Quote:
Originally Posted by drakesword
1) Correct/Proper wiring both CAN and Power.
Going to rewire most of the can-related components at VCU. We were using 6 conductor wire. Going to switch to two. 120 ohm terminator soldered to wire not crimped.
2) Ground isolation of components.
Again as above. possible ground loop through our can wires
3) CAN Interface used, Serial or 2CAN
Serial
4) Type of JAGS used, Black, Grey, Both
Black (5)
5) The programming language used
Java.
6) The control mode(s) used.
Speed and Position
7) encoders, pots, limit switches are being used.
Us-Digital(I think) encoders 200 pulses per revolution. Don't know what switches we use but we have 2 on our gripper.
To address potential blocking we will make individual threads to set output to each motor.
|
Our team's usage was similar:
1) Correct/Proper wiring both CAN and Power.
I wish we had load tested our robot's wiring, but it appears that no one ever tested the wire with a static resistive load...it's on our agenda for future builds. As such I can't be sure there wasn't a bad wire or crimp somewhere...however...please be aware that since last week our robot has seen competition in Palmetto and from what I understand no wiring issue presented with the robot configured to do only %VBUS over the CAN. Course that doesn't mean there isn't still an issue in that wiring somehwere...just not any major show stopping issues.
We tested our practice robot (same build as the production robot in the crate) for many hours at our test field and of all those hours we had less than 4 total lock ups that could not be cleared by any means other than a complete power down. That is statistically small...but troubling enough that the team is traveling with PWM cables and code to bypass any issues related to using them.
2) Ground isolation of components.
Our robot is using telephone wire between the Jaguars. All the wires on the interface follow that bus from the input at the serial port on the cRIO to the output where the terminator is. Hugh had suggested this as a possible ground loop, but we had to stop messing around with this for the moment and so we never broke this connection. Also that cable is not twisted in anyway, but it is unshielded, stranded 26AWG.
Our Jaguars are bolted to plexiglass, so they aren't connected to the frame ground.
As a result of the shear amount of work involved with the encoders and the drive train using the Jaguars, we entirely stopped using the encoders. When we had them connected, we used the encoder splitter we designed from the other topic and that provides complete isolation between the encoders and the Jaguars so there was no option of a ground loop from that source. When we did use this splitter our target set points for the Jaguar PID loops were velocity...and we did tune the loops successfully. However, in this configuration the number of time outs was greater than if we just ran %VBUS without the encoders. So it was either scrap CAN or scrap the encoders given the time constraints.
We still use some potentiometers on our arm that are connected to the Jaguars. We read those values back through the CAN, then hand them back to the Jaguars conditioned for %VBUS. This change was made by the students to make the arm motion more fluid. It also eliminated any further need for the Jaguar PID loop functions. The potentiometers should not be a source of a ground loop.
3) CAN Interface used, Serial or 2CAN
Serial
4) Type of JAGS used, Black, Grey, Both
Black (7)
5) The programming language used
Java and we use threading...doesn't help us when the CAN decides it won't stop blocking as we have it now. Never tried to avoid the blocking...it may be possible.
6) The control mode(s) used.
Velocity and position.
Velocity was working, position was always jerky...so we ended up using neither.
7) encoders, pots, limit switches are being used.
On the gear boxes but not used: U.S. Digital Encoders 360 count x 2
No limit switches
2 multi-turn industrial sensor grade potentiometers on the shoulder and wrist.