Generic Open Source / Open Hardware design
Red Davies
Greetings,
It seems my keyboard design has been loved by some and "if only it would..." by others. That's cool, I'm here to help.
In conjunction with the wiki, maybe we could make a list of the requirements for a generic keyboard design. I'm willing to put some time in to organize them and discuss how we go about making it happen. Let's discuss on the list and I'll go ahead and start documenting it on the wiki.
I would like to make each of the requirements as modular as possible so people can "pick and choose" those items that they want. For example, making a keyboard touch sensitive may increase the price of the hardware by 75% or more - for those that want it, there's an option for that.
So, as generic as possible, please to be listing what you'd want in an "ideal keyboard" and we'll try as a group to rationalize them down.
Thanks,
Red
metasonix
First, let me mention this Atlantic article that ran last week,
it was barely noticed by anyone in the music gear business.
It deals with alternate controllers.
Second: if you want a key sensor with two degrees of analog
sensing freedom, here is an idea that can be implemented
with off-the-shelf hardware and (probably) minimal cost.
Take one of those tiny Alps video-game joysticks.
http://www.mouser.com/ProductDetail/ALPS/RKJXK122000D/?qs=sGAEpiMZZMsFLoxuNbFfpcZ5HqBisj%2fn5t3JMYiSq6I%3d
Mount it sideways, and attach an arm that pokes upwards,
with a key on the end, to the stick end. It could be as simple as
a metal rod with a hole and setscrew on one end, to attach to
the stick, and a key glued to the other end.
So, when you push down on the key, the resistance changes
on one of the built-in pots, and if you push the key sideways,
the other built-in pot changes.
Sense the resistance changes as needed--an analog circuit can
derive keydown, pressure, and velocity in both planes of motion
for a single key. This wold have an advantage in that it would
be faster than microprocessor scanning.
Or a microprocessor can scan a bunch of joysticks at once and
derive keydown, pressure and velocity. A fast processor with
loads of analog-voltage inputs would be a necessity.
There's probably no cheaper and easier way to do this, at least
that I have found. Mechanical design is relatively simple, and a
large number of these devices can be crammed into a small
keyboard because of the small size and PCB mountability of the
joysticks. The joysticks are 22.7mm wide, but you could cram
them into a smaller space by offsetting them alternately.
The joysticks come in two forms, one with just two pots and
another with two pots and a tact switch that is activated by
pressing the stick down. This keyboard would use the version
without the tact switch. (Or, if you prefer, you could use the
tact-switch version, mount it in the normal way, and put a
key on the end of the stick. Then you push down for a simple
trigger, sideways for one analog change, and forward-back for
another analog change.)
If anyone implements this, don't forget to credit me......
--
E. A. Barbour
METASONIX
www.metasonix.com
Ken Rushton
Ken.
--
You are subscribed to the Google Groups "diykeyboard" group.
To post to this group, send email to diyke...@googlegroups.com
To unsubscribe from this group, send email to diykeyboard...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/diykeyboard?hl=en
Also see the DIY Keyboard Google Code Repository and Wiki: http://code.google.com/p/diykeyboard/
Red Davies
The joysticks come in two forms, one with just two pots and
another with two pots and a tact switch that is activated by
pressing the stick down. This keyboard would use the version
without the tact switch. (Or, if you prefer, you could use the
tact-switch version, mount it in the normal way, and put a
key on the end of the stick. Then you push down for a simple
trigger, sideways for one analog change, and forward-back for
another analog change.)
I've seen these before and I love and adore the idea but the cost of those analogue joysticks is insane. They're over $6 each. I even looked at simple "switched" joysticks and they are likewise too expensive.
Of course, that's too expensive for *me*. If someone really wants to build it I will provide the hardware and software support for it.
I would love to see what lives under the eigenharp keys :-) I bet it's a joy to behold.
