deskthority

I wanted to share a terrific 3d printed solenoid and solenoid driver holder project from f4pv:

https://www.thingiverse.com/thing:7076392

This allows for secure mounting of both the solenoid driver and solenoid inside the F62 and F77 keyboards, as well as being an external solenoid holder for the compact case F62 and F77 keyboards.

As a note, new Model F custom front printing can be done, but there will be a setup cost and MOQ. However, only the keycap colors shown on the project web site can be custom requested. To clarify my email reply, the set is not stocked as it would need to be special ordered from the factory at extra cost and with a MOQ, so I confirmed that it would not be possible to order one set (that may not have been clear in my email).

Production update:

Today we have some more production photos, this time of the powdercoating process of the top inner assemblies and the completed bottom inner assemblies.

2025-06-19 beam spring keyboard production (1).jpg
2025-06-19 beam spring keyboard production (2).jpg
2025-06-19 beam spring keyboard production (3).jpg
2025-06-19 beam spring keyboard production (4) - Copy.jpg
2025-06-19 beam spring keyboard production (5).jpg

Today the (long overdue) blog update has been posted. This gathers together the many updates posted on the forum project threads.

Brand New Model F June 2025 Update: user reviews and photos, F122 and beam spring production, new design plans (Split Ortho, Planck, Preonic, F20, Kinesis, etc.), reorganized manual, updated firmware

https://www.modelfkeyboards.com/blog/

A new video below shows the process to open and close the inner assembly for the larger keyboard models (F104, F122, FSSK). This was just added to the manual as well.
PsBUlTvm8jA

New as of June 2025:  Mac Icons+Text 6 key set 

By popular request - a set combining the Mac Icons and Mac Text key sets. This is a special order set that will arrive later this year, so you'd need to wait for delivery if combining shipping with a current order, or order separate shipping. 

Feel free to head over to the Extra Keys page to see more details and order the set.

Mac Text with Icons.PNG

Update on beam spring inner foam:

Everyone can now order an extra beam spring inner foam set at the below link, which has been updated today to add this item. Same as with the Round 1 beam spring first aid kit, extra inner foam was not a part of the kit as it was with Model F First Aid Kits, so it would have to be ordered separately.

https://www.modelfkeyboards.com/product/beam-spring-extras/

If you are just ordering some extras, please feel free to place a new order and choose free/other shipping at checkout, right before clicking Place Order, to add to an existing shipment.

These are for the Round 2 classic style beam spring keyboards. The cut out one goes below the inner assembly plate and on top of the beam modules (just like with the IBM originals) and the bottom one goes below the PCB. Both are made of the same custom material as the newer batch of XT (non-crumbly style) Model F inner foam.

I also have some Round 1 beam spring foam sheets; these are just for the bottom one that goes below the PCB; feel free to email me if you want to order one.

Here's a photo of one of the first samples of the foam from a couple years ago - obviously it is a little rough compared to the production version.

beam foam.jpg

Yes, as noted from the beginning of the project, all stepped keys are from Unicomp, including ISO enter and caps lock. Their color is slightly different from the original Model F key colors.

As a recap, the factory failed to produce these extra keys to my specifications and they did not function reliably, so the order was cancelled. At some point I hope to resume that project to offer those extra few non-stepped keys (PC AT Enter, ISO Enter, Code) in various project colors.

If interested please sign the Google form: https://docs.google.com/forms/d/1vsamkl_jOqp6wAEfZGeivQZ8ZaibpoT0WQkDG39r08Q/

Posting (with permission) a nice F77 setup. The photo was just sent to me today and was taken after a year of usage by a writer who is working on a novel!

This F77 has one of the most popular add-ons, the solenoid, which is installed inside the case.

F77.jpg

Production update:

The factory has completed the lasering and bending of the B104 top and bottom inner assembly plates (shown below); the next step is the coatings to finish these parts.

After they are coated they expect to start assembly, as early as the next week or so. During the time of assembly they will be making the cases, which is a quick process now that they have all of the laser and bending settings worked out. This is done to optimize warehouse space (another reason they order the boxes and protective foam as late as possible before shipping).

As I mentioned before, after the B104, they will work on the remaining 3 keyboard types (B62/BSSK/B122) and hope to finish up as planned.

