Newsgroups: sci.electronics,comp.robotics,rec.models.rc
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!uunet!in1.uu.net!portal.austin.ibm.com!awdprime.austin.ibm.com!awdprime.austin.ibm.com.!nahshon
From: nahshon@vnet.ibm.com (Itai Nahshon)
Subject: Re: Servo Reversing circuit
Sender: news@austin.ibm.com (News id)
Message-ID: <NAHSHON.95May16144337@rs3bt-01.haifa.ibm.com>
In-Reply-To: deancap@ibm.net's message of 15 May 1995 17:43:22 GMT
Date: Tue, 16 May 1995 11:43:37 GMT
References: <DKN.95May15122459@cimar.me.ufl.edu> <3p83rq$2k4d@news-s01.ny.us.ibm.net>
Organization: IBM Israel Science and Technology, Haifa, Israel
Lines: 42
Xref: glinda.oz.cs.cmu.edu sci.electronics:132713 comp.robotics:20810

In article <3p83rq$2k4d@news-s01.ny.us.ibm.net> deancap@ibm.net writes:
>In <DKN.95May15122459@cimar.me.ufl.edu>, dkn@cimar.me.ufl.edu (Dave Novick) writes:
>>I thought I've seen this before in R/C Modeler, but I could be wrong.  Has 
>>anyone seen a circuit which will take an RC Servo pulse and convert it
>>to it's "reverse", i.e. knowing that 1.5ms is the center, a 1.7ms incoming
>>pulse will be converted to a 1.3ms pulse.
>>

>Maybe a PIC controller could  do it [possibly multiple chanels] and
>even add a preset position for pulse-loss [transmitter/receiver failure]

>Circuit would be 18-pin DIP or SOIC, 3-pin resonator, and maybe
>some Caps. resistors (variable) for the loss preset.

>Could be "warted" onto a servo ??

>Just some musings I actually never tried it :-)

>=====
>Dean
>deancap@ibm.net

I would not recommend using special circuits to reverse a servo. If you
HAVE to do it and you cannot just change your software here's one way
to do it:

Set a CMOS monostable (4538 or equivalent) to generate a 3 ms pulse
on the rising edge of the servo pulse. The original puse id subtracted from
the 3 ms puse and the result is fed into the servo.

Changing the 3ms time will change the servo sernter point.

It's possible to trick a single 4538 (with just a few additional resistors
and capacitors) to do the job. Hint: the other half of the 4538 should have
a time base > 2ms. It's triggered by the falling edge of the original puse
and reset when the 3 ms period ends.

Again. I do not recommend it. I prefer to do it in software (and I do not
mean in a dedicated PIC).

Itai Nahshon
Haifa, Israel
