deskthority

Thanks once more !

My comments follow:

wcass wrote: One thing i noticed is that the bottom row (pin 3) does not cross any columns. All of IBM's designs have every row crossing every column exactly one time and I made sure i did that with the XTant, but I'm not sure how much difference this makes or what it does. My best guess is that it is an attempt to make all of the rows/columns as uniform as possible vis-a-vis Cx and Cp.

This was done in purpose. I have taken a look at the documents and it does not state that the crossing must be coherent across rows and columns. I could avoid the crossing on the last row only.

But I will take your advice and route it in a way that it crosses one time each column also.

I think you see what i was talking about earlier; this being difficult to do with IBM's original capacitive pad design due to the number and location of the model M rivets. You might look at some of the other capacitive pad designs that i suggested earlier in this thread that allow the rows to pass through the center of the switch rather than forcing them to always go around. At worse you will have one trace competing with a hole in the space between rows rather than two.

You mean this -> http://deskthority.net/styles/deskthori ... target.gif

Yes, I read it. Very good documents. Lots of information.

The problem that I see with the alternative design is that we do not want to replace the top plate and it is not made out of metal. This means that the keyboard may be affected by noise. I am actually afraid of the 50 Hz (60 in the USA). This is why I tried to use the IBM design. Also because you have demonstrated already that can be done.

We could attach a long flat connector to the xwhatsit controller and design three PCBs (my feeling is that the triangles will work well). It will be more expensive that's the drawback.


I am thinking also to reduce the width of the X and Y traces. It will reduce the noise also.

I will create a new version with these two changes: (1) last row crossings (2) reduced width of X and Y traces


I post the two pages of the documents that you pointed that talk about routing:

Removed due to errors. I will post the new version again
Apologies

----------------
Reminder
----------------

I have created the following spreadsheet to collect the interest in this project.

Click here

Please write down your interest/preferences.

Attached the updated version (0.32) including of the wcass comments.

The design table does not change.

FSSK
FEXM
Correction

These two tracks were running parallel. I have corrected it as follows:

----------------
Reminder
----------------

I have created the following spreadsheet to collect the interest in this project.

Click here

Please write down your interest/preferences.

I just added my interest to the spreadsheet. Not many have volunteered to test the FEXM, so I signed up for that, plus another tested FEXM and one tested FSSK. That should keep me rolling in F goodness for the foreseeable future!

I may have missed where you said this before, but are the F flippers we order on the spreadsheet going to come from Ellipse's newly manufactured F62/F77 project, or another source? Should we just order flippers direct from Ellipse, rather than putting them on the spreadsheet? I guestimate we're about 3-4 months away from being able to have newly manufactured F flippers in hand.

Thanks again!

Techno Trousers wrote: I just added my interest to the spreadsheet. Not many have volunteered to test the FEXM, so I signed up for that, plus another tested FEXM and one tested FSSK. That should keep me rolling in F goodness for the foreseeable future!

Hi. Thanks for signing in for the FEXM. This will reduce my investment cost.

I may have missed where you said this before, but are the F flippers we order on the spreadsheet going to come from Ellipse's newly manufactured F62/F77 project, or another source? Should we just order flippers direct from Ellipse, rather than putting them on the spreadsheet? I guestimate we're about 3-4 months away from being able to have newly manufactured F flippers in hand.

You have missed nothing. I do not know where to find the flippers but from old boards or Ellipse. For the time being I am collecting interest only. I want to know the people that would support the prototype phase once the PCB is ready.

Regarding the flippers, I guess (hope) that Ellipse will, for sure, tell us his source so we can purchase directly
This is what I plan to do, following what xwhatsit did. Without his work, this thread would have never started.
Actually, this is the just the deskthority culture, isn't it ?

all,

I need help again.

