|Re: Störung auf linker radencoder: entwurfsfehler Kategorie: Pro-Bot128 (von Mansfeld - 31.03.2010 15:51)|
|Als Antwort auf Re: Störung auf linker radencoder: entwurfsfehler von PepeV - 30.03.2010 10:12|
> > > The outputs of Timer T1 PWM X (X1_3 / OC1A and X1_2 / OC1B) couse interference on the collector|
> > > of the left wheel encoder phototransistor T3, which is connected to input port X1_9 / INT7. This is
> > > coused by a design error on the printed circuit board of the drive unit.
> > >
> > > On the top side of that board the wire that runs from T3 to CON2 runs very close and parallel to the two
> > > lines on the bottom side that carry the PWM signals for the motors. The result is that INT7 (left counter)
> > > receives extra pulses from interference and the robot will not drive strait at low speeds. (At low speeds
> > > the pulses from the phototransistors are less steep and therefore more sensitive to interference.)
> > >
> > > I solved the problem by disconnecting the printed wire from T3 to CON2 (cutting it near T3 and CON2)
> > > and replacing it with a shielded wire. But hopefully the designers will find a more elegant solution by
> > > altering the printed circuit board.
> > Hi Pepe,
> > thats a very interesting report. U. Sommer recommends in his book to
> > insert two Schmitt-Triggers between collector of T3 / T4 and the
> > Prozessor pins X1_9 and X1_10.
> > How did you GND the shield: On both ends, sender end or receiver end?
> > (I don't own a scope, I can't watch the result).
> > Regards
> > Fred
> Hi Fred,
> I grounded only one end of the shield to prevent 'ground loops'. I chose the receiver end but I don't
> think it matters very much.
> Maybe you want to read my contribution 'Improving the wheel encoders' too. Together with the above
> this really is a big improvement and you will not need the Schmitt-Triggers anymore.
thanks, I will give it a try.
Zur Übersicht - INFO - Neueste 50 Beiträge - Neuer Beitrag - Suchen - Zum C-Control-I-Forum - Zum C-Control-II-Forum