How Do I Eliminate Multi-Servo Jitters ???
wperko
Posts: 66
Hi,
I'm trying to control multiple servos tied to the BOE w/a BS2 but I get uncontrollable jittering ...
How Do I Eliminate Multi-Servo Jitters ???
This happens with other control systems too ... separate channels and/or parallel on one channel ...
Jitter Example Video
Post Edited (wperko) : 4/14/2008 2:17:15 AM GMT
I'm trying to control multiple servos tied to the BOE w/a BS2 but I get uncontrollable jittering ...
How Do I Eliminate Multi-Servo Jitters ???
This happens with other control systems too ... separate channels and/or parallel on one channel ...
Jitter Example Video
Post Edited (wperko) : 4/14/2008 2:17:15 AM GMT
Comments
1) Inadequate power for the servos causing a drop in the power supply voltage to the BS2 which causes the Stamp to reset.
2) You're not sending control pulses to the servos at least 50 times a seconds. This is a requirement for any R/C servo for smooth movement.
Can you post a simple output routing to servos (parallel operation) plugged into ports 12, 13, 14 and 15 on a BOE with a BS2 module?
Maybe that is all I need ... I'm just using a simple pulseout loop ...
DO
PULSOUT 12, 600
PULSOUT 13, 600
PULSOUT 14, 600
PULSOUT 15, 600
PAUSE 5
LOOP
Post Edited (wperko) : 4/6/2008 5:05:57 PM GMT
Actually, I have tried this with various cycles from 5ms cycles to 10ms cycles to 15ms cycles to 20 etc... all the way past 50ms cycles ...
I actually used a two trim pots in series to get a wide range of cycle triggering ... slowly moving one at a time to see if I could find a stable setup ... I never really get a stable jitter free system.
I have used seperate power source capable of 2A service ... I still get jitter.
The jitter is minor most of the time when things aren't changing ... but when I try to move just one servo on one port the whole structure jitters like a caged wild animal.
"...the whole structure jitters..." ? What "structure" ? I.e., is the jitter caused _by_ the one servo that is not turning smoothly, or are you referring to some other kind of jitter? And, when moving more than one servo, do you also get the jitter ?
And, how are the pots connected to the circuit (in series with what)? How much load is being put on the system by those pots ?
PAR
Four Servos on Four Different BS2 OUTPUTS
PULSOUT 12, 600
PULSOUT 13, 600
PULSOUT 14, 600
PULSOUT 15, 600
PAUSE 5
But even if the servos are all attached in parallel on one port the problem is the same. And yes as I stated before, I've tried this with seperated power to servos and controller and even using 5-seperate power supplies one for each component! It's all the same jitter.
If I have the program send a new PULSOUT 14, 600 and there is no other changes in any settings all the servos jitter (causing the whole structure to shudder/jitter like a caged wild animal)... This occurs for all robotic structures I've built ... whether being a robo arm, a snake, a bipdeal etc... when I cause just one servo to move in programming or by a pot change ... just a little bit ... all the other servos jitter.
Then even if I am not making any changes the structures I built from multiple servos will jitter and sometimes very wildly.
It doesn't matter if I use a pot or a program code change ... the structures jitter ... the pot is a RC circuit tied to ground ... negligible load.
PAR
If you don't output that control signal, then all bets are off. Repeating it every 5 mSec is NOT outputting that control signal.
Now, with a BS2, the "PULSOUT" values is in 2 uSec ticks. So 600 is 1.2 mSec, which is okay.
A servo is designed to hold a position against a very small force. It could be you're trying to use servo's which are too small for the forces you're trying to control. It could also be that your power supply is too small -- a servo can pull 500 mA against a large force.
It's also possible that when you put in the code to read a potentiometer, that you're taking too long, and getting outside the 50 mSec refresh interval -- which will cause jitter against external force.
Yes, the grounds are tied ... the servo signal wouldn't do anything if they were not!
I'm looking for a solution not 20 questions ...
PAR
Again, read all the information first ... then give me a solution ...
Servos move between about 0.5ms to 2.5ms pulses on a 30ms-50ms trigger.
I tried shortening and expanding the trigger through several values between about 5ms up through over 50ms to see if there would be a "sweet spot" where the servos didn't jitter ... there was none.
The servos are only carrying themselvs as a load ... that's almost nothing for a standard size hobby servo.
I want to know how to send signals out to a bunch of servos without getting any jitter ... not answer 20-questions.
You haven't said what current your power supply can put out, what voltage you're running your servo's at, or what physical force your robotics projects require.
Since you haven't, 20 questions is appropriate.
BECAUSE -- if you really are using that code, and you've 'bumped' the pause to 20 mSec, then you already HAVE the "simple test code" you keep asking for -- that should work.
Sorry, I use DVD MPEG video ... if your Mac can't view standard codec videos then you need to upgrade.
MS Windows PCs and Linux machines don't have this problem.
I asked for a solution ... not 20-questions about my setup.
As I explained a long time ago ... I am using several different setups with batteries and power supplies with up to 2A available for the simple circuits.
-Phil
We don't have a working copy of your video nor a schematic nor a complete listing of your program. It's not helpful to blame others for not having the "right codec". I have a Mac with the latest stuff and I run Windows XP under it and keep that up to date as well. I can't get your video to play.
Good luck finding out what your problem is. As I mentioned, it's not likely to have anything to do with the Stamp itself or the servos since operating 2 or 3 servos smoothly and simultaneously with a Stamp is straightforward using code essentially the same as what you've shown.
I am waiting for someone to announce: "This has just been a test; you can now return to your normally scheduled entertainment."
Before that happens let me offer an excerpt from a poem by Sylvia Plath, called
Love is a Parallax Analysis
'Perspective betrays with its dichotomy:
train tracks always meet, not here, but onlyin the impossible mind's eye;
horizons beat a retreat as we embark
on sophist seas to overtake that markwhere wave pretends to drench real sky.''Well then, if we agree, it is not odd
that one man's devil is another's godor that the solar spectrum is
a multitude of shaded grays; suspense
on the quicksands of ambivalenceis our life's whole nemesis.
So, wperko you are anything but ambivalent.
but me thinks you are thrice wrong.
No morons here... nor gorish intent.
No replies that were anything but honest.
A poor substitute for time well spent.
You have misinterpreted the remarks and remakes
of good replies in two different threads.
These are real people... owed an honest apology.
But you will turn heal, and flee
instead.
Rich
For those anxiously awaiting the final release of my 96 bit pixel... I've been forced out of my office by the mysteries of child rearing [noparse]:)[/noparse] In my head, the pixel was done before I started. So... any day now[noparse]:)[/noparse]
I don't know.
I just don't know.
Sorry
I was able to view the 1.2 MB, 5 sec video on my Mac. It might be because I have "Perian" installed.
Here you can see the power supply he is using.
Rich H
Jim