Alt-BEAM Archive
Message #06112
To: beam@sgiblab.sgi.com
From: Richard Piotter richfile@rconnect.com
Date: Thu, 16 Sep 1999 19:43:04 -0500
Subject: [alt-beam] Re: is beam dying?
NOPE!!!
First off, BEAM is alive and well. It just is in a rut. I've seen them
before, and then something new comes along and everyone get's all
excited, and then they get over it and sudenly everyone starts
questioning BEAM technology again. It's a cycle, don't worry about it.
I've actualy seen it LESS busy!!!
Second, the posts about MP3 Players are off topic, but unfortunately,
now and then, little things like that pop in here or there.
As for convergence and load induced feedback... You NEED efficient
motors and close link between the motor and Nv for it to have any
noticable effect. I actualy saw a rather ingenious device. I'm not going
to talk about it in detail, cause it's still being tuned and worked on,
and the builder asked to keep it's specifics under wraps till he's done,
but let's just say that it is an AMAZING demonstration of load induced
feedback, and I had the great oportunity to see the device! Like I said
though, it requires an efficient motor and little buffering between the
Motor and Nv. I hope he has finished it! If he has finished, I'll have a
chance to see it soon! :)
If you can't afford efficient motors, you can bias the Nvs using
potentiometers attached mechanicaly to the motors. If you are using
servos ($10 from Tower Hobbies and many other places), you will get it
with a potentiometer built in, but it's value is rather small, and
doesn't do much to bias Nvs. Primarily, I use that method to shorten the
process driving THAT motor. It acts as a dynamic electronic leg stop. If
you are willing to modify the pot, you can cut a section out of the
middle and use it as a process shortener.
James G Watt wrote:
>
> In article <19990916230750.11028.rocketmail@web124.yahoomail.com>, Evan
> Dudzik writes
> >is beam dying?
>
> I hope not because I just found out about it a few week ago.
>
> I got some bits to make a simple microcore based walker but I'm having a
> little trouble with it.
>
> I've built myself a 4 neuron microcore on a breadboard (and a freeform
> version) and I thought it was performing correctly. The microcore uses a
> 74hc14 and I'm using a 74als245 to drive my motors. It all appears to be
> firing correctly except that as far as I can tell there's absolutely NO
> feedback changing the behaviour of the microcore what so ever. I was
> under the impression that the timing was supposed to vary depending on
> the load on the motors? Did I misunderstand something because my circuit
> appears to run as regular as a metronome no matter what load is attached
> to it?
>
> I built the thing following the instructions here:
>
> http://www.xs4all.nl/~vsim/amiller/microcore.html
> --
> James G Watt
>
> http://www.kinetic-arts.demon.co.uk
> Part of the Video Arcade Game Collectors WebRing.
>
> You are in a maze of twisty little passages, all alike.
--
Richard Piotter
richfile@rconnect.com
The Richfiles Robotics & TI web page:
http://richfiles.calc.org
For the BEAM Robotics list:
BEAM Robotics Tek FAQ
http://people.ne.mediaone.net/bushbo/beam/FAQ.html
6113 Thu, 16 Sep 1999 18:06:48 -0700 [alt-beam] Re: power_smart_head fest "'beam@sgiblab.sgi.com'" Wilf Rigter Very impressive presentation of a non-starter circuit.
Well we must do something about that!
I will duplicate the protoboard layout precisely and report back with some
fine adjustments but in the mean time experiment a little!
Going from memory and my reasonably good understanding of the circuit, the
simplest way to get some action is to remove the 0.01 cap closest to the
motor. That will turn the circuit into a monocore head with the motor
buzzing but not moving when both LDRs are lit and turning left or right with
a speed proportional to the LDR imbalance. Very similar to the Bicore head
but unlike the Bicore, no diodes are needed when using LDRs (note to Rob!)
Next put the 0.01 cap back and remove 5.6M. The circuit will still "snooze"
when the LDRs are carefully balanced but will now turn continuously left or
right when unbalanced. The final adjustment is to strike a compromise
between these two behaviours and have it snooze when balanced, turn slowly
when slightly unbalanced and turn fast when greatly unbalanced.
BTW The 1K resistors are optional when using medium resistance LDRs (like
the ones you are using which look like the current model of Radio Shack
LDRs). So for most applications, the circuit can be simplified by removing
the 1K resistors and connecting the LDRs directly across the 5V supply.
More to come....
regards
Wilf Rigter mailto:wilf.rigter@powertech.bc.ca
tel: (604)590-7493
fax: (604)590-3411
> -----Original Message-----
> From: Scott Burns [SMTP:s-burns@uiuc.edu]
> Sent: Thursday, September 16, 1999 9:16 AM
> To: beam@sgiblab.sgi.com
> Subject: RE: power_smart_head fest
>
> At 06:12 PM 9/10/99 -0700, you wrote:
> >Hello Scott,
> >
> >Tell us more: what components? which schematic? what motor? matched LDRs?
> >
> >best regards
> >
> > Wilf Rigter mailto:wilf.rigter@powertech.bc.ca
> >
>
> OK, I've described it in detail (including scope readings) at
>
> http://www.designbyalgorithm.com/circuit.htm
>
> > > -----Original Message-----
> > > From: Scott Burns [SMTP:s-burns@uiuc.edu]
> > > Sent: Friday, September 10, 1999 7:05 AM
> > > To: beam@sgiblab.sgi.com
> > > Subject: Re: power_smart_head fest
> > >
> > > Dear Wilf:
> > >
> > > The virtual beam workshop sounds great. I tried building the PSH and
> was
> > > able to get the square waves to the motor to change duty cycle with
> > > changes
> > > in relative LDR illumination, but they were always in phase. The motor
> > > never saw a voltage differential across it. I'm anxious to figure out
> what
> > >
> > > I did wrong.
> > >
> > > Regards, Scott
> > >
> > >
>
> ------------------------------------------
> Prof. Scott A. Burns
> Dept of General Engineering and Civil Engineering
> University of Illinois at Urbana-Champaign
> 104 S. Mathews, Urbana, IL 61801
> Phone 217/333-1618, FAX 217/244-5705
6114 Thu, 16 Sep 1999 18:59:55 -0700 [alt-beam] Re: So...what did I miss? "Hendrickson"
> So what did I miss? A friend told me a new extreme machine episode about
BEAM
> aired yesterday on TLC. I checked the website and it should be rerun on
> Saturday.
>
It was a rerun of an episode that aired a few months ago (2 maybe). It's a
cool show (I have it taped).
-Brian (beamboter)
P.S. No I haven't fallen off the face of the earth- just got busy with other
things.
Home