Hi all,
Since my goal is to put the robot arm to the industrial environment, I am thinking about maintaining the accuracy of my AR4 robot arm when doing the long period of time, especially the encoder part since it will generate feedback so that the robot will remember the position without keep re calibrating itself
I have seen Chris Annin use the shielding cable for the encoder wire connection, but from what I have researched, the shielding cable will have no effect of preventing interference if it is not grounded itself. If anyone also notice this and what is you guys solution to solve this problem?
The encoder in AR4 has differential outputs or complementary signals for noise immunity. When noise is present on a differential pair, the differential receiver cancels the noise aka common mode rejection. Not sure why AR4 only used single ended A and B signals. In that case a shielded cable can help if noise is a concern. A shield if not grounded has no value.
@david-t-phan Thank you for your reply I wonder if the encoder in AR4 already can cancel noise itself, then why I keep getting error on the motor encoder everytime they are turned on during runtime. I have to turn off all the encoders themselves to make the robot be able to run the program. However, this would just simply a waste for the encoders on the motor since I do want to use them to get position feedback. Again, I acknowledge that the shield cable without grounding have no effect in protecting sensitive signal. That is why I confuse whether Chris Annin also know about this because in the manual instruction, the shield cables of the encoders and the limit switches are only shielded, but not grounded.
@nhan-truong hard to troubleshoot noise related issue over email. Suggest to put a probe on the encoder line(s) to check for noise source which can come from power supply in your setup. The encoder is neither cancel or generate noise. You can try to external line driver to convert differential outputs from encoder to single ended. But that requires wiring modification from encoder and external line driver device.
I believe the noise are coming from the motors themselves while on power. Since the power supply coming to the motor is relatively large, it is very possible that the fact that the wire of the motor going along with the signal wire of the encoder in the braised sleeve have cause the noise that affect the sensitive signal of the encoder transmit to the computer program.
My effort now is to just replace every wire that are not shielded to completely shielded Ethernet cables, along with continuously grounding all of the shielded cables to ground, including the box and the robot itself. I will try this and we will see how the signal of the encoder would improve
Nhan,
I am still sourcing electrical items including wiring. Any update on your results?
Jeff
I've just started upgrading my AR2 to an aluminium AR4. I had this same question re screen grounding. I've decided to provide a connecting wire to the screen section of the encoder wire bundle & decide later whether to ground or not. A section of screen which is ungrounded could act as an aerial? A screen grounded at both ends is not generally a good idea?