It is great to finally see the beam spring keyboards coming together!
2025-06-04 B104 TIA production units ready for finishing (1).jpg
2025-06-04 B104 TIA production units ready for finishing (2).jpg
2025-06-04 B104 TIA production units ready for finishing (3).jpg
2025-06-04 B104 TIA production units ready for finishing.png

The stems for the new relegendables were sourced from Unicomp, which did not separate their stock by color. I did request all one color like the early Model M's but they were already mixed together.

Of course the relegendables are intended to work with the paper templates that will cover all visible sides of each keycap, so the stems would not be visible.

Posting (with permission) another great F122 setup.

This one is special as it has an old hi-fi Pioneer badge (for a cool retro styling) and Unicomp Esc and APL keys.
F122005.jpg
F122002.jpg
F122004.jpg

Thanks for the feedback; I have noted to try to keep tighter case margins for the forthcoming Model F designs. Since the cases may be 3d printed or laser cut, this may be easier to achieve.

It seems like most folks like the case margins to go about 1mm just over the bottom of the unpressed keys, and to keep the case margins at most 1mm away from the keys? Because of the Model F's usage of a curved plate, case design is a bit more limited.

Regarding the case margins, the F122 and Model M cases' function keys can be closer to the keycaps because (with the former) the case tops do not have a "wall" that goes down to and just about contacts the top inner assembly. This wall would contact a key when it is pressed down, if there is not enough margin, since the wall is at a different angle than the key (the case is die casted straight and it is best to have a good few degrees of draft angle and avoid undercuts). Without this wall the case could be closer to the keys if preferred. And in the latter case, PVC (IBM Model M plastic case material) has a 0.5% shrinkage rate while aluminum shrinks by as much as 5-7% and it can shrink unevenly, similar to PBT. Better to have too much margin than for keys to get stuck when pressed down. Maybe I should remove all such walls for future designs, if it will allow for tighter margins.

Another way to reduce the margins is wcass's 4 layer Model F PCB innovation, which would reduce the PCB area by moving the traces to another layer. This may make the forthcoming compact designs even more compact.

Great news from Rico - split communication is now working on the Leyden Jar controllers!

Please see Rico's update post here and head over to the github pages linked below:

https://www.keebtalk.com/t/the-leyden-jar-a-tentative-pcb-replacement-for-the-brand-new-model-f-keyboards/17489/45
https://github.com/mymakercorner/vial-qmk/tree/leyden_jar
https://github.com/mymakercorner/Leyden_Jar_Diagnostic_Tool
https://github.com/mymakercorner/Leyden_Jar

Check the leyden diganostic tool signal level monitor - maybe a key is being registered as pressed down.

Maybe the solenoid L bar itself needs adjustment as it's getting stuck. Does the cylinder move freely by hand and not get stuck? Are you seeing 6 to 9 volts where the solenoid is plugged in, when keys are pressed? Then it may be due to the solenoid.

The solenoid should work by default without any configuration or keypress combinations. I recommend erasing the eeprom in the leyden jar tool and reflashing the firmware. Since the solenoid is able to fire, it may be an issue with your USB port (not enough power, USB hub, etc.). I recommend trying another port, another computer, and maybe trying your other Model F keyboard if all else doesn't work.

With permission I am sharing the first photo of the Hebrew English key set on the new F122!

There are a couple dozen available key sets and it is nice to see photos of some of the rarer key sets.

Hebrew English.jpg

thrillscience kindly see my replies below:

1. The issue is "the horizontal gap because the bottom plate hasn't been slid over enough" because the plates may not line up properly and could be more likely to separate. The plates should be able to slide so that they cannot travel any more and the sides of the tabs of the TIA make contact with the BIA cut out areas.

2. Yes, I'd like to take a photo one of these days showing the clamps, but feel free to search for "F122 model f clamps" to see how these larger keyboards are put together. It takes a few minutes with the right tools, or an hour otherwise! Some of the clamps should be inserted so that they are compressing the area between the function keys and top row of numbered keys, so that those TIA tabs rise enough above the BIA to clear it and allow the plates to slide together.



ms264556 the recommendation is to print out the templates on paper and cut out each one to fit the key. That way the legends can all have the same paper color, the key stem is no longer visible after the template is installed, and one can change the legend in the future. Unicomp provided the stems (they are regular Model M stems for two-piece keys) and noted that they are all assorted colors; no way to have all the same color.

