Proximity / limit switches

[3d_danny]

I am putting the final touches on my Alpha and installed limit switches yesterday.

I would be curious to know why Shopbot put a normally open circuit for a limit switch?

All of the cnc machines that I ever built or worked on had normally closed switches on limit switches as a failsafe. If the switch is normally closed, it has the ability to always check its own wiring for damage(open circuit), connection problems or faulty switches. If the wiring has damage and goes open or the switch malfunctions, the machine will stop until the problem is repaired. With a normally open circuit, you are somewhat 'flying blind" and hoping the circuit is functioning.

Hitting the stops is a disconcerting sound and it will always be somewhat of an concerned moment during a homing run without NC switches.

[gerald_d]

Dan, until relatively recently, even the E-Stop was a normally open circuit. That is now changed. I get the impression that the "limit" switches are used more for finding a fixed position than what it is for safety. If you really want safety switches, then you should maybe put them in series in the (later) E-stop circuit?

[dvanr]

I agree Dan,

The way it is now the Hall effect switch is wired in with unshielded cable. Which seems to be causing some grief as well.

There is a way to wire a N.C. switch in with a 47k resistor and get the same output signal as a hall effect switch. The benefit would be as you pointed out " failsafe" and also less prone to false triggering from emi noise.

[scott_smith]

Dan,
If you change to NC contacts you will need to use two inputs to the controller rather than one. Currently the program runs the X off the limit before checking the Y on the same input. The sensors are wired in parallel.

[fleinbach]

Scott,

Changing to NC contacts would not necessitate two inputs. A normally open circuit can be run in parallel a multitude of times. Each trigger mechanism whether solid-state or manual is detected when it closes.

You will get the same results with a normally closed circuit having 2, 3, or more switches. Each time any one switch opens it will be detected by the sensing circuit.

[scott_smith]

Wire them in series. That would work. Wiring nightmare, but it would work
Thanks Frank and sorry all.

Not sure if the oninput command in the zero program could be setup with the opposite logic. But I’m fine with my NO contacts.

[dvanr]

If the logic doesn't work for the oninput command or you can't change it, then use a pull up resistor at the signal pin (input port).

With the switch closed , the controller will see zero, switch opens and it sees a high ( 5 volts).

[ted]

Interesting thread and some good comments. It may make sense to move towards a normally-closed system for the limits. We'll put it on the list.

Historically, the limits and prox switches have been optional on ShopBots and that is the reason that they have not been viewed as needing to be obigatorily present (as in requiring the NC confirmation of function) in the way the present E-Stops are. They have been optional to keep costs down, and because, quite frankly, before the PRTalpha, you couldn't really hit the end stops hard enough to do any damage. The switches have typically been primarily used as proximity switches for homing and setting up the tool. [Note that the current homing application reduces speeds so that even if the switches were not present or failed the colision would not be severe.]

The speeds of the PRTalpha tools, of course, put us in a different ball park, and Dan's comments are well taken. One thing that is worth pointing out is that the 'Software Limit Checking System' works well, though it does require using (and keeping current) the Table Base Coordinates. ShopBotters tend not to bother with this, but if you want to make sure you do not slam into the ends, this is the best way because you are stopped even before you get up to full speed. Let the computer watch out for you ...

Ted Hall