[sdiy] Software problem hardware solution? + US patents

[Byron G. Jacquot]

>I am using a cheap MIDI keyboard (Fatar 610) and a Roland sound module
>(JV-1010) and have the daunting prospect of getting into editing
>parameters in the sound module using a PC to achieve one very simple
>thing: I want to reduce the overall velocity sensitivity of the keyboard
>so that I don't have to bash away at the keys like King Kong to get an
>even tone when Im practicing/working stuff out. I haven't looked inside
>the keyboard but what 'element' determines the velocity of a key struck
>and shouldn't I be able to amplify it in an analogue way before it goes
>digital to fool the MIDI generator into producing 'louder' notes?
>Alternatively there may be a box someone has designed with a MIDI in/out
>socket and a pot that can do just this task. Simple enough but quite
>handy don't you think? Would love to build or buy if there's a schematic
>on the web.

The sensing of velocity is usually done in the digital domain: there are two
switches per key.  As you press a key, one of the switches closes just a
little before the other, and the microcontroller times the difference, and
turns that into a velocity value.  Any hacking on the keyboard end would
involve some firmware changes, or possibly modifying the mechanism of the
two switches.

There used to be some little boxes from Anatek, I think it was, that did
various MIDI utilities, and velocity mapping was probably among
them...probably something to get around the DX7's max of 100.  A
microcontroller and about 50 lines of assembly code could do it, as well.

I'd be surprised of the JV can't be tweaked to respond the way you want, as
well.

>Secondly I'd like to try and get a US patent or at least 'lodge' one.
>However I don't have the bucks to employ a patent lawyer. Now I don't
>live in the US but I know many of you have experience with this and
>remember hearing someone say that it's not that difficult to obtain,
>although defending it may be the hard part. I have the UK patent for my
>idea. Are there any sites or info to help me do this? Official sites
>would be OK (though I can find these myself) but more inspiring would be
>those aimed at the DIY'er and first timer, or even anecdotal advice. You
>can email offlist if this seems OT.

I may have been one of the folks who gave you the impression that gettuing a
US patent is easy, but that was assuming you'd be using a patent lawyer.
It's more or less a matter of properly documenting what you have.  The
history check & referencing of prior art is very important, and I have no
idea how a DIYer would streamline the process.  I've seen both sides of the
patent defense process, and demonstrating how you fit into the overall
scheme of things will be important, should you need to defend your patent.
If there are flaws in your documentation, you could have problems upholding
your patent.  

Not that a determined individual couldn't do it...a good library and/or web
search would probably be a good start.

Byron Jacquot