go_away

Author Topic: Servo Problems  (Read 288 times)

0 Members and 1 Guest are viewing this topic.

Offline vipulan12Topic starter

  • Robot Overlord
  • ****
  • Posts: 176
  • Helpful? 0
Servo Problems
« on: July 12, 2014, 06:19:05 PM »
hey guys, I was working on a biped robot that uses the UNO and 10 servos, and by changing the angles i should be able to make the robot walk however for some reason my servos tend to twitch like crazy
i am powering the 10 servos using 8 AA batteries(6 volts in parallel) the arduino is not powering any of the servos
the arduino is being powered by a nine volt battery (seperately)

i first tried grounding everything together, that worked but the twitching came back
i also tried changing the batteries which also worked for a little while but the twitching returned

a friend said that I should power the servos using the batteries and the arduino in parallel(could this potentially damage my arduino?)

what exactly is the cause of these twitches?
they mainly occur at  the moment the servos are powered(just by flipping the switch pn the battery pack)

Offline Billy

  • Full Member
  • ***
  • Posts: 114
  • Helpful? 2
Re: Servo Problems
« Reply #1 on: July 13, 2014, 02:39:33 PM »
i first tried grounding everything together, that worked but the twitching came back
i also tried changing the batteries which also worked for a little while but the twitching returned


In every system where there are issues that can be attributed to noise, one of your first actions should be to stabilize the supply. Think of the batteries as the power source for your power supply. The power supply at each part being bypassing capacitors.
http://en.wikipedia.org/wiki/Decoupling_capacitor
On each of your power lines, and including at the battery, try install a 10uF electrolytic cap in parallel with a 1uF ceramic cap. Install these at the batteries, and if possible at the power connectors to the servos.

You said you power the uC from a 9 volt battery. That is OK but the negative terminal of the 9V battery must be connected to the negative end of the AA battery pack. That assumes you do not have any isolation between the servo circuit and the uC. Isolation is an advanced topic that you should consider if you still have issues.

Final thought: your description makes it sound like a noise issue, but of course it's possible you have a SW issue and the pulses to the servos are simply irregular.

Offline vipulan12Topic starter

  • Robot Overlord
  • ****
  • Posts: 176
  • Helpful? 0
Re: Servo Problems
« Reply #2 on: July 13, 2014, 05:22:44 PM »
Thanks for the advice, I ll try that and see what happens
what do you mean by  isolation or the servo pulses being irregular(do you mean the servos might be the problem)?

Offline Billy

  • Full Member
  • ***
  • Posts: 114
  • Helpful? 2
Re: Servo Problems
« Reply #3 on: July 14, 2014, 10:59:12 PM »
what do you mean by  isolation or the servo pulses being irregular(do you mean the servos might be the problem)?


Isolation is a means to leave two circuits unconnected electrically, but allow for information (signals) to pass from one circuit to another. There are several types but include using light to send signals in an optocoupler (optoisolator), or to use magnetic field in a signal transformer or digital isolator. I have even seen a large 50 Hz 3 phase power source isolated from the 60 Hz 3 phase power supply by using a large motor to turn a large generator. In that case the signal was mechanically transmitted.
Here is a link you should have read prior to posting your question, but I'll you slide by on this one: http://en.wikipedia.org/wiki/Galvanic_isolation
It's not the best wiki page I have seen but least they took the time to include diagrams.

Your cell phone allows your parents to send a text saying to come home without you having to actually see your parents, so that would be called social isolator.  Angry mom or happy mom, all you see is the words sent by mom. You don't see the upset look on her face because you're late for dinner AGAIN! You only get the clear, quiet, no-attitude message to come home.  Same with electrical isolation; you get the signal without the electrical noise that is on the other circuit. Angry circuit or happy circuit, you only get the message sent.  Motor (servo) circuits can be noisy (angry) and that noise interferes with the communication between uCs (mom) and other electrical parts (you).

My comment on the pulses being irregular meant simply that the pulses from the uC may not be the same distance apart or not the correct width every time the uC sends one. Maybe your SW is thinking about something else when it should be thinking about sending pulses.

I don't know anything about Ardinos so do not know how the pulse trains are generated. If they are generated by flipping bits in an interrupt routine then maybe you're taking too long in the interrupt and missing the start of the next scheduled call, or maybe you're modifying a variable that gets used in the interrupt when the interrupt gets called, so it's not the value you want it to be. Can't help you there.

 


Get Your Ad Here

data_list