deskthority

Please refer to the mother GB thread which can be found here

Introduction

Note: this post will be updated. As such it shall be considered work in progress. A change log is shown at the end.
Last change 2025.01.01 (yyyy.mm.dd)

The Sensing PCB is considered priority 0, being 0 the highest priority.
As such, at this stage it is important to record constraints that would affect specifications of other components.

Constraints

• Conductive specifications, impacting the change of capacitance of the Sensing PCB and therefore the sensing of the controller.
• Compatibility with IBM model M boards to ensure easy FSSK/FEXT-like projects

Alternatives currently considered

• Injection molding - under investigation
• Filament 3D printed
  • conductive material
  • 3d printed part with conductive paint
  • 3d printed part with coating (graphite)
• Resin 3D printed
• Epoxy with a percentage of graphite

This thread should replace the already existing give and take: 3D printed model F flippers - TBD porting here the information from the original thread

--- change log (yyyy.mm.dd) ---

------------------------------------------------------------------------------------------------------------------------------
The following spoiler section contains the first message of the original thread which was named "give and take: 3D printed model F flippers (closed in favor of FSF component)". I had to use this first post as introduction and I did not want to delete it
------------------------------------------------------------------------------------------------------------------------------

Hello again,

Here is the 3D model of the F flipper which I have made
Next step is to print it.

(check following posts for the 3D files)

The flippers with springs are available for $1 each, same price for many years. Everyone is not paying for just the materials but for the years of R&D and tooling to bring these keyboards back to life.

PLA filament seems to get brittle over time, which is why 3d printed parts have a much shorter expected life than injection molded parts and they do not appear to have been successfully used in Model F keyboard projects. Why save 50 cents now and have to spend hours to redo all of your work in a couple years from now when flippers break or do not work reliably?

Just a note that the shipping price on the project web site is for keyboards. For small extras, the shipping is much lower, depending on weight. The site notes to email me for a quotation based on what you want to get.

Ignore ellipses desperate plea for your money and 3d print the part lol

The guys a scumbag and wouldnt get away with half this shit if he lived in the EU e.g. changing cut off dates for orders every month

3d print this shit.

Hello Ellipse,

Ellipse wrote: ↑

05 Dec 2024, 04:44

The flippers with springs are available for $1 each, same price for many years. Everyone is not paying for just the materials but for the years of R&D and tooling to bring these keyboards back to life.

In my opinion, you should be sharing the R&D with the community.
It is a matter of fairness. Give and take.

Ellipse wrote: ↑

05 Dec 2024, 04:44

PLA filament seems to get brittle over time, which is why 3d printed parts have a much shorter expected life than injection molded parts and they do not appear to have been successfully used in Model F keyboard projects. Why save 50 cents now and have to spend hours to redo all of your work in a couple years from now when flippers break or do not work reliably?

Just a note that the shipping price on the project web site is for keyboards. For small extras, the shipping is much lower, depending on weight. The site notes to email me for a quotation based on what you want to get.

You may be correct with the 3d print vs injection molded parts.
But you also said "they do not appear to have been successfully used in Model F keyboard projects". Perhaps is because no-one has ever seriously try

The alternative is to produce in China: please share the information you hold (CAD, springs specs/source and how to purchase from China) with the community to help in producing the flippers in the same way the community has done with the critical components you have (and still are) using in your products. (Give and take)

BTW: The numbers are as follow: some $5 vs (approx) $150 in most of the countries:

• $1 x 104 = $104
• Shipping would be some $25
• Customs $20

$1 is simply too much for a DIY project. This thread is about solving this issue. I personally have enough flippers for all my keyboards. I personally do not need any more.
Once more ... give and take. I may success in giving the flippers to the community; I take Rico's Leyden Jar, including all what I have learned from his open design and shared conversations.

But let's stop this ($) topic here. I will not enter in how to share profit in this post. I mean, how to fund R&D. We leave this to a different one

-- PD: You will find a "give" example as attachment. I am not sharing the Fusion360 files as the design needs some adjustment where the spring joins. I will share them when the design is proven.

