Re: IBM FSSK Build Log
Posted: 24 Jun 2021, 09:27
The question is: do I floss the grease?
You're making it sound like this isn't something we've already tried
If anything, I'd say the reverse is true. You have to remember that Euler buckling is very different from ordinary Hookean spring compression. But I'd say that, overall, the feel changes less than you'd think.Go-Kart wrote: 24 Jun 2021, 06:40 I could imagine the floss mod increasing the force required for the tactile event ever so slightly.
You do the intellectual heavy-lifting, so we don't have tooBjerrk wrote: 24 Jun 2021, 10:15 Of course, if you dip the springs in glowstick juice, they'll get even lighter!
... I'll see myself out ...
I do not thing that grounding is a real issue if the keyboard is mounted the right way.pyrelink wrote: 09 Dec 2021, 20:43 Would an integrated controller help with the grounding situation at all? I feel like grounding has always been one of the bigger struggles with getting proper key registrations and no interference. It would definitely simplify the installation process though, and I imagine with proper placement we could reuse the existing connector hole in the case.
Hey i$!idollar wrote: 09 Dec 2021, 19:07 So following the guide that you sent me (thank you pyrelink) I managed to compile and install QMK in my two QMKs.
It has not been an easy process. To start with, the keyboard definition is not equal to the one that is distributed with the sources. I cabled the controllers differently. I used an in-between connector to ensure that I would work better with my prototypes.
I had to remap the rows and columns ... twice. The two FSSKs which I have use different PCB versions. When I designed the FSSK I ordered a first set of two PCBs. It turned out that the design missed a connection which I could patch (literally) with a wire. The second version solved this issue and added additional keys that could be used for the future. It was also used to create the fext
Anyhow, after some effort, I can tell you that I am typing with the prototype this post. It works, I believe, much better.
I have two possible concerns with that:As I said, if there is interest, we could also design a PCB with the controller integrated.
For the FSSK we should really support Unicomp's new Mini M, and the additional bottom row keys.added additional keys that could be used for the future.
The issue is mine, your software is correct.For model Fs the simplified assumption is that either a ribbon cable is used, where wires can't cross each other (or that the user takes care to keep the ordering of the wires). This works out well with most model Fs, cause there's only one way to wire then correctly. Some times people accidentally cross two wires when soldering,
Code: Select all
#define LAYOUT_all( \
k_2_2, k_2_1, k_2_4, k_2_3, k_1_3, k_2_6, k_2_8, k_1_8, k_2_7, k_2_5, k_2_10, k_1_9, k_2_9, k_2_12, k_1_11, k_2_11, \
\
k_1_2, k_1_1, k_4_1, k_1_4, k_4_4, k_4_3, k_1_6, k_4_6, k_3_8, k_4_7, k_1_5, k_4_5, k_1_10, k_4_10, k_4_9, k_1_12, k_4_12, k_4_11, \
k_3_2, k_6_1, k_3_1, k_3_4, k_3_3, k_6_6, k_3_6, k_6_8, k_3_7, k_6_7, k_3_5, k_3_10, k_6_10, k_3_9, k_3_12, k_6_12, k_3_11, \
k_5_2, k_6_2, k_5_1, k_5_4, k_5_3, k_6_3, k_5_6, k_5_8, k_5_7, k_5_5, k_6_5, k_5_10, k_8_10, k_5_9, \
k_8_2, k_7_1, k_8_1, k_8_4, k_7_3, k_8_3, k_8_6, k_7_6, k_8_8, k_8_7, k_7_5, k_8_5, k_7_10, k_8_9, k_8_12, \
k_7_2, k_7_4, k_7_8, k_7_7, k_7_9, k_7_12, k_8_11, k_7_11 \
) \
{ \
{ k_1_1, k_1_2, k_1_3, k_1_4, k_1_5, k_1_6, KC_NO, k_1_8, k_1_9, k_1_10, k_1_11, k_1_12 }, \
{ k_2_1, k_2_2, k_2_3, k_2_4, k_2_5, k_2_6, k_2_7, k_2_8, k_2_9, k_2_10, k_2_11, k_2_12 }, \
{ k_3_1, k_3_2, k_3_3, k_3_4, k_3_5, k_3_6, k_3_7, k_3_8, k_3_9, k_3_10, k_3_11, k_3_12 }, \
{ k_4_1, KC_NO, k_4_3, k_4_4, k_4_5, k_4_6, k_4_7, KC_NO, k_4_9, k_4_10, k_4_11, k_4_12 }, \
{ k_5_1, k_5_2, k_5_3, k_5_4, k_5_5, k_5_6, k_5_7, k_5_8, k_5_9, k_5_10, KC_NO, KC_NO }, \
{ k_6_1, k_6_2, k_6_3, KC_NO, k_6_5, k_6_6, k_6_7, k_6_8, KC_NO, k_6_10, KC_NO , k_6_12 }, \
