Control of a Cole-Parmer Masterflex peristaltic pump

B

Thread Starter

Bob Lewis

The pump I will be controlling is a the model number Masterflex EW-7523-40. This is a pump made by Cole-Palmer. We are going to be setting up a color control system on a laboratory papermachine. We are using several devices that use RS-232C communications. We are using a serial-ethernet terminal server due to the layout of instrumentation and controlling pc hence the desire to control the pump through a serial port. The pump has a DB15F connector on the back. I know I need some type of external i/o device to interface the pump with the serial connection that has DAC, ADC, and relay capabilities.

I need to be able to trigger 3 contact closures all of which have a common ground. The pump output is selectable from 0-20ma, 4-20ma, or 0-10V. The input has the same choices and finally I have the option of either an NC or NO contact for monitoring if the pump is running. If I had the time and experience I imagine I would build some type of custom PIC to handle the job. What I am looking at doing is using a device that has analog i/o and then adding on relays to handle the contacts.

This is all new to me so I was looking for any suggestions you might have. Specifically I was wondering if there was a way to handle the contact closures other than using the relays and how to handle monitoring the NC or NO contact. Thanks in advance!

Bob Lewis
 
C

Curt Wuollet

Sounds like a job for an SBC with the right IO.
But you can do this with the minimum effort
by using a full OS so all you have to do is
write the actual control bits. The Sixnet RTU
would work but is probably overkill. The comm
flexibility needed argues against a PLC. I think
your first instinct was right, but it would be
fastest to write in a language/environment you
are familiar with. For my part, that would be
a Linux PC or an SBC powerful enough that it
programs just like a PC. Perhaps Diamond
Systems or Tri-M Engineering. They can supply
a board alreasy booting and ready to program.
Spendy, but it would drastically reduce the
programming time VS learning something new,
so the total cost would be minimized,
Your choices might well be different, but I
would have a high degree of confidence that
it could be done and any surprises handled
with that approach. I would have much less
confidence with secret (proprietary) stuff
unless I'd solved similar problems with it.

Regards
cww
 
Top