Alt-BEAM Archive

Message #05012



To: Wilf Rigter Wilf.Rigter@Powertech.bc.ca
From: Dave Hrynkiw dave@solarbotics.com
Date: Fri, 02 Jul 1999 18:41:20 -0600
Subject: [alt-beam] Unicore ideas/ideals --> was: Ideal Unicore


At 07:01 PM 12/06/99 , Wilf Rigter wrote:
>So it appears as if using the differential dc component of high frequency
>Pcore output duty cycle does not provide a corresponding linear change in
>Mcore duty cycle. Instead the range of interaction was narrow and extremely
>non-linear. I am beginning to see that I jumped to some conclusions which
>although wrong could be designed that way with a modification of the Unicore
>circuit and would perform better than the original Unicore.
>
>Any comments?

Yes, YES! You are a madman, and will kill us all with your technology!!! ;>

This is an interesting twist on the online evaluation of the Unicore.

Onto your next message in the chain...:
> It appears as if in BRIGHT light the average dc component of high
>frequency Pcore output which corresponds to the Pcore duty cycle does not
>produce the hoped for corresponding linear change in Mcore duty cycle
>without affecting the Mcore frequency. Instead the range of interaction was
>narrow, non-linear, greatly affects frequency and rapidly converged on the
>Mcore locking up in one state or the other.

I've been thinking about this a bit lately. My experience playing with head
circuitry left me frustrated with tuning issues at the extremes of light
exposure. What would work very well in bright light wouldn't work very well
at low-levels, with the reverse being true as well. The problems in using
the PD's to directly control the oscillation frequency got me thinking
about our old friend, the PhotoPopper.

The following is just a thought experiment (as I am currently travelling
and can't confirm my ideas), so caveat emptor (ie: take it for what it's
worth). At this point, I wish there was some sort of easy-to-manipulate
sketching program everybody could access. Ever use Visio? Something like
that. Back to the task at hand...

Instead of using just PD's to influence the following Mcore, use a pair of
PP eyes, being a photodiode feeding a 1381 & cap, just like the
PhotoPopper. These PP eyes are simultaneously activated, being fed Vcc (an
enable of sorts, I suppose) from another external oscillator source. Use
the output of the "winning" PPE to charge an Nu on the Mcore AND tie the
other PPE low (to keep it from finishing it's charge cycle) AND enabling
the motor driver chip.

The PhotoPopper is very good at tracking light at all light levels, where
using the traditional back-to-back PD's is good only for a much more
limited range of light. Using a circuit that amplifies a difference rather
than changing a frequency seems more ideal in this situation.

This concept is not completely thought out yet, but I thought I would throw
it out there for what it is worth to keep the momentum going.

>This maybe the kind of challenge needed to evolve the original Unicore
>design ( if this hasn't already been addressed) and begs the question: What
>might constitute the behaviour of a IDEAL BEAMant?

Good question. I'm not sure if it needs a precise answer, as I think a
slow, deliberate low-light BEAMant would nicely complement a much quicker,
nocturnal sibling. But if I had to set some criteria down, it would be this:

1a) Lots of light = lots of activity (not nocturnal)
1b) Low Light = Switchable setting from long charge & bursts of motion to
Photopopper action (move when there's power to take 1 step).

2) Motion - both motors turning at same time, although I KNOW one at a time
is much more realistic to expect and results in longer run times.

3) Back up behavior -
touched on left, SPIN to right (both motors CCW)
touched both, reverse with a shorter time-out on one side, perhaps
PD-influenced

4) Light Seeking behavior
Given point 2, smooth arcs.
Given point 2a (alternate motors), pivot on one stationary wheel

Of course, this is just a short list. Add on behaviors of a nocturnal
setting (charge all day, run only when dark), basic communication
(modulated LED) and a living chain/snake (see Beckingham's Photovore) would
all be worthwhile goals. And lets not forget a PIC/uC for advanced functions!

That's my brain-dump for the day...

Regards,
Dave
---------------------------------------------------------------
"Um, no - that's H,R,Y,N,K,I,W. No, not K,I,U,U, K,I,_W_. Yes,
that's right. Yes, I know it looks like "HOCKYRINK." Yup, only
2 vowels. Pronounciation? _SMITH_".
http://www.solarbotics.com

------------------------------------------------------------------------

eGroups.com home: http://www.egroups.com/group/alt-beam
http://www.egroups.com - Simplifying group communications



Home