Home
7 minute read
The HOME tab
provides you with a control interface which gives tou the ability to axes manually or automatically. In this section will provide a explanation of all the functions in the HOME tab
.
Main Controls [1]
The Main Controls represent the state machine of the system. The states can be changed by pressing Disengage
, Engage
and Switch Off
. After the system is powered on, it will automatically start in the OFF_S
state. You can find the current system state in the top right corner.
To activate the system press Disengage
button. The system will enter the DISENGAGE_S
state Servo-drives will be powered but movement commands are not tracked.
From the DISENGAGE_S
state you can go to the ENGAGE_S
State by pressing the Engaged
button. the system is activated!
In case of a E-stop error the system will recover in the OFF_S
state.
Mode Switches [2]
If the system is in the ENGAGED_S
there are a couple Mode
options to control the servo-drive.
Manual Jog [3]
Manual Jog can be used in Manual
mode and enables the you to jog each axis separately by means of the joystick
.
In Engaged_S
state and Active
mode it is possible to use Manual Jog
.
You can move each axis with a joystick
:
- Moving up resulting in a clockwise rotational movement,
- Moving down resulting in a counter-clockwise rotational movement.
The Joysticks are aligned from left to right respectively axis 1 till 6. The current target position is displayed underneath each joystick.
In this example the system is put in Engaged_S
state the modes are set to Active
mode and Manual
mode. The values of the Manual Jog outputs are changed by means of moving the joystick.
Signal Generator [4]
Signal Generator can be used in Manual
mode and enables the user to control one or multiple axes by means of a Signal Generator
.
Per Default a Sine signal is generated with the Signal Generator. This can be changed in Motorcortex Desk
.
The Signal Generator can be configured by a set of parameters: Fadetime
, Amplitude
, Frequency
and Offset
for each axis. The output of the signal generator is displayed in the output
column.
In this example the system is put in Engaged_S
state the modes are set to Active
mode and Manual
mode. The signal parameters have been defined before the signal generator is enabled.
Network Inputs [5]
You can use Network Inputs
to control Advanced Setpoint Generators with are adjustable in: Velocity, Acceleration and Jerk parameters. You can use Network Input
if system is in Engaged_S
and the mode switches are set to Active
, Auto
and Network Input
.
While in Network Input
mode, Using Motorcotex Desk
you can parameterize the Setpoint Generator
under the following path: root/Control/jointAutoSetpointGenerator0x/
. In this path the following parameters can be adjusted:
input
- This is the input of the Setpoint Generator [rad
].maxAcc
- Here you can adjust the maximum Acceleration [rad/s^2
].maxJerk
- Here you can define the maximum Jerk [rad/s^3
].maxTarget
- Here you can define the maximum Target Position [rad
].MaxVel
- Here you can define the maximum Velocity [rad/s
].output
- This is the output of the Setpoint Generator [rad
].
In this example the system is put in Engaged_S
state the modes are set to Active
mode, Auto
mode and Network Input
mode. In the side frame you can see the Desk-tool. within the parameter tree
under path root/Control/jointAutoSetpointGenerator0x
the settings maxACC
, maxJerk
and maxVel
of the SetPointGenerator need to have a value other than the default value of zero. after this you can define an input
value.
Motion Player [6]
The Motion Player is part of the auto mode and enables the user to play a motion-file, which is uploaded after defining its location in the configuration.json file.
With the motion player the user can define custom motion profiles defined in a .csv
file. These files can be loaded to Motorcortex Generic App by storing the .csv
file in a predefined default path as specified in the config.json
: "MotionPlayer": "/etc/motorcortex/config/motionfiles/"
. You can change the motion file path in the config.json
to you own liking.
The motion file
format should meet the following criteria in order for the motion player
to read it. A example file is stored in /etc/motorcortex/config/motionfiles/00_ExampleFile.csv
.
Motion file criteria:
- The motion file name should have the following format
xx_yyyy.csv
.xx
represents the motion file index number in the range of 00..99 as theyyyy
value can be changed to every name you like. - The first row of the .csv format should have the value
loop = z
.z
represents the repeat function of the motion player, for no repeats replacez
with0
. For the Motion Player to repeat automatically the motion file replacez
with1
. - The second row of the .csv file contains column information
# Time Axis1 Axis2 Axis3 Axis4 Axis5 Axis6
. - The data in each row is
separated
with aspacer
to indicate the columns. Note: all other spacing signs will result in motion player failure.
In this example the system is put in Engaged_S
state the modes are set to Active
mode, Auto
mode and Player
mode. The 00_ExampleFile.csv
is loaded
, Index Profile: 0
selected. Repeats light is off because in the .csv file loop = 0
. Motion file starts after pressing Play
. While in play the Progress [s]
starts to count and output data is shown in player output
.
Positions [7]
In Positions the Target
and Actual
position of each axes is displayed. The Target
position is defined as the summation of the Manual Jog and Signal Generator value. Because it is a summation it gives the possibility to Manual Jog on top of the signal generator.
Mini Plot [8]
In the mini plot the Actual
position value of each axis is plotted in a different color.
Virtual Twin [9]
A 3D visualization of 6 axes is used to displayed movement of each axis based on its Actual
position.