Load cells to PLC solution

Does anyone know a load cells to PLC solution?

We are using “Survivor” and it malfunctions very often. I am thinking to propose a PLC instead of. I have learnt a little bit about Cells. It doesn’t look straightforward. It needs a signal conditioner and summing junction box in my case (we use 3 cells).

I’ve already found AEC blender solution with A-B M1500 controller, but it still is not simple. It uses digital input to count pulses I guess. I couldn’t upload that program.

So, does anyone have bad or good experience with design and using that? It would be too bad if I develop worse solution than it is.

Thank you.

Alex,

I have used this company in the past with a AB 5/03 along with their waver saver card.

http://www.hardyinst.com/

Many of the major scale companies offer multiple output options, such as devicenet, ethernet, profibus etc....

A number of PLCs have analog input cards that will take a strain guage signal directly. Most strain guage suppliers have signal conditioners that provide a 4-20 mA or similar analog output from the strain guage. These analog outputs can connect to virtually any brand of PLC with analog inputs. Most signal conditioner manufacturers have units that will provide excitation for the strain guage and convert the mV output to a 4-20 mA signal. Summing the three signals is easy in almost any PLC.

You have lots of choices. Each has advantages and disadvantages.

I use Magpowr (magnetic power systems) load cells and their din rail mount TSA amplifier connected to PLC analog inputs. They're simple to set up and versatile. They can each sum two load cells. All (except one) of my applications use a separate amp for each load cell, and any summing and error checking is done in the PLC. I have a Nexen stand alone tension control system that uses dual load cells per sensing roller. On this system, if one load cell fails, my readings get cut in half and the tension is doubled, but the PLC is unable to detect it ... just something to think about.

http://www.magpower.com/pdfdocs/Instruction_Manual_-_TSA_Tension_Sensor_Amplifier.pdf

Could you describe your application in a little more detail? You mentioned a blender, are you using the load cells to measure the weight of a blending vessel? a brief description of the process could be useful.

I am close to a distributor for the Hardy Weigh Scale solution here in New Zealand, so I have no possible commercial interest in this...but I have seen numerous customer applications for it and it works exceedingly well. The more demanding the application in terms of speed, accuracy and ease of setup, commissioning etc, the more worthwhile the modest extra cost is.

Having the weigh solution appear right on the backplane eliminates all the speed and integration issues, plus the Hardy instrument seems to deal to vibration ( a very common cause of weighing issues) rather nicely. I always recall one of the first systems I saw installed...a 15 tonne flour hopper with a small vibrator near the throat. With the vibrator running (and the throat closed) the total hopper weight would bounce around for a moment or two and then settled down. Then after that we put a 20cent piece on part of the framework...and the last scale digits accurately weighed it. Pretty impressive!

It is a batching system in plastic extrusion to load raisin.

Recipe of five ingredients.

Total weight of batch is 50 kg.

To control I have now:

- ML1000 – 5 hoppers loader;

- ML1000 – extruder loader;

- and “Survivor” panel – scale.

Materials dropping one by one and then a batch is “At weight” Survivor wait until extruder loader mixed that stuff and emptied bin. Often Survivor doesn’t provide “At weight” signal and augers are feeding over and over again. Material overflowed, augers overloaded. Also we have problems with hoppers vacuum pump overloading.

In my thinking a one PLC+HMI to control all 3 parts would be better idea:

• I could control scaling process;
• I could assign hoper loading priority according to the recipe;
• I could keep truck of operation to analyze causes of tripps and provide technological report.

My bosses are Ok with old system but I am looking forward to use PLC as much as possible. I need strong reasons to overcome conservatism. What do you think: Is my reasons strong enough?

Thank you.

As a general rule, every time I have moved from external devices to PLC control, it has resulted in major improvement overall.

Yes, the more data the PLC logic has about the machine it is controlling, the better it can perform and self-diagnose.

I see it as if I'm removing a blindfold from the Minds Eye of my plc programs when I move relay logic or external controllers into a main PLC.

If these machines are to be kept in the same arrangement for the near future, there is no reason to keep separate ML1000s. Or, you could make them communicate, for now. If these machines are to be separated in the future, they would each still stand alone.

Often Survivor doesn’t provide “At weight” signal and augers are feeding over and over again

Click to expand...

You should be able to halt that problem with a timeout timer added to the ML1000 right away.

As for the upgrade, (Disclaimer: All of my load cell experience is measuring tension of a moving web.) I would not want to have to depend on communication with a separate scale, and it seems entirely feasible and tempting to roll your own. It would not be a huge waste of money to try this, and once you got it working, it would be maintainable with common off-the-shelf parts.

