EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: listofopions on April 30, 2018, 01:46:41 am
-
So to start this whirlwind off, let me begin by saying I'm a bit of a keyboard nut (I'm typing this on a model F bigfoot, with a soarer`s converter) so be warned! :P
This is my first post on this forum, so thanks for the platform Dave!
I've begun the design process for an inductively sensed keyboard, the mechanism is partially inspired by the ITW magnetic valve
(https://patents.google.com/patent/US4263582A/en (https://patents.google.com/patent/US4263582A/en) https://www.youtube.com/watch?v=VbFWqH_yOPI&t=570s (https://www.youtube.com/watch?v=VbFWqH_yOPI&t=570s))
which inspires the matrix, and the hp pulse transformer which inspires the sense mechanism
(page 44, https://docs.google.com/open?id=0B4Wyz4bY7XjLLV9hUnB2WGFqck0 (https://docs.google.com/open?id=0B4Wyz4bY7XjLLV9hUnB2WGFqck0))
the hypothetical method of operation is as follows (I wish I had more work in EE, as I'm a mechatronics engineer,
so please tell me if this is sensible, and what changes need to be made):
the controller selects one of 8 columns to ground, and selects one of 16
rows to power with the pwm; both via a 74hct4051 multiplexer. the pwm
cycle is one pwm signal and one inverted (net ac current of 5v max per the
voltage dividers on the pwm pins), and is begun shortly after the row and
column are selected. if there are keys depressed in the selected column
there should be an inductive short on the transformers and therefore a
higher voltage on the rows of the column not selected by the row
multiplexer. this voltage is sensed by comparitors that are biased by a
known set-point that is established for the comparitors via a windowed
dac, and resistors R1, R2, and R3.
the attached kicad schematics should help (I hope) with the explanation.
anyway, my main problem at the moment is the design of the sense transformers for the keypad (under the assumption that matrix sense mechanism is sound)
I am assuming that I can use the design from the HP pulse transformer keypad, which is double spiral transformer (the other end of the coils is via-ed in the center
to the other side of the pcb) I've got the fast field solvers loaded and installed on my computer and I have some idea of how to use them (thank god for
documentation!) but I would appreciate, and desperately need some guidance on the design of the sense transformers, like the entire design!
and if that's not the main problem with this design let me know, any advice or input is much appreciated!
-
Oh I should also mention that I need to have some help with layout! I am assuming that for the parts used I can just simply use the datasheets suggested layout.
I am a little worried about the way I have the AGND, and GND coupled through a capacitor, as that would cause the grounds to a different potential, but because this is
an inductive load that the multiplexers are driving, I think I would need to separate the driven inductors from the rest of the logic. I've also gone ahead and attached a
pdf of the kicad schematic.