[Coco] Artificial Intelligent (AI) Opponent

[Gene Heskett]

On Tuesday 02 January 2018 11:34:42 Dave Philipsen wrote:

> A lot of applications don’t require PID. You just use a feedback
> amplifier to drive a servo, essentially just the ‘P’ of the equation.
> But when you need to be more precise or perhaps overcome certain
> variations in the operating environment then PID works great.  Back
> then I had to buy a book with a chapter on it to understand how to
> write the program. We used 12-bit ADCs and calculated with integer
> arithmetic (no floating point).
>
> I designed some servo boards for hydraulic valves that drove large
> cylinders used to move a motion base simulator ride and that did not
> require PID.
>
> On the other hand, I worked with a company in Korea on a huge motion
> base simulator ride that carried upwards of 2-3 dozen people and I
> believe PID was used for that (by means of a PLC). The company
> specialized in military aircraft simulators but they got involved with
> the entertainment business mostly because no one had ever made a
> motion base that large for entertainment purposes.  The hydraulic pump
> was about the size of a medium-sized truck!
>
>
> Dave
>
> > On Jan 2, 2018, at 8:51 AM, Allen Huffman <alsplace at pobox.com> wrote:
> >> On Jan 2, 2018, at 8:48 AM, Dave Philipsen <dave at davebiz.com>
> >> wrote:
> >>
> >> Cool. So the feedback on the fire nozzle was some sort of flow
> >> sensor? And I’m eager to know how you would have incorporated a PID
> >> loop into a garage door opener!
> >
> > My previous job was working on oil industry "flow computers" and the
> > made big use of PID loops. I’d never heard of them before then.
> >
> > Now I wonder if we look up the old A-BUS interface stuff if they
> > would be discussed for hardware those boards could be hooked up to.
> >
I think for something like that, where erosion of the edges of a several 
hundred dollar spool valve could mess up the pid's performance, today I 
think I'd look for brushless servo motors big enough to do that job for 
each "cylinder". With todays very readily available pwm to dc servo 
drives, I think you wouldn't have near the maintenance the hydraulic 
system needs, even simulating a 8.0 earthquake for 20 people. Jon Elson, 
Pico systems, has such a drive, rated 160 volts at 20 amps for the 
motor, for about a $130 bill on your front stoop. Only slightly thicker 
than a pack of 100 mill smokes, I'm using 2 of them to drive the 1hp 
spindle motors on two of my machines, with response times in the 
millisecond category if the tool hits a tough spot, or if theres too 
much noise in the feedback it speeds the motor up, and drags it down 
makeing a sound like the headstock bearings are eating the ball cages 
and total bearing failure is only hours away. But it not, its the 
backlash in the gears making all the racket. I found that out recently 
when I installed a second encoder, 1000 edges a rev. directly on the 
motors rear shaft, but still using the index from the old encoder. That 
nicely turned a 264 edge encoder into a 7161.61 edge encoder with a 
vanishingly smaller quantization noise. I am still turning up the Pgain, 
but its easily 10 times as "stiff" right now at maintaining speed under 
load as before.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>