Comunication COM1: and a CNC milling machine (timeout?)

[Stephen Quinn Barncard]

Welcome to the world of machine interface. I've done this sort of 
thing a lot, in fact my first hypercard project was a 
software-hardware interface to a bank of 128 cassette machines in 
1988. And I built the hardware end as well and it was still difficult 
to get them to talk to each other.

It's a bitch.

Even at slow rates, you'll find yourself ripping your hair out at the 
unresponsiveness of old serial interfaces. Many of them run without 
handshaking of any kind. Sometimes the connections fail and you have 
to restart both the target machine and the computer. You'll find 
yourself padding commands with delays just to get stuff to work.

And if you don't know the right commands, you could be out of luck.

One thing I might suggest is that you hook up the machine with it's 
regular controller, but tap the send line with a breakout box and 
attempt to view the commands that are sent with a terminal program 
(instead of Rev for now). If echo is set on the target device you 
should see both ends of the 'conversation'. This will give you a 
start.

After that testing commands using a terminal program rather than rev 
will speed up your development. Sometimes people who write the 
imbedded software for controllers will put in a help menu - try 
sending an "H" to the controller.

And if you have to develop this thing with your dad in the room, it 
will take forever. What you need is your own machine on your own 
bench to do this in your own time. I've found that working at a 
distance is almost impossible with this kind of project.

Plan on spending a LOT of time on this one. The REV interface will be 
the fun and easy part.

good luck


sqb



>Hi Dar, Sarah and all,
>
>Developing this stack is a bit complicated, as I need to write it in my
>office, compile it and then test it in the company of my father.
>I got the manual of the machine today but have not found any useful hints on
>communication with a PC.
>