AL5D arm, questions about jitter and slow motions

Lynxmotion Robot Arms
1

Hello everyone,

I would like to get your feedback about two issues I encountered using AL5D arm.

I’m working on visual servoing,sending robot commands according to images sent by an embedded camera on AL5D end effector. My algorithm needs a stable image and give me small velocities to integrate.
I designed a software to control the robot end effector in velocity mode and it works well on relatively fast movement, as soon as I try to give it a pretty slow movement speed the motion become unsmooth and jitters a bit. (I assume that issue is due to servomotor which have not enough resolution but I’d like to get your feedback)
I also noticed that AL5D is unable to achieve certain stable position, with no movement command (robot should not move) the end effector jitters a bit and I hear digital servos switching between on and off constantly.
I have the same issues happening using lynxmotion RIOS.

I’d like to get rid of these jitters and unsmoothness movement so I’s like to hear your feedback on these.

Thanking you in advance.

2

The AL5D uses analog servos so I’m not sure what you mean about switching between on and off. The servos have a speed that they operate very well at. This is where the arm is operating in a range the built in PID is setup for. It will not operate perfectly smooth at any speed. It’s just how they work. Adjusting the mechanical dampeners may help.

3

Under powering the arm may cause some of what you are seeing. Servos take a good bit of current when under load.

4

True, you might get better performance with a higher current supply.

5

Hi, thanks for this great forum, I’ve been browsing and sneaking about for a little while. Time to stop lurking.

I have been reading up on the precision discussions and improving the power of the arms with “better” servos. However, most of the discussions are somewhat older and seem to be about the older arms instead of the AL5D. I have some questions on this topic:

  1. Would replacing the included servos with stronger digital servos (like 5990) a) reduce jitter b) improve precision and repeatibility?
  2. When replacing servos could I use any servo with the same dimensions?
  3. I understand these are not industrial arms by any measure. With this new range what would be the bottle neck for continuous 8hr/day, 7 days like operation? Just the servos or the actual arm mechanics?

Thanks for your help.

6

Apparently I was searching for the wrong terms. Searching for accuracy yielded viewtopic.php?f=6&t=7392.

To summarize (and please correct me if I am wrong):

  1. Would replacing the included servos with stronger digital servos (like 5990) a) reduce jitter b) improve precision and repeatibility?
    Yes, both should be better.

  2. When replacing servos could I use any servo with the same dimensions?
    The 5990 should fit in the standard servo brackets. Further info on replacements in the thread i linked above. If the case dimensions are the same, can I assume a servo will fit, or can there be differences in where the spindle exists the case?

My third question stands though. I am not about to populate a factory with AL5D’s (hell, it’d be my first robot), but the arm looks so rugged I do wonder if “only” the servos would be the bottle neck for continuous use or if the actual arm would break down at some point.

7

Sorry for missing this. Well I saw it this morning, but forgot to reply when I got to work.

The arm is sturdy aluminum construction. The servos have a potentiometer that will eventually wear out. The manufacturer does not supply any data concerning mean time between failure. This is why we don’t recommend the arms for 24/7 use.