The next step is to create a real gerber in a PCB cad. I would like to do it with Kicad for varous reasons:

1) xwhatsit is available in kicad
2) It is GNU
3) it runs in my debian

I have exported the design from the "famous" deltacad into a a "DXF" file and using dxf2svg2kicad converted it into kidcad PCB.
(reference http://mondalaci.github.io/dxf2svg2kicad/) .... but the capacitor PADs are missing.
Does anyone knows a better way of doing this ?

wcass once explained the process that he followed:

I don't use KiCAD much anymore, i use DipTrace. I find it easiest to do almost everything using DeltaCAD then import into PCB software. KiCAD's import functionality sucks where DipTrace's is very good and i find DipTrace's user interface more intuitive. ...

He also offered to do the PCB for us ... but I wanted to try myself.

I would prefer to avoid using DipTrace ....

EDIT - I have no problems to share the DeltaCAD file. If anyone wants to try a good procedure in parallel, PM me of use the attached.

EDIT 2 - I have tried to import the DXF into Diptrace but nothing shows.

Haven't used KiCad, but I can recommend gerbv to visually validate the gerbers.
A google search does return various tutorials on how to export gerbers from Kicad.
Sorry I can't be of more help.

chzel wrote: Haven't used KiCad, but I can recommend gerbv to visually validate the gerbers.
A google search does return various tutorials on how to export gerbers from Kicad.

Thanks. But is only a viewer. It does not read any of the possible DeltaCAD output

Sorry I can't be of more help.

Any idea/help is always well come.

I'm curious as to your aversion to diptrace. It does run on Linux after all.

fwiw, I have a completed schematic for the xwhatsit controller and I will tell you that validating that by hand between kicad and diptrace was hell. I'm about a quarter of the way through routing it on a smaller board swapping out the qfn for soic.

I'm happy to share the files with the disclaimer to assume they're junk until I get my prototype back and tested as working.

idollar wrote: Regarding the flippers, I guess (hope) that Ellipse will, for sure, tell us his source so we can purchase directly
This is what I plan to do, following what xwhatsit did. Without his work, this thread would have never started.
Actually, this is the just the deskthority culture, isn't it ?

That may be optimistic. I don't think Ellipse will feel any obligation to disclose his manufacturer, but instead will treat his work and any of the volunteer stuff he's been asking for as his own property, subject to profit. I'd be pleasantly surprised if he made an exception. Unlike many of us, he's not just in this for fun.

Also, he's said many times that he's doing a single production run. I don't think there is any plan for continued availability for Model F spare parts from him.

Muirium wrote:

idollar wrote: Regarding the flippers, I guess (hope) that Ellipse will, for sure, tell us his source so we can purchase directly
This is what I plan to do, following what xwhatsit did. Without his work, this thread would have never started.
Actually, this is the just the deskthority culture, isn't it ?

That may be optimistic. I don't think Ellipse will feel any obligation to disclose his manufacturer, but instead will treat his work and any of the volunteer stuff he's been asking for as his own property, subject to profit. I'd be pleasantly surprised if he made an exception. Unlike many of us, he's not just in this for fun.

Also, he's said many times that he's doing a single production run. I don't think there is any plan for continued availability for Model F spare parts from him.

So that means no future for model F unless some vendor wants to carry the board right?

Custom parts runs cost a lot of money. Ever since IBM abandoned Model F, it has had no right to a bright future. That's something we have to handle for ourselves.

I think it's quite extraordinary that there is enough interest to do even the one time revival. Can anyone think of another piece of technology that has been out of production for a full 30 years, yet retains enough supporters and new converts to enable a brand new production run?

I'm personally planning to order 300 flippers to cover my FSSK and FEXMs, with a few spares left over. I hope they won't be too expensive!

Yep Techno they should be about 50 cents each flipper with spring in bulk. Same price for the barrels.

I have the new flippers and springs in hand and they have been tested good with capacitive PCBs and will be tested more in the coming weeks.

As I mentioned on my thread, for barrels and flippers to be in continuous availability for the near future, some people would need to order a few hundred or few thousand extras during the two month window where I will be tallying up the order totals, and then be willing to ship out individually to the community.

idollar, orders will be handled through a web site that links to PayPal checkout (PayPal account is not required). The web site is not yet fully finished.

bocah - that's right - for a vendor to carry even just ~100 units it would cost them $35,000, which is a big risk on a project that has attracted about 230 pre-orders in 6 months. I am hoping that Model F's become more popular, but low-volume production is extremely expensive.

