Analog Instructions or Methods to Control Proportional Hydraulic Valve

Hi,
I am an inventor looking for a way to automate a block-making press I made out of a log splitter. It makes blocks for building homes from recycled paper. I am leaning toward Allen Bradley products because of their huge installed base and because used PLCs and parts appear to be readily available on Ebay. My press video is located on YouTube at http://www.youtube.com/watch?v=yIqTBQA8_Yo


I've learned some things by asking a similar question on this forum before, but am still a bit stymied because some of the answers seemed to indicate that I should have to buy expensive add-ons to the PLC to accomplish what I want to do. It seems to me that what I want to do should be fairly straightforward PLC work. Maybe I'm missing something.



What I want to do is:
1.) move the fully retracted ram at normal speed forward for several seconds,
2.) slow it down well before it approaches its stop (block thickness) point,
3.) stop it at that point within +- 1/32 inch
4.) retract it a tiny bit (an inch or less) to take pressure off door,
5.) (door opens)
6.) move ram forward again normal speed a set distance to push block out of press,
7.) ram then stops and retracts all the way, stops and waits for press to fill

What I need is an idea of the type of AB analog instructions that would be used to talk to a solenoid-operated proportional hydraulic control valve. I would hope that I could use simple reed switches and timers for start, slow down, stop, reverse, stop, forward, stop, retract, stop. What I’m hung up on is the AB analog instructions to the proportional valve to change speed, reverse direction, etc. The valve will most likely be +-10V variety.

Anything to get me going in the right direction to find the PLC analog instructions needed to control the valve would be appreciated.

Do I really need expensive add-on speed controls, etc.?

Thanks.
Barry

This is a pretty simple application that can be controlled in open loop but it does look like you should have some sort of position feedback so you know when to slow down or stop.
You need an analog output -10v to +10v for the valve and an analog input 0 - 10v for the position sensor. A small PLC should be enough.
It can probably be done much faster and more accurately with a real motion controller but I think that would be overkill.

There aren't any analog instructions. You change the analog output by writing to the analog output register. The analog output register may be scaled so that -16384 counts is -10volts and 16383 counts is 10 volts. You can write any other number in between. The output range will depend on the PLC. Some analog outputs are only 12 bits so the range would be -4096 to +4095 for -10v to +10v. The position can be read from the analog input register. It will also be a number that will be 0-4095 for a 12 bit input or 0-16383 for a 14 bit input. Then the analog counts need to be scaled using the SCP block to get positions.

Could Reed Switches Be Used to Trigger the Valve Changes?

Thanks Peter. I'm starting to understand. I recognize your name from an earlier question too.

Hopefully one last set of questions. Can the position sensors be reed switches instead of an analog device? On my press, a rod is connected to the back of the piston. The rod travels with the piston and is pretty stout and stable. It protrudes out of the rear of the press in a sheltered area. (I show it working in my video.) I'm using it now as my visual "thickness sensor".

I was thinking of mounting a magnetic collar on it, which would move with it, and positioning reed switches along its travel, which duplicates the distance traveled by the piston inside the press. So I thought the reeds could fire the speed change and the 1st stop, maybe a timer could trigger a short reverse and a resume forward. Then a reed would trigger block out of press and retract. Is that at all workable?

Barry

thinksabunch said:

Thanks Peter. I'm starting to understand. I recognize your name from an earlier question too.

Hopefully one last set of questions. Can the position sensors be reed switches instead of an analog device? On my press, a rod is connected to the back of the piston. The rod travels with the piston and is pretty stout and stable. It protrudes out of the rear of the press in a sheltered area. (I show it working in my video.) I'm using it now as my visual "thickness sensor".

I was thinking of mounting a magnetic collar on it, which would move with it, and positioning reed switches along its travel, which duplicates the distance traveled by the piston inside the press. So I thought the reeds could fire the speed change and the 1st stop, maybe a timer could trigger a short reverse and a resume forward. Then a reed would trigger block out of press and retract. Is that at all workable?

Barry

Click to expand...

Sure you could probably accomplish what you need to do with reed switches, but by the end of the project you are going to wish you had sprung for the analog sensor

Having 4 or 5 reed switches that the press depends on to stop safely and properly just gives that many more opportunities for failure.

Plus if you want to fine-tune any positions of the ram, you have to stop the process, get in and fiddle with an adjustment, then start back up, check, adjust, check, adjust, etc. If you have analog feedback, then a simple rung edit (or HMI adjustment) can dial - in your process on the fly.

Some of the cheaper options of linear transducers that were suggested in the earlier thread could probably be had at or lower than the cost of a handful of switches.

Just my $.02

Good luck,
Dustin

Which linear transducer would work?

My problem with selecting a workable linear transducers is how to install them and their cost. I've looked at string pots and been warned away from them. They actually seemed most appropriate for what I'm trying to do, but someone said they go slack in heat and change length over time. I'm working outside in Arizona sun. It can and does hit 120 once in a while.

The stroke of the piston is 30 inches - so shorter-acting transducers won't work and long ones cost a bundle. I cant install anything inside the hydraulic cylinder- it would be impractical. Someone suggested a linear decoder but I would have to figure out a way to accurately machine a three foot long, gear-toothed rod for that. Expensive and subject to misalignment I'm told.

Anyway, if there is a silver bullet out there in terms of workable analog input position device, I would like to hear about it. It is also true that I don't fully understand some of these options, so maybe the solution is in front of me and I just don't recognize it.

Thanks,
Barry

You could use 1 limit type switch attached to the ram and have a series of fixed dogs to trip the switch at different locations of travel the logic should be pretty straight foward

Limit Switch and Fixed dogs

I like this concept but how does the PLC know which dog is tripping the switch? Isn't the switch confined to a single input? Is it possible to set up some kind of sequence which the PLC keys to specific dogs? I've read and studied Phil's book and Bolton's 5th Edition but I don't have practical experience.

Actually, you could use 2 limit switch, one for each direction. They could be wired to up&down counter. The counter must be retentive.

Barry,

I once built a wood splitter that I used on my farm until I got too old to cut wood. For what you are doing, use the prox or reed switches located at different positions along the cylinder. Mount the switches so that they are easy to move slightly in either direction (slots or a track mounting). It will be easier for you to adjust the switches than to make changes to the PLC program.

After reading all of your posts, one important question should be answered: Why do you need such a close tolerance for your blocks? I know that most building blocks and bricks have a lot of variation, 1/8 to 1/4 inches. Also, I think that if you are pressing water out of paper, that there will be a lot of uncontrollable block re-expansion after the press (that is probably independent of how close you pressed to a certain dimension).

The sequence would keep track of which dog should come next you could have a first scan bit reset the sequence and retract the ram on power up Keep it simple

I'm thinking proximity switch is the way to go. Each one goes to a separate input on your PLC, so you write your program accordingly. They can be easily adjustable ( think slotted mounting) and should repeat plenty good for what you're doing

proportional valves are expensive, so why not stick with your lever or coil activated controls.
The endpoint of pressing is a simple rollerswitch
then retract for 1 second no problem where it stops, open doors and push all till the end of the piston.
for the door i would also use a hydraulic cylinder to keep it closed.