s subsequent Bluetooth version as an option would be ideal!) I could see people getting creative with LEDs (as Phosphorglow did on Model Ms) under the keys, indicating special functions like layers as well as basic Caps and Num locks.

Model F barrels and flippers are hard to source. (I’ve a good few extra for myself, for XTant and then for Model MF, but this is unusual.) Wouldn’t we be best off with a layout that’s 100% covered by one single donor XT? It doesn’t have to be the only one—this is Model F remember, and I’d want to HHKB it up with split backspace, like my Kishsaver—but 83 keys just might be worth consideration.

I know the XT case rather begs for more keys—so many oddball big keys worth of space to fill!—but 6 keys in the nav island + 4 arrow keys is pretty much enough for me byeond the 60% block and a well placed Fn key.

I'd also be up for using the exact number of barrels that ship with an XT.
If anyone knows anybody able to make curved steel plates for cheap give us a shout!

As soon as I get my hands on another XT (gave mine away, doh!), I will start experimenting at home with a flexible top plate with screws.

andrea-i wrote: Also, I'd love to see __red__ designing the pcb on video!

Scheduled - see thread in Workshop.

andrea-i wrote: I'd also be up for using the exact number of barrels that ship with an XT.
If anyone knows anybody able to make curved steel plates for cheap give us a shout!

Many HVAC shops have a slip roller which is used to put a curve on sheet metal. I got lucky and found a shop to do this for free.

I have a quantity of curved steel plates from M122s. These could be cut down to make XT plates? I also have large quantities of Model M flippers and springs, as well as stabilizers.

elecplus wrote: I have a quantity of curved steel plates from M122s. These could be cut down to make XT plates? I also have large quantities of Model M flippers and springs, as well as stabilizers.

M122 plates will not do us any good as they are back plates and we need new top plates. F122 top plates might work - if only we had buttress type (122/AT) barrels for them. Unfortunately, we have pin type (XT) barrels.

wcass wrote: if only we had buttress type (122/AT) barrels for them. Unfortunately, we have pin type (XT) barrels.

If only there was someone in the community who had moulds for the barrels, flippers, a source of springs, and willing to actually sell and deliver them today... :-/

Pardon my ignorance , but here I'm seeing 83 keys, considering we might need stabiliser barrels for at least backspace, enter and 2 shifts, that adds another 4 barrels, even if we add another two win keys (yack!), isn't the XT layout made of 104 keys?
raw json:
[{a:7},"","",{x:0.25,a:4},"¬\n`","!\n1","\"\n2","£\n3","$\n4","%\n5","^\n6","&\n7","*\n8","(\n9",")\n0","_\n-","+\n=",{w:2},"Backspace",{x:0.25,a:7},"","",""],
["","",{x:0.25,a:4,w:1.5},"Tab","Q","W","E","R","T","Y","U","I","O","P","{\n[","}\n]",{x:0.25,w:1.25,x2:-0.25},"|\n\\",{x:0.25,a:7},"","",""],
["","",{x:0.25,a:4,w:1.75},"Caps Lock","A","S","D","F","G","H","J","K","L",":\n;","\"\n'",{x:0.25,w:2,x2:-0.25},"Enter",{x:0.25,a:7},"","",""],
["","",{x:0.25,a:4,w:2.25},"Shift","Z","X","C","V","B","N","M","<\n,",">\n.","?\n/",{w:2.75},"Shift",{x:0.25,a:7},"","",""],
["","",{x:0.25,a:4,w:1.25},"Ctrl",{x:1.25,w:1.25},"Alt",{a:7,w:7.5},"",{a:4,w:1.25},"AltGr",{x:1.25,w:1.25},"Ctrl",{x:0.25,a:7},"","",""]

An XT has 83 keys, 83 barrels, and 83 flippers. The barrel under the space bar is different though, so should be replaced. You are correct that extra barrels are needed for Backspace, Enter, shifts, and Win keys.

ok thanks, I was too lazy to count and the internet is not always giving you the best answers : D

So, if we take out 5 barrels from the right cluster and make it into a standard tkl layout without winkeys, we should be good.

Some more food for thoughts, has anyone with a good SLA printer tried 3d printing the xt barrels out of resin? I think the resolution of those machines could be good enough to produce something that doesn't feel scratchy.

andrea-i wrote: has anyone with a good SLA printer tried 3d printing the xt barrels out of resin? I think the resolution of those machines could be good enough to produce something that doesn't feel scratchy.

yes and yes

I wonder why only the F62 and F77 are built again. Wouldn't the AT be a good third option?

thorough wrote: I wonder why only the F62 and F77 are built again. Wouldn't the AT be a good third option?

When I asked Ellipse about this 2 years ago he said he wanted to focus on a smaller selection of products to make it easier to build/sell etc. Having more models just complicates the already herculean efforts in the project