Jørgen wrote: I failed to recognize that the gear change only too place on the OFF to ON transition, this may work. How do I (from the .comp code) prevent the cycles to continue while the gear change takes place?
You would probably need to re-code the .comp as a state machine. State machines in .comp look a bit like this, typically
Code:
...
parameter unsigned state
...
;;
FUNCTION(_){
switch state{
case 0: // Waiting for spindle off
if (! spindle_on){
state = 1
}
case 1; // Waiting for spindle on
if (spindle_on) {
if (some sort of logic to choose gear 1) {
gear1 = 1;
gear2 = 0;
gear3 = 0;
gear4 = 0;
state = 11;
break;
}
if (some sort of logic to choose gear 2) {
gear1 = 0;
gear2 = 2;
gear3 = 0;
gear4 = 0;
state = 12;
break;
}
....
case 11: // Waiting for gear 1 engagement
if (gear_1_engaged) {
state = 21;
}
break
case 12: // Waiting for gear 2 engagement
if (gear_2_engaged) {
state = 22;
}
break;
...
case 21: // gear 1 speed output
speed_out = speed_in * scale_1;
if (! spindle_on){
speed_out = 0;
state = 0;
}
....
}
The "..." show places where code for the other gears needs to be added.
The interface to the original Fanuc drivers was 0-10V and the encoders were connected to the Fanuc controller. After spending a long time to repair the drivers and debugging the Fanuc/Mazak interface logic it tuned out that the controller also was defect. I then decided to replace everything and started with the servo drivers. The drivers are now based on the YAPSC project having STEP/DIR interface, and seems to work ok.
Rather than struggle with limited parallel port IO (and 5V signal levels) you might want to look at something like a Mesa 5i25 / 7i76 combo to add fast step generators and a lot more IO. It will make things easier. Or carry on using the parallel port and connect a Pico USC to the parallel port to achieve the same thing. (16 digital inputs and 8 SSR outputs)
