Help with a Flip/Flop Type of Logic

[chelton]

This would work. Cylinder would always retract first if in unknown position.

 

[chippieboy]

Not bad Chelton, but if the cylinder coasts and stops at the fully retracted position, we might have an issue. This machine is a printing press (not that that matters) but the hydraulic pump is powered by one of the spinning cylinders/shafts and when the machine stops it will coast sometimes stopping on a limit switch.

 

[chelton]

Not bad Chelton, but if the cylinder coasts and stops at the fully retracted position, we might have an issue. This machine is a printing press (not that that matters) but the hydraulic pump is powered by one of the spinning cylinders/shafts and when the machine stops it will coast sometimes stopping on a limit switch.

Do you mean the cylinder coasts pasts the retract limit? The code I posted will not be affected if the limit switch is made when the machine stops.

 

[drbitboy]

Not bad Chelton, but if the cylinder coasts and stops at the fully retracted position, we might have an issue. This machine is a printing press (not that that matters) but the hydraulic pump is powered by one of the spinning cylinders/shafts and when the machine stops it will coast sometimes stopping on a limit switch.

Hmm, I don't see the problem. Chelton's code will never start up toward either limit switch if that switch is a 1.


Maybe there is a misunderstanding of what is meant by

Cylinder would always retract first if in unknown position.

i.e. the misunderstanding involves missing the significance of the highlighted clause i.e "if in unknown position" means "the ram position is in between the limit switches, i.e. both limit switches are 0?"


The only possible problem would be, as Chelton asks, if the ram coasts past the retract limit switch that is an (optical?) prox switch and the switch signal becomes a 0, but if that is the case, then any algorithm proposed to solve this thread would have the same problem with one switch or the other: having both limit switches as 0s mean that the ram is between them is, I am pretty sure, a necessary condition for this to be automated.

 

[Phrog30]

Hmm, I don't see the problem.

Me either. Whether it stops on the switch or goes past.

 

[chippieboy]

Yup guys, now that I slow down and take a good look, everyone is correct. I'm guessing the cylinder will retract first because the retract rung is first in the scan. Is that correct? Now just out of curiosity I will set up a test PLC and try them all but Chelton's will probably go into the real program. Thanks for everything !!!!!

 

[drbitboy]

Yup guys, now that I slow down and take a good look, everyone is correct. I'm guessing the cylinder will retract first because the retract rung is first in the scan. Is that correct? Now just out of curiosity I will set up a test PLC and try them all but Chelton's will probably go into the real program. Thanks for everything !!!!!

Yes, the order of the rungs determines which direction is chosen for the "both limits are 0" case, assuming the PLC obeys the "left-to-right, top-to-bottom" mantra*.

And I agree with your choice: four contacts and two coils cannot be beat!

* Apparently Siemens does not, in certain situations, according to a recent thread.

 

[RoTaTech]

Just to throw a little figurative monkey wrench into your program: if you are using a double-acting valve, then when you (nearly) instantaneously switch one output on and the other off, there is a brief period when there is magnetic field on both sides of the solenoid, so they are both pulling, ergo they are fighting for the spool. I usually put a small delay between the turn-off A and the turn-on B, and vice versa - it is kinder to the valve.

 

[drbitboy]

Just to throw a little figurative monkey wrench into your program: if you are using a double-acting valve, then when you (nearly) instantaneously switch one output on and the other off, there is a brief period when there is magnetic field on both sides of the solenoid, so they are both pulling, ergo they are fighting for the spool. I usually put a small delay between the turn-off A and the turn-on B, and vice versa - it is kinder to the valve.

Ooh, noice.


So a TON on a branch in parallel with each coil, with each .Done bit driving a solenoid side?

 

[chippieboy]

Yes, RoTaTech does have a valid point and that scenario is correct with the valve.

 

[RoTaTech]

Ooh, noice.


So a TON on a branch in parallel with each coil, with each .Done bit driving a solenoid side?

That's exactly it, Dr. BB.

 

[plvlce]

I usually put a small delay between the turn-off A and the turn-on B, and vice versa - it is kinder to the valve.

Out of curiosity, how long a delay do you usually use?

 

[kalabdel]

3 way 2 position valve with no limits and timers for control.

 

[Phrog30]

3 way 2 position valve with no limits and timers for control.

Same. I'm sure some valves are different, but most you don't need delays for the valves sake. If anything you need dwells and delays for the mechanics. Same thing for FVR starters.

 

[OkiePC]

Same. I'm sure some valves are different, but most you don't need delays for the valves sake. If anything you need dwells and delays for the mechanics. Same thing for FVR starters.

I agree with RoTaTech. A small delay after de-energizing one solenoid before energizing the opposite solenoid of the same spool is gentler on the solenoids and the spool. This practice was recommended to me by MAC Valves. I used a 50ms delay.