OT : Thermocouples

Ok

So this may sound wild but we have a bunch of idiots here who can never seem to plug a theremocouple into the correct jack.

On each line there is a extrusion dies that has about 12-20 zones and a TC for each zone. When we replace a TC we put a brass tag on it with the corresponding zone number on the tag.

Somehow they still plug them into the wrong jack and the jacks are numbered by zone also BTW.

When this happens they are not getting the correct reading for that zone which in most cases cause the zone that is being controlled to overheat and wharped dies and all kinds of stuff.

On the lines that have plc controlled zones i limit the setpoint and i limit the output. I also look at the feedback reading and if it is out of range i stut that zone down this saves the equipment but maintenance is still called out to finf the problem and most of the time it is just that they have it plugged into the wrong place.

So now to my question does anyone know of a company that makes some kind of keyed thermocouple plugs and connectors so that the idiots can not physically plug the TC into the wrong port?

If it will not go in on the first couple of tries then maybe they will look at the tag?

Any other suggestions or methods are welcome. This is something the plant manager and Maintenance manager wanted me to research as it is about 30% of our reative work orders dealing with heater related problems. Of that 30% i would say half is related to incorrect TC Connections.

Thanks

Da Kid

If you make it idiot proof, they just find a bigger idiot.

However, I would suggest a couple of options. I believe Omega Engineering http://www.omega.com/ has thermocouple selector switches.

A better solution is to have a solution that leaves the thermocople permanenently wired in. If you boss objects to the cost, find out how much the downtime and warped dies are costing.

If you are just getting local indication on some of your T/Cs then buy a bunch of digital indicators from Red Lion etc.

If you have some tied into controllers buy a bunch of 1/4 DIN controllers from Red Lion or others.

The cheapest solution overall is probably to get some inexpensive PLCs and HMIs from AutomationDirect or your preferred supplier, wire all the T/Cs to inputs, and display on the HMI. That will also allow you to control, alarm, tie to SCADA or MIS, and provide a path for growth and refinement.

It's like the old joke
Man: "Doc, it hurts when I do this."
Doctor: "Then don't do that!"

If the plugs are the problem, get rid of the plugs!

The use of conducting metal other than the thermcouple element wire does create a junction but if the junction is isothermal then one leg is positive, one leg is negative, each cancels the other.

So, theoretically, you do not need thermocouple connectors.

The key requirement for a thermocouple connector is to be isothermal, that is, same temperature gradient at both terminals, so that its in-line junction does not add or subtract to the temperature being measured at the hot end junction.

So, you can use other copper or nickel plate copper connectors for thermocouple junctions.

Between mini and regular sizes for different type T/Cs and RTD quick connects you've got about 10 different plugs/jacks of different colors.

But if your shop people ignore tags, they're just as likely to ignore color coding, if the plug mates with the jack.

So, use an M12, a canon audio plug/jack, go catalog shopping.

A pharma has this same issue for its CIP thermocouples.

They did what Tom suggests - permanantly wire the T/C's and their HMI picks a subroutine program that selects which AI's are used for that given setup.

A PLC would definately help. Once a die is installed and all connections made the PLC could heat one zone at a time. It could watch the temperature feedbacks and raise errors if no zone showed a rise or the incorrect zone showed a rise. Then repeat with the other zones. It could take a while but it would be a type of self checking.

A even more intelligent program, assuming at least one feedback reacted for each zone heating cycle, would remap the inputs for the affeted zone thus un-idiocizing the machine.

You might also look at Harting QD plugs. They make many type and sizes and most can also accept coding plugs to prevent mis-match. www.harting.com

I think i did not make my post as clear as it could be.

These dies have power and TC plugs on them because they are moved to other lines and changed out a lot.

I am looking for something with some sort of keying pin for each zone where it will not plug into the wrong connector kinda like using keyed harting connectors for controls.

I was under the impression from what i have read a t omega that you can not put in a inferrior metal and not effect the temp reading from the TC? This is why they have TC terminal blocks and such from my understanding?

From the jack back to the plc card is 50-70 feet on most lines. From the jack to the die it is 10-18 feet.

Hardwiring is not a option as TC stays with the die so it has to be unglugged and they move 2-3 times a week on average.

