Off-the-Shelf Non-PLC Interface

[grover1]

Greetings.

I have a packaging machine that has only a couple of spare PLC DC inputs and outputs. I need to implement a new function that will individually monitor sealing cycle air pressure delivery on six pancake cylinders. I will be using six SMC ISE40 pressure switches to deliver PNP output signals that will turn off when the pressure exceeds the set point. My problem is having only two DC inputs into the PLC (A-B 5/30 w/DH) when I need six. An expansion chassis is not practical and adding anything with software, such as a small PLC, would entail purchase of more software and the corresponding software controls (i.e., documentation, validation, etc.) required in this industry. I was hoping for an electronic solution, perhaps something that could logical AND multiple inputs and generate one output when gated. It would of course be great to have it contained in a nice little enclosure with DIN rail mounting!

Ideas anyone?

 

[Bitmore]

Sounds like the problem could be solved with some external relays, you can perform the logical operations of AND, OR, and NOT with a few relays.

bitmore

 

[bernie_carlton]

Do the sensors provide enough current to energize a small relay? Check out AB 700-HL Interposing relays (.2 watts for the coil). Then series relay outputs for an AND circuit.

 

[The_Wanderer]

Do you need to know which of the 6 cylinders is out of range, or only that one of them is out? Sounds like the latter, in which case the others are right-- a few relays and you're sorted. Otherwise, there may be trouble...

Just FYI, I've had problems with those ISE 40's being a bit unreliable and not very accurate. We still use them to monitor incoming air pressure, but there is a lage accpetable range there. If you need accuracy, you may need to look elsewhere.

 

[Tom Jenkins]

If you haven't already bought the switches consider using an old fashioned switch with "mechanical" SPDT contacts. Just wire through all six NC contacts in series back to the PLC. If any switch fails or if the wire is broken then you will lose the "Pressure OK" input to the PLC and you can stop the process.

If you wanted to you could wire the NO contact of each switch to its own resettable counter, and then you would even know which switch tripped. Or, some switches (I think Murphy switch and Dwyer for example) have a telltale or follower needle that shows the highest point reached by the switch. Not ideal, but if you don't want to spend any money you have to accept compromises.

Now, what are you going to do with the spare input?

 

[Bitmore]

Now if your still interested in something "smart" look at this...

http://www.industrialnewsroom.com/fullstory/450970

bitmore

 

[akreel]

Do the sensors provide enough current to energize a small relay? Check out AB 700-HL Interposing relays (.2 watts for the coil). Then series relay outputs for an AND circuit.

And if the sensor output current is an issue, solid state relays with a good-old electromechanical tied to their outputs could work.

 

[Eugen]

I think with 2 inputs and nothing smart you can discriminate between 4 inputs (or 4 different scenarios) using binary logic (maybe 2 relays) unless .... there somithing I dont see in the other postings. Or maybe u can free onother one (input)??

 

[Bitmore]

Well grover1 looks like you got some explaining to do! But chances are the past few post have sent you on the way to a solution, how's the project going?

bitmore

 

[grover1]

Thanks for the great response, everyone!

Rats. I sure was hoping some Bullwinkle out there was going to pull a rabbit out of their hat and produce just what I was looking for (a DIN module six in and one out). That doesn't seem to be the case, so I guess I'll proceed with the next best thing, our electro-mechanical buddies, the relays.

Six DPDT relays relays should do the trick. I can connect one leg of each serially to the PLC and the other leg individually to a status LED. The LED won't latch anything, but most of our failures aren't expected to be intermittent.

I'm very sorry to hear that the ISE40s aren't reliable. I selected it because it has an analog output as well as two digital outputs. We verified that they were within their published specification, however, I may have to increase the calibration cycle to see if we experience any reliability issues.

As for the IDEC, that would be too much like what I'm trying to stay away from (on this project at least).

Thanks again, and keep new ideas coming!

 

[akreel]

Rats. I sure was hoping some Bullwinkle out there was going to pull a rabbit out of their hat and produce just what I was looking for (a DIN module six in and one out).

Well, I kept it to myself because relays *should* be a slightly less expensive and easier option, but if you felt like learning new software...

Check out the Crouzet Millenium II. It's essentially a DIN-Rail mounted PLD (similar to the IDEC). You can program it to do what you need. You'll pay for the hardware, but the software is free (I think).

AK

 

[Doug_P]

FLEX I/O?

 

[Grover]

The machines I work on are used for pharmaceutical packaging. This particular machine will remain in use for only 12-18 months. When any type of programmable or configurable hardware is involved, the project becomes significantly more complex and time consuming because of the validation requirements imposed by the FDA and by the company. These processes are very important to ensure the drugs meet safety, quality and efficacy requirements. Software design documents, process verifications, standard operating procedures for maintenance technicians are just some of the documents that would need to be generated for this very simple hardware change. (Ever wonder why drugs cost so much?) Before you think this is overkill, people can die or be seriously injured from products that weren't manufactured properly. Just the simplest problem, like a number 8 that looks like a 3 from a clogged lot code printer head that the machine vision system missed can result in an entire lot of packages (500,000 perhaps) being reinspected (by hand!) or destroyed. How about a paint chip, an o-ring seal fiber or any one of a myriad of foreign substances found in just one bottle of pills or tube of gel? Toss the lot. Getting it right the first time is critical. I'm sure many of you can relate, working in food, beverage, chemical or medical industries.

This is why I'm staying away from programmable logic on this one. It would be great to use the Crouzet or IDEC, but since it can be programmed in more than one way, the project becomes too complex for the return on investment. I had hoped for some discreet logic module, but relays will be more than adequate, albeit antiquated and mechanical. I still have to validate the changes and update equipment drawings and schematics, but I get to avoid an entire layer of design and testing by leaving out programmable hardware.

Thanks again, everyone.

Grover/Grover1

 

[Tom Jenkins]

It has just occurred to me that a safety relay, of the type used for light curtains and such, might be what you need.

 

[stevez]

I'm with you Grover

Grover,
I work for a large pharma company.I really sympathise with you.
I don't think guys from other industries know what we go through to make even the smallest change on a machine or in a process.

Good luck ,

Steve