PLC programming is primarily about time

No Slip

• Encoder to Motor: calculated/estimated encoder PPM is proportional to motor RPM
• Motor to Conveyor: motor RPM is proportional to conveyor linespeed speed, inch/s
• Conveyor to Failed Can: failed-can speed is equal to linespeed
• Failed-Can to BSL Index: the movement of a modeled failed-can's 1-bit index in the bit-shift array is proportional to can position

The PLC only tracks the position of failed cans;

• a passed can is no different than no can i.e. than an empty spot on the conveyor

Fixed Geometry

• Constant distance
  • from the position of a can at the camera inspection station at the time of that can's fail detection,
  • to the position of a failed can where the reject pusher initially contacts that can to reject it

Fixed Time Dynamics

• Inspection camera
  • Camera timing relative to inspection proximity sensor will be variable, but stochastic with a consistent mean and a small variance
  • Camera timing is independent of linespeed
  • There is a positive time delay, ΔTinspect, between the time the prox sensor detects a can entering the fixed position of the inspection until the camera returns a failed-can signal to the PLC
  • That time delay ΔTinspect, plus the linespeed, determines the position on the conveyor that is modeled by the first (right-most) bit in the bit-shift array.
    • That position modeled by the first bit determines the distance, and number of bit-shift array elements modeled, a failed can must traverse to get to the reject pusher.
• Pusher
  • Pusher movement and timing are independent of linespeed
  • There is a positive time delay, ΔTpusher, between the reject signal from the PLC and the moment when the reject pusher initially contacts a failed can
  • The reject signal from the PLC must be sent at that ΔTpusher offset time before a failed can reaches the position of the pusher's initial contact with the failed can

Adequate resolution in the PLC encoder counting and bit-shift array

The distance a failed can travels between when the PLC receiving the failed-can input signal, and when the reject signal from the PLC should be sent, is equal to

• the distance from the inspection trigger prox to where the pusher's initially contacts the fail can,
• minus the product ((ΔTpusher + ΔTinspect), s ×linespeed, inch/s)
• and that shortened distance is what is modeled in the bit-shift array elements' positions, from element 0 to some element N.

The value of the index in the bit-shift array where a 1-bit should trigger the reject signal from the PLC is linear with the calculated RPM, so

• IF we know, or can empirically determine
  • the highest and lowest operational speeds of the motor
  • the two indices in the bit-shift array at which to trigger the reject signal at those two speeds
    • HISPEED_OFFSET_SP for the highest operational speed of the motor
    • LOSPEED_OFFSET_SP for the lowest operational speed of the motor
• THEN
  • a linear model in the PLC can calculate the index of bit-shift array to observe to trigger the reject signal
    • the linear model is based on
      • Measured motor speed (from encoder counts vs. time)
      • highest and lowest operational speeds of the motor
      • HISPEED_OFFSET_SP and LOSPEED_OFFSET_SP

original/

• Original code and documentation
• Annotated (i.e. enhanced comments) version of the original code

versions/

• Versions of ports of original MicroLogix 1100/RSLogix 500 code to Micro820/CCW, plus some other test code
• See versions/README.md for details

edge_case/

• Demonstration of MicroLogix 1100/RSLogix 500 edge cases in failed-can signal timing that cause original code to fail
• Also two solutions
• See edge_case/README.md, plus comments in code, for details