Re: [NTLK] A programming question (Newton as a controller for an electric motor)

A couple of thoughts:

The newton has no parallel inputs, nor any Analog to Digital
converters (audio-in could be rigged as such in a pinch I guess).
You'd have to do the scutwork of converting analog inputs from car
sensors into digital values, and taking digital output for the cruise
control and generating the varying analog output to the throttle
OUTSIDE the newton.

You probably already realized this (you mentioned that techy friends
could build a dongle for such a purpose), why not make this dongle
intelligent in it's own case? The only way the dongle can communicate
with the newton will be a serial protocol - so the dongle will have
to have some sort of micro-controller (picaxe, atmega, basic stamp
etc.) to communicate with the newton.

I would wager that integrating the actual cruise control
functionality directly into this micro-controller would be safer and
more useful as I would trust a micro-controller to have much higher
reliability than a newton for uptime / realtime response, and would
always be in the car - allowing you to use cruise if the newton isn't
available.

This reduces the complexity of the Newton's tasks to polling /
displaying / logging the status data coming from the micro-controller
- giving you gauge readouts etc, but not controlling anything in the
car itself.

Yes, it would remove the nifty factor of having your car be Newton
Controlled (Headline: Electric Bug Powered By Newton Technology!),
but the safety / utility / ease of implementation is likely worth
giving up the nifty.

J. Tyler Nichols

On Aug 11, 2006, at 1:11 PM, Chris Browder wrote:

> On 8/11/06, Florian Mosleh <indigo@13th-floor.org> wrote:
>> Hi, I have a question about this idea.
>> On most cars, isn't cruisecontrol a system whereby the amount of
>> acceleration is continuously adjusted in order to maintain a constant
>> speed?
>>
>> It seems to me that simply maintaining a constant power throughput
>> would
>> not help you maintain a constant velocity at all. If you were
>> trying to
>> code an app that allowed you to do cruise control, I would think that
>> you'd need to feed the Newton metrics about the vehicle's current
>> speed
>> and acceleration and, have it increase the power to the motor when it
>> notices a drop in velocity, thottle back when it increases, etc.
>
> I like what you're thinking, but from what I've seen of both a Ford
> Focus and a 2006 Jetta, they only keep the car speeding up, they do
> not slow it down if it goes down hill. It seems logical that it would
> throttle back the fuel and slow the vehicle to maintain the speed, but
> it simply adds fuel to keep the speed up should it drop. VERY
> aggitating, set the cruise at 35, go down a steep hill and watch the
> speedo climb to nearly 50.
> --
> This is the NewtonTalk list - http://www.newtontalk.net/ for all
> inquiries
> Official Newton FAQ: http://www.chuma.org/newton/faq/
> WikiWikiNewt for all kinds of articles: http://tools.unna.org/
> wikiwikinewt/
>

-- 
This is the NewtonTalk list - http://www.newtontalk.net/ for all inquiries
Official Newton FAQ: http://www.chuma.org/newton/faq/
WikiWikiNewt for all kinds of articles: http://tools.unna.org/wikiwikinewt/