[UA_Iron]

You're looking to have relay control synchronized with movements or other actions?

Is it coordinated motion that you're trying to do; things like perfectly circular movements requiring two or more synchronized axes is what I'm talking about. Or are you going to points in space, and dropping components off based off of other triggers (repeatable points)?

If you don't need coordinated motion, then a PLC is probably your best option. If you need coordinated motion, then you will need a motion controller to synchronize the axes.

At my work we use a 2 axis pick and place X-Y robot stage. We use Schneider Mdrive Plus2 stepper motors in the machine communicating over RS422. An ASCII command line is sent to the robot where to move (pulling location information from an xml file), this command line contains all the motion parameters: distance to move, acceleration, velocity etc. The integrated controllers in the motors handle the rest of the motion routines - converting distance to steps, looking at I/O from sensors, checking movement with encoders etc. Since the axes are not synchronized, once one axis has moved, the other is allowed (or controlled) to move.

The above requires a software front end at some point, while a PLC could achieve this without this front end as well as dealing with event based relay switching etc.

I'm not sure if mach3 can handle I/O coordinated with movements... but since it has I/O capability, I don't see why it couldn't - afterall, this is how the limit switches stop the machine or axis. There are other ways to solve this problem and still use mach3 though. If your stepper or motion drivers are capable, you can program I/O and routines from them, while driving them with the pulse train from mach3.

[Neosec]

You're looking to have relay control synchronized with movements or other actions?

Is it coordinated motion that you're trying to do; things like perfectly circular movements requiring two or more synchronized axes is what I'm talking about. Or are you going to points in space, and dropping components off based off of other triggers (repeatable points)?

If you don't need coordinated motion, then a PLC is probably your best option. If you need coordinated motion, then you will need a motion controller to synchronize the axes.

At my work we use a 2 axis pick and place X-Y robot stage. We use Schneider Mdrive Plus2 stepper motors in the machine communicating over RS422. An ASCII command line is sent to the robot where to move (pulling location information from an xml file), this command line contains all the motion parameters: distance to move, acceleration, velocity etc. The integrated controllers in the motors handle the rest of the motion routines - converting distance to steps, looking at I/O from sensors, checking movement with encoders etc. Since the axes are not synchronized, once one axis has moved, the other is allowed (or controlled) to move.

The above requires a software front end at some point, while a PLC could achieve this without this front end as well as dealing with event based relay switching etc.

I'm not sure if mach3 can handle I/O coordinated with movements... but since it has I/O capability, I don't see why it couldn't - afterall, this is how the limit switches stop the machine or axis. There are other ways to solve this problem and still use mach3 though. If your stepper or motion drivers are capable, you can program I/O and routines from them, while driving them with the pulse train from mach3.

A bit of both actually. I'm looking to have two axes move a part in space and bring it to a two axes CNC router (spiral saw). The spiral saw will cut a circular hole. The part will rotate to another side (by another stepper) and the saw will cut another hole. Rinse and repeat 4 to 6 times. The part will then be places in a stack and the machine will get the next part from a stack. The parts are picked up by vacuum (the need for the solenoids) checked at another station for double or triple stack (stuck together when picked) using an apposing vacuum source (another solenoid) then brought to the spiral saw... etc, etc.

I think the hardest part of the whole process is if there is a double (two or more stuck together) when the pick occurs. It would require jumping to a sub routine to deal with the error before proceeding with the cutting and placing process.

Five axes in all, plus three vacuum solenoids and a spindle. May need a blower too to deal with the debris.

I see breakout boards like this on eBay, as an example, that have plenty of outputs which, if I'm not mistaken, can be controlled with M-code. I've also read some things that mention the G-code can do conditional routines. I've never written a line of G-code in my life but I think since I'm only moving to about 16 points in space and cutting a single hole over and over, it shouldn't be to bad.

I don't know anything (much) about PLCs. I just have a feeling that CNC (Mach3) works better (easier?) with stepper motors and PLCs work best with analog and digital IF - THEN - ELSE logic sort of things. I may be very wrong. PLCs definitely seem to be more expensive to set up 5 axes stepper control, but the conditional issues would be much easier.

Neo

[UA_Iron]

A bit of both actually. I'm looking to have two axes move a part in space and bring it to a two axes CNC router (spiral saw). The spiral saw will cut a circular hole. The part will rotate to another side (by another stepper) and the saw will cut another hole. Rinse and repeat 4 to 6 times. The part will then be places in a stack and the machine will get the next part from a stack. The parts are picked up by vacuum (the need for the solenoids) checked at another station for double or triple stack (stuck together when picked) using an apposing vacuum source (another solenoid) then brought to the spiral saw... etc, etc.

You may be able to use both, you'll just need certain triggers to trigger the next process to happen. A sensor switching high or low is always a good point.

Instead of an opposing vacuum source to check for double or triple stack, could you just use a pneumatic cylinder to knock them off? I prefer pressure over vacuum, especially in a dusty environment - you'll never need to service to suction cups with a pneumatic cylinder.

They make integrated relay boards for Mach3 breakout boards
CNC4PC

I think the hardest part of the whole process is if there is a double (two or more stuck together) when the pick occurs. It would require jumping to a sub routine to deal with the error before proceeding with the cutting and placing process.

Punch them all out of the way with a pneumatic cylinder over some kind of slide leading to a bin!

Five axes in all, plus three vacuum solenoids and a spindle. May need a blower too to deal with the debris.

I see breakout boards like this on eBay, as an example, that have plenty of outputs which, if I'm not mistaken, can be controlled with M-code. I've also read some things that mention the G-code can do conditional routines. I've never written a line of G-code in my life but I think since I'm only moving to about 16 points in space and cutting a single hole over and over, it shouldn't be to bad.

I would not buy any USB based breakout board besides a smoothstepper. A lot of the stuff on ebay is crap too, buy good boards that people can vouch for on the forums, it's not worth the headache to save a few bucks and buy the ebay stuff.

A USB or ethernet smoothstepper allows you to connect multiple breakout boards so you have a ton of I/O capability.