Measurement Sensors and PLC's

[ander88]

Hi Everyone,

I've been an electrician for just over 6 years in a packaging processing plant.

I have been asked by my manager to design a system using measurement sensors, moisture sensors and an output printer.

I have found various sensors that will measure these and give me a 0-10V or 4-20mA outputs as well as switching points etc.

What my question is and it is probably not an easy one to answer as its quite a big ask....


Ill try and describe the application.

We produce corrugated board and a machine loads these sheets onto a pallet creating a stack.
During production water, starch and PVA are used to create the corrugated part. Sometimes the boards come out mishapen instead of flat.

We want to use two measurement sensors looking down onto the top of the pallet to measure the distance between the sensor and the board then calculate if there is any difference between the two values.

If the value is within a set acceptable range we would like a printer to output a label saying pass, otherwise the printer to output a fail label.

The same with the moisture meter. Read the moisture in pallet and compute to give a pass or fail.

I have no experience atall with PLC programming so the question is can anyone recommend a PLC that can do something like this that we can program easily etc.


Many thanks, Michael

 

[itsmoked]

Wow.. I don't know where to start.

You are talking about a very tall order for a guy who's never even programmed a PLC. My experience sez your project is going to crash and burn many times over. There are so many show stopping gotchas lurking around this design. You guys aught to spend YOUR time specifying exactly what you want instead of trying to cobble this up.

Figure out what you want for a human interface.
The range of acceptance for your boards and how to change them.
The available power supply.
What electrical outputs you might want for controlling other machines.
What signalling you might need because X boards in a row failed.
Whether you need to look at the whole sheet or can instead scan the sheet as it goes past a point. In other words, can you line scan it instead of looking at the whole board.
What kind of enclosure you need.
And for cryingoutloud!! What you can use instead of a ridiculous printer to print out stoopid pieces of paper. Talk about a major failure point for a production application! This alone demonstrates you guys need some local talent to help you with this project.

So spec exactly what you guys want,(see above), then shop for someone to help you with it. It will cost far less in the long run.


Keith Cress
kcress -

http://www.flaminsystems.com

 

[danw2]

Wow. And Keith didn't mention that the measurments could be a project in and of themselves.

"moisture in a pallet"? How the probe get into the pallet? One shot, or does moisture continuously change over time?
Are the pallets stationary or moving?

What about the distance measurement? How good does that need to be? ±1"? 0.0001" How quickly is the measurement needed?

 

[controlsdude]

I guess a big general question is the room that this equipment is in humity and temperature constant? I remember in my days of paper converting that the rooms had air washers (fancy ac) that kept the room humity and temperature constant. Otherwise, not sure you will even pull this off with spec on equipment?

 

[jraef]

This is more of a sensor issue than a PLC issue in my opinion. The flatness sensing is already available as a canned product, I have used them for similar applications (roofing tiles). It can provide an analog output into a PLC, or is can be configured with an internal bandwidth and give you a simple "Go-No Go" or "Pass-Fail" discrete signal that you can look at with a PLC. The sensor is usually called an OD Sensor, for Optical Displacement. Sick is who I'm familiar with, but there are plenty of competitive products out there. Then the printing aspect is a simple canned solution as well, there are a bevy of ink-jet case/pallet/crate printer companies out there already, tried and true technology, easy to integrate. HP even has canned ink-jet printers for industrial applications now.

Moisture on the other hand is going to be more of a challenge. Moisture typically requires sampling of some sort or another. If you had a drying room and want to monitor the moisture content of the air, that's relatively simple. But detecting a moisture level in a product moving by a sensor is something that's very specialized. The big industry leader in that, for applications similar to yours (pulp, paper and converting) was Measurex, who was bought by Honeywell. After the Honeywell purchase I lost track of them and their products, but I'm sure Honeywell did not abandon that technology, it was the industry leader. The only other one I have ever seen in use that was non-contact was this one from Moisture Register Products, a Near Infra Red (NIR) sensor that looks at and compares wavelengths of reflected NIR energy. I'm sure there are more out there though.


"Will work for salami"

 

[Skogsgurra]

I think you guys are working yourselves up a bit.

The setting is quite well defined: An industrial environment with no extremes (paper converting). Available power is standard. And if a printer is needed to label the stacks (or sheets) - let there be a printer. You cannot always use a jet head to print directly onto the product. But if possible, why not?

Sensors for humidity are plenty. And so are distance sensing lasers or triangulating devices.

Jeff mentions a few papermaking household names, which I am sure you recognize, but there are a lot more. Keyence have been quite innovative lately.

As for the PLC, I think that something just above a micro PLC (like the LOGO!) is needed. I said just above. Main thing is that it can communicate with the sensors and that it can control the output device (printer or similar).

Tell your manager that this will consume time and that you will need to go an introductory training course to be able to work efficiently. Or that you need to take in a consultant.

Also tell him that the long-term best solution is that you do the job. You will then not be dependent on outside resources when/if problems occur.

Final word: Don't just get it running - write down what the thing does. And how. Name everything. Documentation is important.



Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ander88]

Hi Everyone,

Thank you for all the reply's.

Here are the sensors we have provisionally chosen,

Distance:

http://www.ifm.com/products/uk/ds/O1D155.htm

Moisture:
Not chosen yet.

We already have a couple of the distance sensors in the factory on different applications.
The moisture sensor is still being researched, we have a couple of options but all we have looked at give a 0-20mA output.

The measurements will be taken on a stationary pallet.
I will suggest the jet head printer to my manager (we have a few of these in the factory) but I think he would rather have a label printed to stick to the pallet.



There are no harsh environments. The area to take measurements will be in the quality control area which has temperature and humidity control so will be constant.

As a factory any kind of power is available, but a single phase supply will be used with a 24V transformer.


Time is not an issue, I'm open to learning.
I can already program in C+ , HTML, PHP, some Java so I presume I will be able to easily pick up some programming from an introductory course.

I've had a look at the Logo PLC and it looks good. If my memory serves me correctly these are the same PLC's that are in our heating control system.

Will this be able to communicate with a touch screen interface?

Thanks again,

Michael

 

[Skogsgurra]

I repeat "I think that something just above a micro PLC (like the LOGO!) is needed. I said just above"

The LOGO! is too rudimentary. It is just a "programmable relay" - you need a real PLC. An entry level one.

You will have to do the rest yourself. Like finding out how to communicate with a printer and so on.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[controlsdude]

If I was in your shoes, I would contract out the hardware design and build of the panel. Probably have plant specs to go by and usually a controls house is better able to deal with the cad work and panel builds then your plant. Do the software yourself since that is your area of expertise as you say. Its been my experience that software people dont have a clue about hardware sometimes, but maybe your different. Better to focus on what you know then bite off more than you chew.