Not a bad project at all. Not an expert in EE and 3D printing but conductive filaments have very high resistance, I personally would have gone with nylon... filament and use conductive paint or something similar but this could just work perfectly fine for this use case.

ergya wrote: ↑

05 Dec 2024, 10:03

Not a bad project at all. Not an expert in EE and 3D printing but conductive filaments have very high resistance, I personally would have gone with nylon... filament and use conductive paint or something similar but this could just work perfectly fine for this use case.

Thanks a lot for contributing !

I thought about conductive paint already. I have some experience with it.
I also thought about addesive copper: https://www.amazon.de/-/en/gp/product/B ... OA8H&psc=1

If we're talking about https://proto-pasta.com/products/conduc ... 1265211476 - it should work, despite not being really conductive - it's more of a "dissipative" kind.
I plan to make flat MX-compatible model F and looks like I'll need to 3d-print the prototypes (because I don't have $75k for the 3 molds required - barrel, flipper, and MX-key-stub) - might as well get the spool of that and print.

The adhesive copper won't work - adhesive will dry out in couple years and fall off.

Hello,

I have succeeded in creating a model which works. Check the compressed video (I will post a better one, I do not have time to do it now):
Attached you can find the fusion360 file Version 5 (zip)

I also include pictures of the design, made for 3D printing and the way to print it. I also include the print project (zip)
To avoid any kind of support I have used glue in the 3D printed bed.
The printer is a bambulab XC1. I have used PLA in this test
The result is solid and good.

DMA wrote: ↑

12 Dec 2024, 03:03

If we're talking about https://proto-pasta.com/products/conduc ... 1265211476 - it should work, despite not being really conductive - it's more of a "dissipative" kind.
I plan to make flat MX-compatible model F and looks like I'll need to 3d-print the prototypes (because I don't have $75k for the 3 molds required - barrel, flipper, and MX-key-stub) - might as well get the spool of that and print.

We can test this filament, indeed.
We can skip the mold for the barrel. I am 90% sure that it can be 3D printed.
Regarding the MX-key-stub, I will not need it but it can also be designed to be printed.

DMA wrote: ↑

12 Dec 2024, 03:03

The adhesive copper won't work - adhesive will dry out in couple years and fall off.

I am not sure ... It can be covered with a thin layer of epoxy.

Let me repeat a question for which we did not get the answer:

Elipse: (Give and take) why don't you please share the information you hold (CAD, springs specs/source and how to purchase from China) with the community to help in producing the flippers in the same way the community has done with the critical components you have (and still are) using in your products ?


Cheers

i$

idollar wrote: ↑

12 Dec 2024, 08:13

Hello,

I have succeeded in creating a model which works.

Wow, man, LUDICROUS SPEED!

idollar wrote: ↑

12 Dec 2024, 08:13

We can test this filament, indeed.

Since you already at the "printing" stage - I'll defer the testing to you

idollar wrote: ↑

12 Dec 2024, 08:13

I am not sure ... It can be covered with a thin layer of epoxy.

From my experience, the bottom must be pretty flat, so this will need to be a pretty thin, pretty even layer of epoxy, which must withstand repeated hits. Sounds quite hard to me.

idollar wrote: ↑

12 Dec 2024, 08:13

Let me repeat a question for which we did not get the answer:

Elipse: (Give and take) why don't you please share the information you hold (CAD, springs specs/source and how to purchase from China) with the community to help in producing the flippers in the same way the community has done with the critical components you have (and still are) using in your products ?

LOL you'll never get an answer to that one.

Bro actually asked him to share something yeh you aint getting an answer from him, sorry buddy

[dupe]

OK, I 've bit the bullet and bought a resin printer ($700 after christmas discounts) and a bottle of silver-coated-copper conductive paint (MG Chemical 843AR, ~$40 for 55ml). Could've bought 2 ellipse's keyboards instead, I guess - but I hope my kid will have more fun with a 3d-printer.