Additional transparent relegendable template:

With permission I am reposting a transparent relegendable template:

credit to geekhack member Cyber_engineer – Twitter @Mates_ENGINEER

Very nice template!

Leyden Jar update and request for help:

So far over 130 F122's and their Leyden Jar controllers have shipped. After some firmware updates the reports are good so far!

Here are some plans for the Leyden Jar controller:

Need help - Split communication between halves, with one half powering the other (for future split ergonomic layouts):

"Well for the moment one half is working but I have trouble making the split feature work. I need time to understand what is wrong with my code: as I have a custom matrix scanning this looks to complicate stuff (all the project I have seen look to use standard QMK matrix scanning). I need to make split working and also try different settings like full duplex (testing half duplex at the moment), various detection of left and right parts"

Rico has kindly published this test code for split communication with the below link. If anyone can help it would be much appreciated to make these controllers even more powerful and useful with various keyboard projects. Rico confirms that the controllers work even with non-capacitive keyboards. https://github.com/mymakercorner/vial-qmk/tree/leyden_jar_impostor

I would like to use a "USB C cable to connect the two ends instead of a TRRS cable" as the latter may result in accidental shorts if the keyboard is powered on during disconnection.

4 layer PCBs:
Using 4 layer PCBs as innovated by wcass to remove the trace area around the PCB, for a smaller PCB (better for the future compact designs mentioned in the Google form above.

PS/2 support:
Rico has started this but it is not yet working. If you'd like this feature and can help, please do! https://github.com/mymakercorner/Leyden_Jar

RedESC please see my previous F122 update post from a couple months ago (March?) regarding the queue and shipping timeline.

Round 2 is expected to all be delivered together, including the cut out mod boards.

Thanks; please keep the custom notes coming!


Project milestone - this week the 6,500th keyboard shipped, for the new Model F and Beam Spring keyboard projects! Who would have known that we would reach that when the project started and I was getting quotes to make 50 F62 and 50 F77 keyboards!

Thanks for sharing photos of your modded F122 Ordinary Witch!

I like the terminal-style customizations that remind me of the F122 originals.

Can folks please share what they do with the custom keys in Vial? To give people ideas on the possibilities. Are they macros? Key combinations? Do they open specific programs?

F77E39 I can confirm that the two issues were discussed over the years. Regarding paint loss, this is disclosed on the project web site as you mentioned, and the paint formulation changed years ago to a tougher powdercoating with no reports of premature wear (the oldest keyboards with this new powdercoating were the compact F77 and compact F62 keyboards that were powdercoated and die cast instead of CNC milled and anodized; the new cases started shipping in early 2022). In summary every keyboard besides the two originals does not have this issue because the issue was due to the old style paint used on all the original style zinc cases; once current stock is depleted that powdercoating will no longer be available, but it is the only option if you want one of the original project boards. It is not an issue related to a possibly defective batch as there are far more reports of keyboards from that batch in perfect paint condition after years of daily usage.

Regarding a few reports of scratchiness:

For those few folks reporting scratchy keys please do some more research to see what is the matter or what some possible fixes may be (see the links below). Out of thousands of boards shipped since 2019 this issue has only been reported by a handful of people and only on the DT forum, so I am inclined to believe that this is more of people noticing differences compared to their other keyboards, which may be described as having a smoother feel, whether IBM or not.

I just went back in my emails to see previous discussions. In two of the earliest discussions, folks reported scratchiness with off-center pressing in an original IBM XT Model F keyboard, and it persisted with one example even after a thorough restoration including ultrasonic cleaning of the parts. Otherwise I can't find any reports of new production 1U keys being scratchy outside of the several reports here.