All zones control and go to the plc TC input cards.

Are there any keyed M12 connectors? I would need up to 20 different ones and we would have to put male ends on them as we currently buy stock TC with male connectors already on to save time.

Harting has type J pins available for the inserts.

Also, if you are plugging power and TC then you can use one plug system pre Die and place coding pin in the plug/receptacle to make it "idiot" proof.

Are these T/Cs that the operators are unplugging/replugging as they set up the line? If so then don't you have a set up inspection procedure? Sometimes we look for a technical solution when a policy solution might work just as well. A set up should be inspected and signed off. If its not right but it was signed off then the signing employee has to answer for it. If the line is run without the set up signed off then the set up employee has to answer for it.

The Plc Kid said:

I was under the impression from what i have read a t omega that you can not put in a inferrior metal and not effect the temp reading from the TC?

Click to expand...

Intermediate materials are allowed and they don't affect the reading provided that the junction temperatures at both ends of the intermediate material are the same. Omega actually covers this topic in their technical literature.

This is why they have TC terminal blocks and such from my understanding?

Click to expand...

Not quite. Its so they can make money selling you something... and also to give you a solution for when the junctions are a different temperature.

Kid

I think the solution to your problem is as suggested ie two man rule.
Extruder dies and heaters do have to interchange onto different machines. Saw a case where a guy was able to plug a worn plug into the wrong power source and zap himself.

One guy connect em and other does independent check.

Standard policy and practice on submarines, nuclear weapons etc.

Some people ie you and me are perfect and we NEVER wired a motor wrong or made other silly and stupid mistakes.

Dan Bentler

When I worked in the platci industry we had similar issues with our molds. What we did was use Tomas and Betts hoods and plugs Here is the web page: http://www.tnb.com/contractor/docs/posekon.pdf

These plugs can be pinned so you can use the same and it will not allow for anyone to cross the heaters and TC's. We basically made all the molds "plug in and play". Tokk about a year to get it done but it made life alot easier once it was done. We would put the base on the mold and hard wire it and put the hood on the machine.

Note: the plugs are techinically not TC rated but it never became an issue. Pins are copper with a zinc coating so cunductivity is good.

One other thing we did was create a test jig in the tooling shop so the mold could be tested prior to going to the floor. This was needed since the tooling shop wired the new haters and plugs to the molds and we wanted them correct before the mold went to the floor.

Our last step was to color code the plugs for the really dim witted ones. Some of our molds would have 12 to 14 heating zones. Since the guys on the floor could not count we made sure they weren't color blind also.

I wish I had some pictures. I will say this elmitated 90% of the silly issues.

One other thing. You will see in the catalog that there are many different styles. Make sure the TC plug is nothing like the power plugs. Pins can be removed and swapped around and the last thing you want is to cross power with TC.

There are already policies in place to check things as they have been done. This is done by the line tech but most of the time done half a%&ed.

Getting these people to follow a policy or procedure correctly is like pulling teeth.

Getting management to make them do it is like pulling lion's teeth.

Is it possible to clamp or dress the cables on the machine so that they are just long enough to reach each T/C?

Then they must use the longest cable for the T/C that is furthest away and so on down to shorter and shorter cables. If they put one in the wrong spot then there will be one left dangling that doesn't reach the empty T/C.

Of course, this presumes that all the T/C's are different distances away from a hard mounting point. Also, with them "just long enough" there may be more of a tendency to fall apart. And, there is the possibility that they just yank on the cable until it breaks or comes free of the clamp

Maybe you should post a picture of the die.

Alaric said:

Intermediate materials are allowed and they don't affect the reading provided that the junction temperatures at both ends of the intermediate material are the same. Omega actually covers this topic in their technical literature.

Not quite. Its so they can make money selling you something... and also to give you a solution for when the junctions are a different temperature.

Click to expand...

I concur. Alaric is absolutely correct.

Thermocouple wire is NOT needed at junctions. Thermocouple switches do not use thermocouple wire, they use plated copper.

It isn't clear to me how the Tomas & Betts connectors polarize or key one to another. It seems one can put select certain pins in a multipin connector, which might create an open circuit if #8 were mated with #5, for instance. But it looks to me like #8 will mate with #5, unless there's some key that I missed that prevents such.