Resin - because
a) I wish something close to injection molding tolerances and
b) because of the additional post-curing step (I've read that models come out of the vat sticky, which should allow for some penetration into the layer that contacts the PCB - plus nothing prevents me from painting the conductive layer AND THEN painting some uncured resin on top of it and curing it that way)
The initial idea was to insert "dust the PCB-facing flipper surface in conductive carbon black and paint that over with uncured resin" step between "wash" and "cure" steps - but after searching amazon, two problems became painfully apparent:
1) from customer reviews, looks like lots of things selling as "carbon black" are actually "polypropylene black" and not conductive at all.
2) particle size of "10-100 nanometers". Just for comparison, laser printer toner is 8-12 micrometers, and that stuff is already quite evil should you spill it (or just breathe towards it, really).
Another issue is that carbon black isn't that conductive, really - the fact that carbon black is mixed into bulk of the plastic really helps OG model F filppers, but in a thin layer it will likely be at the edge of what CS can pick out on "model E"-style/connected pads, and certainly way below what xwhatsit on F-style/floating pads can discern. Silver-coated copper gives 0.008 Ohm/sq vs carbon's ~100 - about 10000 times better. Silver would give another 4x improvement in conductivity, but would cost $180/55ml, so I decided not to go there.

Hopefully, I'll have something to show in a month or so.

Hi,

Thanks a lot DMA !

Any investigation to get the flippers done is very welcome.
Creating these flippers is the first step to manufacture our own and affordable Model F keyboards !

Cheers

Hello,

The following video tutorial came to my attention:
Electroplating the flippers should definitively work !

The process seems simple:

• 3D print (already done)
• Paint with graphite - https://www.ebay.de/itm/387291730590?_s ... BM7pzHuPpk
• Polish - https://www.ebay.de/itm/374010318902?_s ... R9TU8rj6ZA and https://www.instructables.com/Affordable-Rock-Tumbler/ or https://es.aliexpress.com/item/10050031 ... ry_from%3A
• Electroplate

I need to check this in January


Cheers

..also, since I'll have whole 55 grams of supposedly-acetone-based conductive paint along with a hundred-ish model M flippers - I can actually try and make a model M filpper capacitive by painting the underside.
it is extremely likely that PCB will need to be slightly reworked (basically, by moving the pads up almost to the round dots) - but if the point of FSSK is to have capacitive sensing and NKRO, not the sound of the model F (the feel is already virtually similar TBH), it can be achieved by spending $40 for the paint and several hours grinding out the nubs. That 55ml can is supposed to last ~21700 cm2 (3360 square inches for you americans), which is what, 20 keyboards worth? No need to electroplate - electroplating is for decorative parts, not supposed to be banged on PCB for thousands of times, plus it can produce fumes way more toxic than acetone.

OK, I've got the conductive paint. It's acetone-based, settles really quick, looks pretty mesmerizing - like a snow globe - and is unexpectedly gold in color. I'm tempted to repaint my silver wingnut gold! Whole thing settles in 10 minutes, and dries in 3. You can also see how good this stuff is by counting warning signs! It's also REALLY fluid - touching surface of the liquid sucks those golden flakes into almost a centimeter of horse hair instantly. After drying on the brush, golden dust just flakes off the horse hair - so, I guess, adhesion won't be that great. I'm giving it about a day to fully dry out, then we'll see if it any good and how durable the coating is.

Overall, I think it will definitely work for painting uncured resin-printed filppers, letting it dry for 5 minutes, then covering with a layer of uncured resin and then curing the prints. Tomorrow we'll know if this is a viable technique for converting model M filppers into capacitive role.

Great !

Elipse: (Give and take) why don't you please share the information you hold (CAD, springs specs/source and how to purchase from China) with the community to help in producing the flippers in the same way the community has done with the critical components you have (and still are) using in your products ?

idollar wrote: ↑

21 Dec 2024, 10:14

springs specs/source and how to purchase from China

The way I currently understand it, purchasing springs from China is easy, you just need to purchase a ton (like, LITERAL ton, 1000 kg) of them at once. And if you're buying this much - specs not needed, just ship them a sample.
The reason for this is that spring-winding machine is a bitch to setup, but once set up it's extremely fast, like thousands per hour.