PhillipW is not the first person I have heard rave about the Hardy weigh scale. They must be good. More money, guaranteed to work.

Definitely go to a processor with analog capability, a little more memory and better communication if you're going to centralize the control.

I have used Hardy Instruments and Mettler Toledo products and both have great products. It may take some time to get up to speed if you haven't worked with scale/load cell products before but work great with few problems long term.

Making a call regarding PLCs is easy based on the application. If OEMs are doing it they are probably using PLCs. There are tons of companies that do batching / weighing etc... systems with load cells and PLC/HMI systems out there.

The PLC is the right solution for this application.

One way you may peak the interest of your bosses is to do a cost payback. Is there a cost to not getting a consistent mix? Does a batch ever need to be disposed of? Are you adding too much of an expensive ingredient? How does this tie to the QA/QC requirements of the products?

You may find it difficult to assign a value to these questions but it is how management thinks when making decisions.

If "Survivor" is the one I am familiar with then its a Rice Lake weighing system and should be "almost" as good if not as good as most products.
NOTE: Survivor I am familiar with is a truck scale.

I do not know the process you have but it is common to use "weight controllers" with plc's....primarily a dedicated processor for weighing. The primary issue with any weight controller is calibration and the ability to deal with vibration etc...ie are there external conditions that could cause improper scale readings.

It appears that in this application plc's are being used for "control" but your main issue is getting the right signal from the "weight controller"...as OkiePC mentioned you could determine time period the weight should have been added and use a timer to halt the process if signal is not reached in the correct time period....BUT this is a band-aid.

Personally, assuming I have interpreted the application, I would determine if the "weight controller", Survivor, is setup, calibrated and functioning properly....you may need to re-calibrate and insert an averaging routine if the controller has that cability. Many controllers can use 10 (or more) weight examples updated in millliseconds to provide a more accurate weight and eliminate interference from things like vibration.

To convert the process you may have to change load cells or purchase an amplifier (or signal conditioner) to convert the signal to an analog input the plc can use.

This is just my opinion but I would try to make the system work "accurately" as it is and if that is not possible then look at alternative methods....another controller may be less expensive to use and offer more features. I would obtain as much data as I could on bad batches, frequency controller does not accurately signal etc.

Using the above method and information you should be able to determine what would be needed to improve production rate, what the cost of the new equipment would be and at the improved production rate how long it takes for the new equipment to pay for itself...the shorter the better.

In other words do not just look at "incorporating a plc", look at what is best for the system and the company.

BTW: I am also fond of Hardy Instruments and Mettler-Toledo products. For smaller applications I also like Doran Scales...no relation.

I always recommend using a Standard brand name external weigh controller using a proven connection method such as RIO or D-NET for two reasons:

1 - Field calibration, unless you are an integrator who happens to lug weight cart full of 50lb test weights around, chances are a scale company will be looking after the testing and calibration of the scales. If the scale company shows up and there is Toledo , or Hardy or Rice Lake weigh controller, that they can calibrate. A PLC backplane solution requires access to the PLC.

2 - Weight conversion. No resolution is lost via a digital transmission. The weight on the scale indicator display is seen in a PLC register. How many times I have heard when an analog output is transmitted to an analog input , " How come the reading on my display does not match the reading on your screen ?"

Ian Phillips
flomation engineering

1 - Field calibration, unless you are an integrator who happens to lug weight cart full of 50lb test weights around, chances are a scale company will be looking after the testing and calibration of the scales. If the scale company shows up and there is Toledo , or Hardy or Rice Lake weigh controller, that they can calibrate. A PLC backplane solution requires access to the PLC.

Click to expand...

I don't like to depend on outside sources for this. I keep calibration weights, so any technician can verify calibration and calibrate at any time. It's easy to program an HMI to allow this so even some operators are capable of handling it. I do agree that for precision, a dedicated controller interfaced to the PLC would be superior to a roll yer own approach. Next time I have to do a new one, I'll definitely investigate one of the backplane type modules.

It appears that in this application plc's are being used for "control" but your main issue is getting the right signal from the "weight controller"...as OkiePC mentioned you could determine time period the weight should have been added and use a timer to halt the process if signal is not reached in the correct time period....BUT this is a band-aid.

Click to expand...

A well placed bandage may stop the bleeding and earn you some points with management ... prove your skills. Don't stop all the bleeding, though, or they won't need to upgrade!

One of the nice things I like about the Hardy Inst. set-up I use is there is no need to calibrate with an external weight, it has a chip built in it that automatically calibrates the unit when directed to do so and then you just check it with a know weight from the QA office.