On Tue, 23 Sep 2014 19:20:17 -0500, Karl Townsend wrote:
You need to ping Tim Wescott who sometimes posts here - he's the guru.
What? Who? Huh?
Maybe I've been sleeping under a rock, didn't know this.
I've been fighting a buzzing at stop in my Galilmc.com controlled AMC
servo amps and DC servos. 9 out of 10 times the servo will stop
smoothly, the tenth it will oscilate about 1 or 2 encoder counts so
rapidly it makes a buzzing noise.
Adding the low pass filter to PID helped (see below for manual
description) but i ended up reducing P,I and D by huge amounts to stop
it. Now the system is sluggish. FWIW, Galil has servo tuning software
to automatically tune your system. that worked for many years till this
issue popped up.
Comments?
user's manual excerpt snipped
The low-pass filter and reduced gains sound an awful lot like a bandaid,
not a solution.
If it worked for years and now it doesn't, then something has worn or
broken, or something got "upgraded", or it's been doing it all along and
you only just now noticed. Control systems only work as well as the
plants that they're driving -- they can't do magic if something is worn
or broken.
What's attached to the servo motors? If it's a ballscrew or some bolted-
up assembly, my first guess would be that over time slop has developed in
the system that allows the motor to wiggle a bit before it comes on to
the load. I base this both on the fact that it used to work and doesn't
now, and the fact that backlash and slop can be a position-dependent
thing.
It could also be something electrical -- a broken ground connection
between the Galil board and a motor, or a power supply whose capacitors
have dried out and allow motor current to couple in to the supply
voltage, or worn-out brushes on a motor.
Either that or you've added something with a bit of spring, or you've
taken off some bracing that lets the motors or their loads "twang" a
bit. Resonances like that can be death to a stable control system.
So as tedious and pointless as it sounds, you may want to go over
whatever backlash adjustment you have, loosen and re-tighten all mounting
bolts, check for loose bearings, reverse any "improvements" you made
about the time the singing started, check operation with a different
power supply (or hang an oscilloscope off the power supply and see what
happens to the power rail when it's "singing"), and all that due
diligence stuff.
For a while I was viewed as a sorcerer by the FLIR Systems service
department, because nine times out of ten I could fix a control system
problem just by having them replace one magic bearing. Then I confessed
that you had to disassemble most of the bolted joints in the assembly to
get to the bearing, and in the process of reassembly you tightened them
all to spec, with Loctite. After that I got summoned to service a lot
less often, but the problems they presented me with were a lot more
challenging.
(You couldn't just say "tighten all the bolts", because that made no
sense to them. So they'd tighten what they could reach and claim to have
done them all. But replacing a bearing -- that's real.)
(Soon after, the jerks upstairs decided that engineers shouldn't be
allowed to talk to service, and visa-versa. I hate big companies.)
Keep in mind that I'm shooting in the dark here, because I know almost
nothing about your setup. But you know that something has changed; now
you just need to figure out what it is, and change it back.
--
www.wescottdesign.com