OK, confirming that painted model M flipper gives about half the signal original model F flipper does - but still about 2x the noise floor, so it's workable.

Interesting analysis DMA! What is the best way to measure signal and noise floor? I like to test PCBs and flippers by just placing a flipper on the PCB pad and monitoring changes in the pandrew signal level monitor utility or Rico's Leyden Jar diagnostic utility.

Ellipse wrote: ↑

25 Dec 2024, 00:00

Interesting analysis DMA! What is the best way to measure signal and noise floor? I like to test PCBs and flippers by just placing a flipper on the PCB pad and monitoring changes in the pandrew signal level monitor utility or Rico's Leyden Jar diagnostic utility.

Fuck you, you never give back to the community so I'm not telling you.

I tried to print some calibration targets on a newly-bought resin printer and not printing anything anymore until I build a fume hood that vents to outside. Stuff like "mars mate" with charcoal filters not only won't really work, but also bankrupt it's owner for the filters
So, see you after fume hood is built and after surgery and recovery - so, somewhere closer to February I guess.

DMA wrote: ↑

27 Dec 2024, 05:28

I tried to print some calibration targets on a newly-bought resin printer and not printing anything anymore until I build a fume hood that vents to outside. Stuff like "mars mate" with charcoal filters not only won't really work, but also bankrupt it's owner for the filters
So, see you after fume hood is built and after surgery and recovery - so, somewhere closer to February I guess.

Seems very promising DMA !
I do not have access to my regular (filament printer) at present. I will plan to test some options in January

Ellipse wrote: ↑

25 Dec 2024, 00:00

Interesting analysis DMA! What is the best way to measure signal and noise floor? I like to test PCBs and flippers by just placing a flipper on the PCB pad and monitoring changes in the pandrew signal level monitor utility or Rico's Leyden Jar diagnostic utility.

Very interesting analysis Ellipse ! What is the best way (efficient) to order the flippers which would allow dozens of new keyboard projects ? I would like to do a group buy, a real one as the members have done before.

[Injection]

Hi,

Introduction

In parallel to the R&D on 3D printing, I will check options to produce the flippers in injection modding. In order to distinguish the two threads, I will label the posts with [injection] or [3D].

Plastic Fair in German

I have found that some months ago there was a plastic manufacturing exhibition near by my place: https://www.fakuma-messe.de/en/
The next session will be in October 2026.

The good news is that site it contains a list of the participants: https://www.fakuma-messe.de/en/list-of-exhibitors/

--> Injection_Q1: May we collaborative write the specifications for the flippers, and eventually the barrels ?
I am specially interested in determining the plastic to be used.

--> Injection_Q2: May you check if there is any fair near by your ? We may have opportunities in other locations

--> Injection_Q3: If you are in Germany, could you please check whether any of the listed companies are in the near ?

Sourcing in China - [injection-china] ?

I have contacted a company in China. There are many.
As we have the 3D design it should be simpler to get a quote.

--> Injection_Q4: Does anyone has any contact that we can used to progress [injection-china] ?

--------------------------------------------------
the forum members have requested "Model F Labs LLC" to share how to source the flippers from China but they refused. Thus the title of the thread.

You can read the answer here: https://www.deskthority.net/viewtopic.p ... 55#p521855


--------------------------------------------------

ignore

Ignore

Hi,

I am in contact with a company to produce the flippers from China.

Their question is:

Dear, what is the material and quantity of your products?

My answer is:

I would say that I need a first batch of 5000 but additional batches may be needed.
You may quote 3000, 5000 and 8000 and 10000

The material for this part shall be a CONDUCTIVE PLASTIC. .
I believe that plastic with CARBON BLACK will work.

The part is moving to change the capacitance of a capacitor underneath. If the capacitance is not changing it will not work

PLEASE NOTE THAT IF THE QUOTE IS ACCEPTED I WILL SEND YOU A PHYSICAL SAMPLE PART TO BE USED. The model that I sent you may not be 100% accurate.

Regards

Has anyone knows if there is a better material to be used ?

Cheers

Torayca ASHT-18A with 18 percent carbon content