So I finally got around to implementing this and it works.  I initially=20
used the Arduino PID library, got it working/tuned and then wrote my own=20
PID code for a PIC18F.  I used integer math, then scaled down to get the=20
appropriate range.  In my final tuning, I used (what I believe is=20
feed-forward) an offset -- I experimentally determined what constant=20
output value would just counteract the spring pulling on the servo and=20
added that to my PID output result. Works pretty nicely.

Much thanks,
-Neil.


On 1/1/2015 7:36 PM, Neil wrote:
> I was continuously spitting out the raw A/D value from the control and
> feedback pots and can see that I have the resolution, and there there is
> some discrepancy between those two.  Range is ~113 to ~870 (raw A/D
> values), and the difference is sometimes around 30 when the motor
> "stalls" for a bit . It seems like it wants to correct it, but can't,
> then every so often it gets enough oomph (I expect from the Integral
> component building up) to move, overshoots a bit and then repeats in the
> other direction.
>
> I am seeing (on the continuous serial dump) the LSB (in decimal)
> toggling by one digit, so I don't think it's noise (since the
> discrepancy is ~30).  But I can add some IIR filtering easily anyway.
>
> Cheers,
> -Neil.
>
>
>
> On 1/1/2015 4:42 PM, Ruben J=F6nsson wrote:
>>> *Basic* tuning (ie: no oscillation, chatter and getting pretty close to
>>> setpoint) with the Ziegler-Nichols method was relatively easy, but it
>>> seems a bit "sticky" -- ie: very small changes to my control pot don't
>>> immediately take effect, and continuing to change the pot a bit more
>>> suddenly makes it "jump" to the correct position.
>> Could it be that you have too low resolution so very small changes to th=
e pot
>> isn't registered? When you move it a bit more, you will get a step chang=
e.
>>
>>
>>> Also, just sitting
>>> with a fixed setpoint, it will move a bit every once in a while (like
>>> 5-10 seconds).
>> Could be noise in the signal from the motor position and/or pot. You nor=
mally
>> will get one to a couple of bits jitter from the A/D converter if your a=
nalog
>> circuit isn't well filtered and separated from the digital circuit. Perh=
aps
>> digital filtering could help.
>>
>> /Ruben

--=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
.