Hi Neil,

Either use a different multiplier for your integral term or different
limits for integral growth, based on polarity.

Cheerful regards, Bob

On Tue, Dec 30, 2014, at 06:37 PM, Neil wrote:
> Hi all,
>=20
> I'm trying to make what is essentially a servo for a one-off project for=
=20
> a friend... I'm driving a window motor (I can control speed=20
> bidirectionally) with a controller/driver and there's a potentiometer=20
> attached for feedback.  For now, I'm using an Arduino to get it running,=
=20
> but I'll get a custom PCB made for a PIC-based circuit later.  BUT...=20
> there's a one-sided force (spring) on the load the motor is moving.
>=20
> For an initial test I tried setting the motor speed proportional to how=20
> far off it was from it's target, but when it gets close, the spring load=
=20
> is too much for it.  Changing the force so it's a bit more at close to=20
> target position helps, but I think I'm just going to use PID.  My=20
> concern is that with the spring force acting only one way, the PID=20
> parameters would be different from one side to the next.  Essentially to=
=20
> stay in a fixed position, I'd need a continuous non-zero speed on the
> motor.
>=20
> How is this (easily) handled?  Do I do separate cases for one direction=20
> vs. the other?  Or do I empirically determine how much speed to send it=20
> to counter the spring force and add that value before sending to the=20
> motor controller?  Of course, this would assume the spring it mostly=20
> constant force for the full range of movement.  But what if it's not?
>=20
> Cheers,
> -Neil

--=20
http://www.fastmail.com - Same, same, but different...

--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.