Using PLC to control hydraulic movement

[ozarkgem]

The OP doesn't need a top of the line control system but he does need one that is good enough. I could do the control in the PLC but I have a lot of experience at this and that makes a HUGE difference. The application isn't that difficult. The problem is that the gains are different going up and going down. The PLC has no target generator.
I just got back from China. Over there they use simple PID control and limit the output to get the speed limit. This way they don't need a target generator. Feed forwards aren't commonly understood or used.

An RMC75S-QA1 would probably be overkill, but I could get it running in just a few minutes.
http://deltamotion.com/products/motion/rmc70/
The PawTawJohn control relies on using a MDT so that is out.

There is a sawmill trade show in Portland this week.

the forward speed and all other controls are manual hydraulic valves. Would like to have seen the trade show. Pretty fair skip from here.

 

[redbarron55]

Perhaps an encoder on the system and a simple counter to position.
He does not need positionability, but repeatability and the count can be determined by moving and counting.
The count can be determined for each position and the hydraulic stop valve be closed at the several points needed as determined.
A simple pointer to the desired count would be all that is necessary.
A PID might be overkill for this job.
If he positioned every inch the count for each inch determined and stored for the counter/positioner.
A hydraulic stop valve would be pretty solidly held in position.
for 12 inch coverage twelve counts.
Those counts may not be evenly distributed due to geometric inconsistencies.
Each could be changed to fine tune the system.
A little math would let you change one and add/subtract the difference from the others.

 

[Peter Nachtwey]

the forward speed and all other controls are manual hydraulic valves. Would like to have seen the trade show. Pretty fair skip from here.

You should need only one valve to control the motion up and down.
You should use a proportional valve with a zero over lapped spool or a real servo valve.

At the saw mill show I am giving a presentation about why a P, PI, or PID is not adequate for best performance.

 

[ozarkgem]

You should need only one valve to control the motion up and down.
You should use a proportional valve with a zero over lapped spool or a real servo valve.

At the saw mill show I am giving a presentation about why a P, PI, or PID is not adequate for best performance.

I know zero about this. What are P,PI and PID?
I though maybe a pilot operated check valve for simplicity.

 

[kamenges]

Originally posted by ozarkgem:

Doesn't have to be deadly accurate...

Can you put some numbers around that? +/-0.25" or +/-0.025"?

Peter Nachtwey works with systems that move very fast very accurately with very small settling times. This is likely not what you are looking for but you really haven't said.

If this is something you have a lot of time to play with and just want to learn something new then we can keep going as we are. It would be helpful if you could put some actual specs out there, though. If this is something that is actually going to make you money, or worse yet is costing you money until you get it completed, you may be better off finding someone local to come in and help you out. Being able to stand in front of something and talk about it has some benefits.

If you don't already have a sensor (you have been talking encoder) or, just as importantly, the linear to rotational transmission that will be required with an encoder, you may be better off looking into the linear sensor options people have been referring to in previous posts. They are much easier to deal with if you don't have the rotary driver already in place.

Keith

 

[ozarkgem]

Can you put some numbers around that? +/-0.25" or +/-0.025"?

Peter Nachtwey works with systems that move very fast very accurately with very small settling times. This is likely not what you are looking for but you really haven't said.

If this is something you have a lot of time to play with and just want to learn something new then we can keep going as we are. It would be helpful if you could put some actual specs out there, though. If this is something that is actually going to make you money, or worse yet is costing you money until you get it completed, you may be better off finding someone local to come in and help you out. Being able to stand in front of something and talk about it has some benefits.

If you don't already have a sensor (you have been talking encoder) or, just as importantly, the linear to rotational transmission that will be required with an encoder, you may be better off looking into the linear sensor options people have been referring to in previous posts. They are much easier to deal with if you don't have the rotary driver already in place.

Keith

1/16" accuracy. I posted the video so you can see the head speed. Not real fast up and down. All the commercial mills in the category like mine use rotary encoders. Must be a reason. I would love to find someone local. I think this is beyond me even with all the knowledge from everyone here.

 

[Geoff White]

If you are talking typical rough cut sawmilling gear it can be a very simple system.

Start wide and always approach the stopping position from the same direction.
Use the encoder to decide to stop feeding when it is the average stopping distance from the desired position.
Monitor how much distance it takes to stop each time and update your average stopping distance (with an exponential average) so the error does not creep with time and mechanical changes.

If it overshoots/undershoots too much then lower and try again - this time with the improved stopping distance.

This is extremely dumb but is robust and easy to implement.

 

[Engineer_MG]

Hi,

I recently developed code for a Hydraulic shear machine for metal cutting (Max. 1 inch thickness) with automatic Rake angle control (vertical movement and at an angle) and automatic stroke length control (vertical movement). We used 2 Linear transducers (0-10 V output , 0-310 mm stroke length)

Rake angle was a variable generated based on material thickness
Stroke length was a variable generated based on material length
We also had a servo drive backgauge for material width

The basic logic for hydraulic controls in this machine:

1. Scale the analog output , say 0-10 V to 0-xyz mm/inch and Linear transducer (going to the PLC as analog inputs). We have an analog scaling function block in the Schneider Electric software.

2. By trial and error figure out the tolerance window (abc millivolts)
Command = 1 inch , converted to xyz Volts(millivolts for better resolution)
Enable hydraulics
Monitor Analog output of the linear transducer via Analog inputs of PLC
If Linear encoder value is within +- abc millivolts then the position has reached the target.

Of course I used Schneider Electric PLC, HMI, Servo, IO . The linear transducers were from Balluf.

Let me know if you have any questions.
Thanks.