Alt-BEAM Archive

Message #05232



To: beam@sgiblab.sgi.com
From: Bob Shannon bshannon@tiac.net
Date: Sat, 10 Jul 1999 10:55:11 -0400
Subject: [alt-beam] Re: Something funny with the 1382 voltage triggers?


Steven Bolt wrote:

> On Fri, 9 Jul 1999, Bob Shannon wrote:
>
> ---8<---
> > Either of these approaches can resolve the problem without
> > resorting to other SE circuits.
>
> I can't help but wonder why you are so reluctant to replace that
> 2-transistor latch/driver, while keeping the voltage trigger.
>
> Note that a monostable would not merely avoid all reset trouble.

Why would a monostable be better?

I can see a number of reasons (especially with a wheel stuck) that a mono
would
be the very last thing I would try.

Why not fix the dang reset problem without adding lots more parts?
Its easier to 'fix' the 'problems' with the popper engine than it is to
design something
new and justifiy it as being better.

> If you stop and think, you'll realize that using the storage cap
> and the load as a crude RC timer isn't such a great idea in most
> applications. The monostable gives you an independent RC constant,
> easily adjustable for maximum efficiency.

I totally disagree.

The popper engine reacts to the motor load, rather like Tilden's implex
idea.
Using a monostable precludes this effect, to what advantage? It means that
the
on time of the latch is no longer dependant on what the motor is doing.

Why is this more efficient? (Here is a perfect example of theoretical
efficiency!)

> You might connect the voltage trigger output to the active high
> input of a monostable (1/2 74HC221, for instance). A cap and
> a resistor determine the output timing (about 0.7 * C * R).

Absurd. You want me to carry around a HC221 as well?

> Use an adjustable resistor to make burst duration continuously
> adjustable; replace with a fixed resistor when you've found the
> optimum value.

I like the SE and motor to control the burst duration as a function of
loading.


>
> 1K connects a BC817-25 to the active high output. The transistor
> drives your motor.
> The smallest supercap (47000uF) would provide sufficient current
> in a light and compact package, and the voltage would hardly drop
> during motor bursts, unless you want them rather long.

Greatest efficiency comes from a ratio of burst lenght, wheel base, and
wheel diameter (given the limits of motor torque of course).

There has been a body of though that suggests that best efficiency is
acheived
with many short steps. This is incorrect. A wider wheel base and a longer

step means that we are spending less energy overcomming stiction and
rolling
friction.

We can preserve a tigher turning radius simply be slececting a motor with
high enough efficiency and torque that the undriven wheel can turn
backwards
about half a turn when the other motor fires. This gives us some pivioting
component
to each step that can compenstate for the wider wheel base without loosing
much energy at all (sue to suspension deflection storing that energy, and
returning it
on the next step).

> If your motor does well at low voltages, efficiency will be
> increased if you use for instance a 2V (instead of 2.7V) trigger
> level, or go even lower. You still get good performance, because
> the voltage won't drop much and certainly not rapidly, while the
> motor is running.

Wrong, this totally ignores stiction and rolling friction. This is a myth
of theorectical
efficiency.

> That's five and a half components for a single SE, or eleven
> for a double version. The 74HC221, all resistors, the 100N cap and
> the driver are SMT, which means that weight and size will be mostly
> determined by the solar panel and motor(s).

A better solution would be to use a comparitor.

A monostable is a bad idea when efficiency matters Steven, I've tested it,
and
rejected it.



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

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



Home