Some ideas for the accelerometer

[Kyle Bassett]

I think the idea is great... I've designed a few control systems using
accelerometers to time the ejection of parachutes on MPRs/HPRs (Medium/High
Power Rockets).  The biggest issue is how to interpret the data.

One of the largest problems we've had to overcome with accelerometers is
averaging the input data over the right time interval.  Too short of an
interval and the jitter effect drowns out any useful data. Too long, and
detecting any useful pulses is attenuated.  An algorithm that varied that
time interval would probably prove most useful.

I am particularly interested in the scrolling function, possibly using a
[tap].  You could [tap] right above the top of the lcd module or right below
it, and depending on the program, it would scroll up or down.  This could
easily be picked up using the accelerometers since the phone is so light, an
acceleration "peak" is easily attained (and I'm only thinking 2D here,
there's a huge amount of information coming in with 2x3D accelerometers).
This would even work while walking or running, because those would register
as "curves" more than a specific peak.  (The algorithm could use the GPS
data to detect walking and attenuate accordingly.)

Do we know the exact accelerometer being used in GTA02 yet?  I could do a
comparison using the accelerometers I have lying around.

@Dean Collins, The IBM Hard Drive knock application is an excellent example
of using this type of input.


-Kyle



On 10/15/07, Alexey Feldgendler <alexey at feldgendler.ru> wrote:
>
> On Sun, 14 Oct 2007 21:57:52 +0200, Ortwin Regel <ortwin at gmail.com> wrote:
>
> > It's also important to remember that the motion of picking up your phone
> > should not lead to denial of the call... ;)
>
> The initial proposal mentioned muting the ringer, not denying the phone.
> It's perfectly OK to mute the ringer if you're already taking the phone to
> your ear.
>
>
> --
> Alexey Feldgendler <alexey at feldgendler.ru>
> [ICQ: 115226275] http://feldgendler.livejournal.com
>
> _______________________________________________
> OpenMoko community mailing list
> community at lists.openmoko.org
> http://lists.openmoko.org/mailman/listinfo/community
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openmoko.org/pipermail/community/attachments/20071015/57947663/attachment.htm