I see a few discussions of scratchiness in original IBM Model M and Model F keyboards (linked below), including some mod attempts that folks have tried, revolving around removing excess flashing. I don't see excess flashing on the new production keys but maybe trimming parts of the stem like the very bottom edge of the keycap or the semicircular area may help with original keys. Are these keys a fraction of a mm too tall and it is touching the PCB when pressed? According to one DT member in 2014: "The other thing to check for - is the bottom lips of the stem don't have any flash on them. I ran into this a couple times with new stems from Unicomp - where a little bit of flash would rub against the spring making it feel scratchy. Quick work with a razor can clean this up."

https://deskthority.net/viewtopic.php?t=20660
https://deskthority.net/viewtopic.php?t=25030
https://deskthority.net/viewtopic.php?t=9128
https://deskthority.net/viewtopic.php?t=23828
https://deskthority.net/viewtopic.php?t=22673
https://deskthority.net/viewtopic.php?t=8343
https://deskthority.net/viewtopic.php?t=15264
https://deskthority.net/viewtopic.php?t=21054
https://deskthority.net/viewtopic.php?t=20043
https://deskthority.net/viewtopic.php?t=24389
https://deskthority.net/viewtopic.php?t=27212

I don't know what the current status of that is from those who have been doing tests. There was some discussion of using a dremel to remove a mold parting line inside the barrel, but I am not sure if that fixed the issue that a few folks reported.

resonator, did you replace them with caps and barrels from the same batch or from a different shipment from years before or after the originals? Have you been using the board for many months without issue after replacing the barrels and keys? Can you try the reported scratchy keys and barrels outside of the keyboard to see if they are the same, and see if you can notice anything different with them? Could it be an issue of tolerances (maybe the key stem was a little too wide and/or the barrel interior was a little too small - you may need a micrometer or caliper to determine this accurately)? Do you see evidence of excess flashing or wear on the keys or barrels?

The above keyboard with an IBM logo is mine.

I just wanted to provide a quick review.

I have been using PCs since the early 90s, got started on old IBM AT/XT stuff I found cheap at garage sales and such, loved systems from the 80s. 20 years in IT doing everything from programming to networks to sysadmin to cybersecurity. Run my own business now.

About a year ago I decided to look into mechanical keyboards after years of using whatever was sitting around and discovered there were buckling spring keyboards available. I had many fond memories of these growing up on my early collection of PCs. Ordered an F122 and a couple Unicomps new Model Ms about a year ago to use for work and hobby. The Unicomps were a big improvement over the cheap junk I had been using, but I was very surprised how much better this F122 is than the Unicomps. Have been using the Unicomps for a year, just got my F122 today.

The setup is going to take you some time, but I was able to get mine from in the box to assembled, tested and mapped with Vial within a few hours. Had to replace a spring on the 10 key + with a spare I had in the repair kit I ordered with the unit, also had to do the spring reseat on F8.

The feel of this keyboard is amazing. I was not expecting there to be such a jump from the Unicomp to these. Less torque required for each keystroke by what feels like 10-20% at least and the action is so smooth. The sound is very different, higher pitch with more of a ping than the Unicomps/Model M. The keyboards I remember getting started on so many years ago were loud and so that does not bother me, I like it. Been happy with the Unicomps but this Model F is just a whole different experience and quite a bit nicer to type on. Also Vial is awesome, and with the macros feature you can do pretty much anything you want with these.

This project is quite an accomplishment. I've done custom hardware and software with commercial sales and I know that is hard to get right, and honestly the fact that I can go order a new condition Model F of any sort in the 2020s is a little crazy. I know from following posts here there have been teething issues to work through but this is a Herculean effort on the part of Ellipse to be able to deliver the product I am typing this on. Well done.


Thanks for your review ClickNClack and for the photos!

Posting (with permission) another nice F122 setup, customized with keys from various key groups on the Extra Keys store page.

This one features one of those third party IBM new production badges (I believe from eBay) that I do not offer. The spot on the keyboard is designed to fit the IBM F122 style badges (the PC AT badges won't fit since they are bigger).

Everyone please share your new Model F photos, or send them to me if you prefer me to share them.

F122 various (1).png
F122 various (2).png

Thanks PeterQuines!

I'm not sure I understand your question. Are you asking if the F77 layout will be made for the beamsprings? Any layout is possible but the tooling costs would have to be covered. Since the BSSK is quite close to the F77, especially with the 3x5 cut out mod, I was not thinking about the B77.

If anyone has requests for future layouts, whether F or beam spring, please do fill in the Google form here:

https://deskthority.net/viewtopic.php?p=524139#p524139

The Leyden Jar firmware has just been updated (same link as my earlier post). Please let me know if there are any issues!