{ KC_NO, k_7_2, k_7_3, k_7_4, k_7_5, k_7_6, k_7_7, k_7_8, k_7_9, k_7_10, k_7_11, k_7_12 }, \
{ k_8_1, k_8_2, k_8_3, k_8_4, k_8_5, k_8_6, k_8_7, k_8_8, k_8_9, KC_NO, k_8_11, k_8_12 } \
}
Code: Select all
#define LAYOUT_all( \
k_2_2, k_2_1, k_2_4, k_2_3, k_1_3, k_2_6, k_2_5, k_1_5, k_2_8, k_2_7, k_2_10, k_1_9, k_2_9, k_2_12, k_1_11, k_2_11, \
\
k_1_2, k_1_1, k_4_1, k_1_4, k_4_4, k_4_3, k_1_6, k_4_6, k_3_5, k_4_8, k_1_7, k_4_7, k_1_10, k_4_10, k_4_9, k_1_12, k_4_12, k_4_11, \
k_3_2, k_6_1, k_3_1, k_3_4, k_3_3, k_6_6, k_3_6, k_6_5, k_3_8, k_6_8, k_3_7, k_3_10, k_6_10, k_3_9, k_3_12, k_6_12, k_3_11, \
k_5_2, k_6_2, k_5_1, k_5_4, k_5_3, k_6_3, k_5_6, k_5_5, k_5_8, k_5_7, k_6_7, k_5_10, k_8_10, k_5_9, \
k_8_2, k_7_1, k_8_1, k_8_4, k_7_3, k_8_3, k_8_6, k_7_6, k_8_5, k_8_8, k_7_7, k_8_7, k_7_10, k_8_9, k_8_12, \
k_7_2, k_7_4, k_7_5, k_7_8, k_7_9, k_7_12, k_8_11, k_7_11 \
) \
{ \
{ k_1_1, k_1_2, k_1_3, k_1_4, k_1_5, k_1_6, k_1_7, KC_NO, k_1_9, k_1_10, k_1_11, k_1_12 }, \
{ k_2_1, k_2_2, k_2_3, k_2_4, k_2_5, k_2_6, k_2_7, k_2_8, k_2_9, k_2_10, k_2_11, k_2_12 }, \
{ k_3_1, k_3_2, k_3_3, k_3_4, k_3_5, k_3_6, k_3_7, k_3_8, k_3_9, k_3_10, k_3_11, k_3_12 }, \
{ k_4_1, KC_NO, k_4_3, k_4_4, KC_NO, k_4_6, k_4_7, k_4_8, k_4_9, k_4_10, k_4_11, k_4_12 }, \
{ k_5_1, k_5_2, k_5_3, k_5_4, k_5_5, k_5_6, k_5_7, k_5_8, k_5_9, k_5_10, KC_NO, KC_NO }, \
{ k_6_1, k_6_2, k_6_3, KC_NO, k_6_5, k_6_6, k_6_7, k_6_8, KC_NO, k_6_10, KC_NO , k_6_12 }, \
{ KC_NO, k_7_2, k_7_3, k_7_4, k_7_5, k_7_6, k_7_7, k_7_8, k_7_9, k_7_10, k_7_11, k_7_12 }, \
{ k_8_1, k_8_2, k_8_3, k_8_4, k_8_5, k_8_6, k_8_7, k_8_8, k_8_9, KC_NO, k_8_11, k_8_12 } \
}
There's also leftshift+rightshift+B See here for the feature https://beta.docs.qmk.fm/using-qmk/adva ... re_commandidollar wrote: 10 Dec 2021, 08:58 I have found very useful to assign the "reset" command to a layer.
What you are commenting and the proposed solution makes lot of sense.I have one concern regarding the FSSK/FEXT project. The currently released files (at least the ones from here) don't have a license attached. Assuming you're the only contributor, could you release them under some open source license? For example a version of Cern OHL v2?
Yes, an integrated controller would be great. Could even implement reed reset support for disassembly-free reflashing. As for the hole in the case, I think it would be cool to be able to use the original detachable cable over SDL. Making a daughterboard that connects to the controller would be a relatively simple process. Just something to think about!idollar wrote: 09 Dec 2021, 19:07
idea: it would be nice to update the PCB with an integrated controller. This will allow to ease the building process. Any interest ?
promise: When the original controller is removed and replaced, the usb cable goes through a big hole. I always thought that I could design a 3d-print a piece to make it work as it should. I promise that I will post here the source to print this part once I design it.
And looking at the part that I designed here:headphone_jack wrote: 10 Dec 2021, 15:27
Yes, an integrated controller would be great. Could even implement reed reset support for aqadisassembly-free reflashing. As for the hole in the case, I think it would be cool to be able to use the original detachable cable over SDL. Making a daughterboard that connects to the controller would be a relatively simple process. Just something to think about!
My approach is to focus on the functional requirements first and ensure that the "cosmetics" are possible at a later stage. If we try to do everything we may end up achieving nothingheadphone_jack wrote: 10 Dec 2021, 15:27
Yes, an integrated controller would be great. Could even implement reed reset support for disassembly-free reflashing. As for the hole in the case, I think it would be cool to be able to use the original detachable cable over SDL. Making a daughterboard that connects to the controller would be a relatively simple process. Just something to think about!