Alt-BEAM Archive
Message #01117
To: Wilf Rigter Wilf.Rigter@powertech.bc.ca
From: Steven Bolt sbolt@xs4all.nl
Date: Mon, 1 Mar 1999 20:19:08 +0100 (CET)
Subject: [alt-beam] Re: New photovore circuit
On Mon, 1 Mar 1999, Wilf Rigter wrote:
> Hello Steven,
>
> Your resistor on one side will certainly unbalance the midpoint of the photo
> diodes if the leakage currents are as high as Andrew Miller suggested.
> Here's a "new" (and use the word cautiously) improved current limited and
> photodiode balancing version of my photovore.You may experiment with pot
> values of 100K -10M.
>
> Sorry bout the credits: I humbly bow to all those who came before me.
You said "(I think) unique," and you're not required to do an
exhaustive search before showing a piece of work :)
It's just that the similarity struck me as obvious. And checking
the date... July 1996. So Andrew beat me to it by a month (unknown
by me). But the use of clocks/watches and the way my photovores
back up without reversing their motors *really* is mine - I hope :)
And I think Andrew missed the point that you're using LEDs as photo
diodes. That also isn't a new idea, but afaik large leakage
currents aren't the problem if you do that - rather the reverse. In
average light, the currents are so small that you'll be dealing
with an effective input resistance in the tens of mega-ohms. That
is likely to make your circuit excessively sensitive to all kinds
of things apart from light; like contamination of the pcb, and even
EM noise radiated in by the environment. That's the main reason why
I use BPW41s. They pull that input hi or lo rather more positively.
As to Andrew's point with regard to `real' photo diodes, yes, the
light intensity range of two BPW41s in series with a resistor is
limited. My clock-motored photovores don't care, as they are
indoor/shadow critters. But SunEater_III had to like the full sun
and you may notice that the series resistor is only 2K7 there, and
the sensor is pulsed to avoid excessive current consumption.
Moreover, the `eyes' of SunEater_III look about 45 degrees down.
This has two pleasant effects:
- It makes the bot try to avoid shadow patches, but if the light
straight ahead is really juicy, it will cross a shadow to get
closer.
- It limits the maximum intensity of light reaching the `eyes' to
something the BPW41/2K7 combination handles very well, even in
full sunlight.
Finally, the midpoint does not get noticeably out of balance until
the light is already too strong for the BPW41/resistor combination.
And even then, the way the photo diodes are mounted - on opposite
sides of the `bot - reduces the effect to a slightly curved
approach. These illustrations & pages:
Linkname: SunEater_II chassis
URL: http://www.xs4all.nl/~sbolt/Other/S_IIIchassis.gif
Linkname: Photovore - search the light, avoid obstacles
URL: http://www.xs4all.nl/~sbolt/e-fotovoor_beweging.html
Help illustrate what I mean.
And don't mind Andrew and me too much. We've been at it too long
and sometimes think we've seen it all, but your use of LEDs did
make me think again about the theoretical arguments against them.
Do have a go and put that circuit on a `bot. The high input-R
problem may not be so bad after all. And perhaps it will help you
to know what trouble you might run into.
Assuming you do try, I don't think that pot in the center will have
much effect. My advice would be to first measure the current
running through those reversed LEDs. It may not be measurable :)
Best,
Steve
----------------------------------------------------------------------
# sbolt@xs4all.nl # Steven Bolt # popular science monthly KIJK #
----------------------------------------------------------------------
------------------------------------------------------------------------
eGroup home: http://www.eGroups.com/list/alt-beam
Free Web-based e-mail groups by eGroups.com
Home