New updates from Rico:
The NKRO option works once more (click Fn+Space+N to enable as it is disabled by default for compatibility reasons)
2nd and 3rd layer shortcuts and mappings for Fn and menu keys now match the xwhatsit Model F firmware shortcuts (see the project website manual for details)

Important note: while in the xwhatsit NathanA firmware you can split and unsplit keys in Vial, for the Leyden Jar firmware you would have to flash the generic firmware for your keyboard (for example, leyden_jar_f122_vial.uf2) to enable additional pads. This firmware is useful for using the shorter spacebar and for using the 5 keys above the cursor keys for those who are adding barrels, flippers, and keys to those spots.

Yes, just order the keys not included in the US ANSI key set from the Extra Keys page and order the $39 layout customization if you want me to take care of it. Note ISO Enter, 2.25U left shift.

tinydarkness the keys are available in pearl/pebble, Dark Gray, and Blue.

Iggy yes, with the layout customization check mark - you can note you want a shorter spacebar, regular spacebar, short left shift, full size left shift, ISO Enter, etc. Also don't forget to order the additional needed keys that are not part of the ANSI set (ISO Enter key, 1U | \, short left shift, etc.).

Important F122 Leyden Jar firmware update to fix phantom key press issue for some keyboards, and other replies:

Nice customization tinydarkness! I like the use of Industrial SSK and terminal keys.

Confirmed, tracking emails are sent as soon as I have the tracking information.

Wow that is fast! Usually they are not that fast!

With permission I am sharing another nice customized F122.

Mopar Blue set F122 - Copy.jpeg



When you have a moment, everyone please test the updated F122 firmware and let me know if there are any issues.

https://www.modelfkeyboards.com/wp-content/uploads/2025/05/leyden_jar_f122_vial_2025_05_10.zip

Rico is working on a new firmware update that will fix a few other small requests too, but I recommend updating to the above firmware if you have any issues with the debounce/with phantom key presses. The above link will no longer work and the file will be deleted after the update is made and added to the regular QMK-layout-files.zip file in the manual, where the various firmware files are all located.

The debounce was not working at all on the original firmware because the debounce time in was set too low for the higher number of bins used in the default firmware (more bins requires a few thousandths of a second more debounce time). More details are below if you are interested.


You have to manually set the debounce time as it is a constant. Setting this time too low results in debounce not running at all, so a setting of 10 ms is recommended as an initial test. Rico notes "What needs to be done for the F122 keyboard is to increase debouncing time so that it is higher than the matrix scan time + additional qmk logic. Total latency is approximately max(debounce_time, matrix_scanning_time) + qmk logic. This means that if the matrix scanning takes more time than the debounce time, no debouncing will be done. With 6 bins we have ~4.3ms scanning time, this is not a huge number. Sure QMK USB polling rate take an additional 1ms and debouncing takes 5ms on top of that, we still end at ~10ms latency which is more than adequate for normal typing sessions and casual gaming (below this value only the very best players can see the difference, and they are more probably inclined to use a hall effect gaming keyboard). To reduce latencies maybe it would be easier to reduce the debouncing values (and check that it does not introduce double press issues), or even better select another debouncing algorithm (there is one that introduces no delay when detecting key presses, only one when detecting key releases). To change the debounce time, the following line has to be added in the info.json file of the f122 board:"
"debounce": 10

The only downside to longer debounce times is a couple thousandths of a second additional input delay. Many mechanical keyboards are about 10ms anyways, about the same as Leyden Jar-powered keyboards.

Here is an interesting solenoid mod, along with a description. I do not know the long-term effects of this, so I cannot recommend it, but it is something to think about! This new F122 user wanted a much lower impact for the solenoid, even beyond the sound from the minimum throw length.

"I also restricted my solenoid a little bit further with some electrical tape. I applied it and stretched it a bit further out using the top bracket. I would say the sound without it has a bit of a harsh/high frequency impact -- metal on metal kinda sound that reverberates through the backplate. Adding the tape sorta rounded it out, making it work much much better with the keys to my ears. Much more subtle and typewriter-y."
solenoid electrical tape mod.jpg

And here is another one! This one is a dark gray F122. Posting with permission as per usual.

dark gray f122 - Copy.jpg