Analog modules working similar to servo ones.

My customer wants to implement a way to drive an hydraulic cylinder, this is what he got:
Module 1756-IF8 as differential input, this module is connected to a balluf transducer 0-10 VDC as cylinder position feedback. Balluf transducer is located inside the cylinder.

Module 1756-OF4 connected to a servovalve to drive hyd. cylinder.

I have not quoted because i'm not sure how to attack the problem, basically I'm thinking something similar to an example I saw the other day where they use a powerflex similar to a motion device. My idea is to create an AOI(s) to handle movements, similar to motion control instructions.

He wants to change speed, accel, decel and position of cylinder.

I do not want any program or something similar, I want some ideas on how to try to attack the problem in order to make an initial quotation.

Normally I would use a 1756-Hyd2 as a solution, but He has this two modules and wants to use them, they were used in the original design from Rockwell in 2004. I designed the rest of the software for the machine and now we're using and ON/OFF hydraulic valve to drive cylinder and use limit switches to indicate position.
I'm using Rslogix 5k, version 19, my cpu is 1756-L63.

Regards,

william

by the time you get done programming the AOIs you will have spent enough time to buy 10 hyd02s at least and the control will not be as good

Peter Nachtwey said:

by the time you get done programming the AOIs you will have spent enough time to buy 10 hyd02s at least and the control will not be as good

Click to expand...

Yes, that's what I thought and that's the reason I have not quoted yet.

widelto said:

Yes, that's what I thought and that's the reason I have not quoted yet.

Click to expand...

Good, so I don't need to convince you but to help you convince your potential customer

PLCs shouldn’t be used to directly control closed loop hydraulic servos systems.
1. PLC scan times are too slow
2. PLC scan times are not consistent. Adding more code to the PLC makes the scan times slower and more inconsistent.
3. PLC analogue inputs and outputs are slow and may not have enough resolution.
4. PLCs do not have a trajectory planner. It must be written.
5. Diagnostics and debugging tools are lacking.
6. The PLC PID is written for process control applications. Time constants are in minutes instead of seconds or milliseconds.
7. PLC's do not have feed forward control to reduce following error.
8. Most important! Writing optimal hydraulic servo control code requires understanding hydraulics, control theory and programming.

Peter:
Excellent explanation.
I'm going to try to convinced my customer to use a 1756-hyd02 module instead, or let them keep on using ON/OFF valves with movable limit switches. They want to save some seconds every cycle and you know that means a lot of money at the end of the day, normally they have a cycle every 35 seconds. They expect to save up to +/- 2.5 seconds every cycle. That means production will increased somewhat near 5 - 6%.

Another argument

Time taken for me to create the program, test and commission to just perform the control is 1 week (pick an number that costs more than a HYD unit) and the performance of the control will NOT be the same as using a HYD02 it will be worse.
this is before adding the application control on top.

This is an obvious "Add the correct hardware to save engineering time"
you get a better result and save time

Thanks Michael.

I'm actually having to implement some rudimentary motion control myself using CompactLogix, a linear transducer, and a proportional valve (even better, both the transducer and valve are going through Point I/O over Ethernet). (We value component cost much higher than engineering time). While it's okay for rudimentary moves it's not very good. Consistency is a real pain as it will do different things depending on when things want to update.

Now to convince them to spend the money to go to a Delta RMC!