And Happy New Year all!

Lets hope somone wants to carry the model F HAPPY NEW YEAR

__red__ wrote: I'm happy to share the files with the disclaimer to assume they're junk until I get my prototype back and tested as working.

1.05" x 2.9".

Now I just need mounting holes of an unknown size...

The next question will be:

"If I want to add that to my PCB then how do I route my rows and columns since you have those resistor packs in the way?

Answer - the back of the PCB is clear all the way up. Enjoy.

__red__ wrote:

__red__ wrote: I'm happy to share the files with the disclaimer to assume they're junk until I get my prototype back and tested as working.

1.05" x 2.9".

Now I just need mounting holes of an unknown size...

The next question will be:

"If I want to add that to my PCB then how do I route my rows and columns since you have those resistor packs in the way?

Answer - the back of the PCB is clear all the way up. Enjoy.

Nice !

But I do not think it will help in the FSSK case ... the space to build the controller is very narrow
(units are inches):
We may need a new layout

And also ... it managed to import the design into Diptrace
We are getting there.

idollar wrote:And also ... it managed to import the design into Diptrace

Screenshot-diptrace.jpg

We are getting there.

Are you wanting to put the design in 0.62"? Remember, you don't want your circuitry inside the metal sandwich.

Mainly because the solder joints won't survive being curved.

The area that I was pointing is mainly flat, but you are correct. I will check if I am correct.

Thanks for the comment !

Attached the latest version of the PCB in JPG from Diptrace, ready for production

TOP
BOTTOM
BOTH with to-layer on top
BOTH with bottom-layer on top
Gerber preview screen dump - TOP
Gerber preview screen dump - BOTTOM
Gerber Files

I will check my files with the recommended manufacturers, I should have the referenced recorded somewhere.
If anyone would like to recommend a site, please do not hesitate

Once I have a quote for the FSSK PCBs, I will contact the people in the "interest list".
Hurry up and register if interested

----------------
Reminder
----------------

I have created the following spreadsheet to collect the interest in this project.

Click here

Please write down your interest/preferences.

Great job i$!
I know it's a bit late for the prototype, but in a next revision maybe you could use thermals on the ground pads. It's a large plane, and might make soldering hard for inexperienced builders with "lightweight" irons.

BTW are the third pair on the bottom row the calibration pads?

chzel wrote: Great job i$!
I know it's a bit late for the prototype, but in a next revision maybe you could use thermals on the ground pads. It's a large plane, and might make soldering hard for inexperienced builders with "lightweight" irons.

Any comment is welcome. It is not late.
It is my first DipTrace project. I just need to know how to do it.

Thanks

BTW are the third pair on the bottom row the calibration pads?

The calibration pad is above the arrows. A little one.

http://www.pcbway.com/
21$ (I have subtracted my 2 prototypes, I hope that everyone would consider it fair) shipped to Europe.
Additional, the shipping from my location to the final, once I test the prototypes to limit the risk.

The above is just the PCB. Nothing more.

idollar wrote:

BTW are the third pair on the bottom row the calibration pads?

The calibration pad is above the arrows. A little one.

So what is that pair doing there? If it's for the extra barrel under the spacebar shouldn't it have the third on the bottom layer? Or I'm misunderstanding something, which is rather common!

You are going to run the FSSK prototypes before the FEXM ones right?
Maybe the other way round is less risky? Since the FEXM is a superset of the FSSK?

Please forgive my ignorance, I see $169 and 21$ in the post above, may I ask which is the final per pcb cost? If it is $21 then I will be sure to add my name to the spreadsheet as soon as I can get a proper computer and not my phone.

Thank you,
neverused

$169 is the cost of 10 PCBs, add to that shipping and expenses, and you get $21 per PCB.