Kind Regards,
Red
Red Davies
2. about one mm more travel finese on the keys - this would allow it to approximate the dynamics of a piano
Can you measure the travel for me now or tell me how much travel you'd ideally want.
3. how about making it so that two could be joined together, tile-wise. Then we could make them as wide as we want?
That is definitely a base requirement for me. Your instrument wants to be narrow and tall. Mine wants to be short and wide. We should be able to just tile the keys appropriately.
Thanks,
Red
metasonix
> I've seen these before and I love and adore the idea but the cost of
> those analogue joysticks is insane. They're over $6 each.
a) in quantities, the price drops. And I suspect many people would
buy quantities of them, to build a large keyboard controller.
If any of you wants to pursue this further, I am willing to pool
and make a large purchase.
b) by the time you've fabricated a photosensor plus specialized mechanical
parts, or a Hall-effect version plus mechanicals, or something with
multiple
electrical contacts, you're rapidly approaching the cost of the joystick.
And none of those approaches give you TWO degrees of freedom, unless you
doubt everything up, and greatly complicate the mechanical design.
c) this is all off-shelf components, except the PC board (easy), the
key-lever
assembly (simple, could be made in quantity at a CNC machine shop
for a few dollars per unit), and the cabinetry. It is mechanically simple
and requires no resin molding or other tricky methods. The joystick is auto
self-centering--no added springs needed.
d) using a Hall sensor, multiple contacts or a photosensor would require
either a LOT of processor power/time, or an analog circuit for each key.
So, the electronic complexity required would be about the same.
e) Do you have a more brilliant idea that can be implemented for less,
at hobbyist part quantities? Hm? Let's hear it.
metasonix
Oops, make that "double"
Ken Rushton
--
Andrew Wagner
mechanical problem, and over-estimating the difficulty and expense of
the electronics problem (for the optical sensor route).
For those joysticks, even if you had a group buy of 6 keyboards,
you're still looking at 3.60 euros per key just for the joysticks.
And then you have double the analog multiplexing and A/D conversion
that you have for the single reflective-object-sensor-per-key design.
Also hiring a professional machinist to do manual labor on each of
your keys individually is NOT going to be cheap. You'll almost
certainly pay more for labor than for materials.
If you want to build something totally new and have the cash and the
time, by all means go for it, but I think your design will be too
expensive for most hobbyists.
BTW, Prof. Haken, maker of the continuum fingerboard:
http://www.hakenaudio.com/Continuum/
was in my department back at UIUC. You get 3 analog degrees of
freedom (actual travel, not just pressure) per finger on that
instrument. Under a layer of neoprene, it uses an array of fine rods
(several per finger) with springs and hall effect sensors on each end.
The most expensive part of the design turned out to be having the
little ends of the rods machined. The hardest part of the electronics
was keeping the darn thing in tune; apparently the hall effect sensors
have quite a bit of temperature dependence. You have to give it some
time to warm up and autocalibrate each time you turn it on.
They actually had a concert where it was hooked up to two huge
"musical" tesla coils on the engineering quad. That thing is hard
enough to play under ideal conditions, I have no idea how the guy
managed it in the sub-zero temperatures we had that evening!
Cheers,
Drew
ClarkBattle
On Feb 13, 2:56 pm, metasonix <sy...@metasonix.com> wrote:
> Gentlemen:
>
> First, let me mention this Atlantic article that ran last week,
> it was barely noticed by anyone in the music gear business.
> It deals with alternate controllers.
>
>
>
> E. A. Barbour
> METASONIXwww.metasonix.com
dominique.waller
Samchillian, something that amazed me. See
http://video.google.com/videoplay?docid=4889538176718826165&q=samchillian#
This instrument is not specifically isomorphic, but can function nearly as
such, if the chromatic scale is programmed as the default scale. I
undertsood that from watching this link :
http://www.youtube.com/watch?v=L8BXMMQZtYE
Dominique Waller
--