Created Date: May 21,2009

Created by: marksji

It could be a lot of things. First ideas that pop into my head are the photo-eye (I know you say it works, but my experience leads me to question it anyway), a software problem (something else inhibiting the output?), and the scan time of the CPU.

Depending on how long it takes the PLC to complete a scan you could get into a situation where things are simply happening too fast; however, I wouldn't put this at the top of the list to check as .5s and .7s is an eternity in the PLC world.

Created Date: May 21,2009

Created by: Wineslob

It could be a lot of things. First ideas that pop into my head are the photo-eye (I know you say it works, but my experience leads me to question it anyway), a software problem (something else inhibiting the output?), and the scan time of the CPU.

Depending on how long it takes the PLC to complete a scan you could get into a situation where things are simply happening too fast; however, I wouldn't put this at the top of the list to check as .5s and .7s is an eternity in the PLC world.

Ty, I'll see if I can check the signal from the eye, though it was replaced roughly 18 month ago.

Created Date: May 21,2009

Created by: milldrone

We have been dealing with a problem with our main casepacker for quite awhile (several years/we just "put up " with it till now), but now it seems to be getting worse.

The casepacker is run by a 205 with a DL230 cpu, 2- D2 16NA inputs and 2- D2 12TR outputs.

When I first read this I was suspicious of one of the relay outputs going flaky. But you have trouble shot the unit, and confirmed that the output indicator sometimes misses a beat or part of a beat. And since the indicators are driven by logic it cannot be a faulty output contact.

The problem is the glue applicator airvalve is controlled by the PLC to apply 2 squirts of glue to the flaps as the box is moved to the end of the machine and into the "press ". It's a .5 sec ON .7sec OFF .5 sec ON cycle. This is so the glue skips the center of the box, and wont get glue on the exposed cans. MOST times it applies the 2 squirts, but about every 20th box or so it misses the rear flap.

Please forgive this next statement if it's not possible. Don't forget we are reading this description on the internet, and we are not there to "see, hear, or smell " the problem (you might be surprised to find out how many problems can be troubleshot "hands on " and not remotely) . Is there a possibility the box is moving faster when it "misses " the second flap? you indicated that the glue is dispensed by time and you indicate the presence of a photo eye signal.

Some details on the photo eye to product might also be helpful.

Is the photo eye some distance before the glue applicator? Or right inline with the glue applicator?

Is your photo eye sensitive to color? Can you spray paint one of the boxes flat black and see how it reacts?

What does the applicator do if the box is half or two thirds the size of your regular box as far as the photo eye is concerned?

BINGO, I could see the light come on for it's two outputs to the airvalve, and sure enough, sometimes it only gave one signal, and, believe it or not, a half signal (resulting in a half squirt of glue), but every time only on the second "squirt " IE: the rear flap of the box.

Can you see both the Photo eye input indicator (on the PLC) and the output at the same time? Does the photo eye PLC input indicator blink when it should be on solid?

Please forgive this next question if it's way off base. Does the product move or jiggle the photo eye wires in any way as it moves past? Does the machine shake or vibrate?

Can you post the code from the PLC?

Created Date: May 22,2009

Created by: milldrone

The AB photo eye that tells the PLC that a box is in place to be glued is set to a "overload " so we get as strong a signal as possible, so I don't think it's that causing a problem.

Now I'm not sure what to look for. Any ideas???

Last night on the drive home from work I thought about this statement. At this point we have no other details on this photo eye other than it was replaced 18 months ago.

Then it dawned on me why this was bothering me. In two previous trouble shooting problems I encountered a "excess gain " issue.

1. The client had a very odd looking opposed pair through beam. Separate emitter, separate receiver. This application was for a case packer also. The problem was the through beam was not always changing state when a box was placed in between the emitter and the receiver. Sometimes it would but on occasion it would not. This opposed pair had very high excess gain and the client had it "maxed out " to try and solve the problem. The reality was the photo eye pair was actually "burning " through the box but not through the product (corn flakes). The client had also just installed a new weighing system that was more accurate, and was now filling to "perfect " weight instead of overfilling. Thus the level of cornflakes was right at the beam height. This why some boxes "tripped " the photo eye and some did not.

2. A conveyor belt had product moving on top with gaps in between. This photo eye and an encoder measured the product length. The photo eye was a non polarized retro reflective unit. Again a high power excess gain unit. The product was about 14 inches from the eye and the reflector was 10 feet further but in a dusty location. The reflector would become dusty and instead of wiping it off, the operators turned up the gain on the photo eye. This worked with some success. The problem was that on occasion a long piece of product would trigger the eye for several short durations. This completely messed up the sorting system. What I discovered was the offending piece of product was squared to the eye and was also somewhat reflective. The operators would not clean the reflector so I placed the photo eye and the reflector at an angle to the product. Problem solved.

Created Date: May 25,2009

Created by: rlcrewse

You definitely want to reduce the gain on that PE. I have had problems especially with diffuse type PE that not only created false signals (a light in the background, someone passing to close to the PE, or even someone walking between an open outside door and the PE during the summer months between 1 and 2 PM!). Another problem with diffuse PE's is that they can be mounted to close to the object you are trying to detect and won't even see it.

But, since you are getting the first flap glued the PE is initially working. If you have the capability to view the program you need to find out what resets the glue cycle and timers back to square one - waiting for a box. It sounds like the timing for the second flap is being reset before the glue timer is finished. If the PE sensing the tail end of the box is resetting the timers see if the PE stays on for the full length of the box. If the 2nd timer reaching the preset is resetting the cycle check to ensure that the box isn't being slowed down, hanging or dragging on something, slipping on the conveyor, or see if the conveyor might be slipping.

Also double check for loose wire connections whether they are at terminals or wire nutted together in a couduit body somewhere.

Good luck

Created Date: May 26,2009

Created by: Wineslob

You definitely want to reduce the gain on that PE. I have had problems especially with diffuse type PE that not only created false signals (a light in the background, someone passing to close to the PE, or even someone walking between an open outside door and the PE during the summer months between 1 and 2 PM!). Another problem with diffuse PE's is that they can be mounted to close to the object you are trying to detect and won't even see it.

But, since you are getting the first flap glued the PE is initially working. If you have the capability to view the program you need to find out what resets the glue cycle and timers back to square one - waiting for a box. It sounds like the timing for the second flap is being reset before the glue timer is finished. If the PE sensing the tail end of the box is resetting the timers see if the PE stays on for the full length of the box. If the 2nd timer reaching the preset is resetting the cycle check to ensure that the box isn't being slowed down, hanging or dragging on something, slipping on the conveyor, or see if the conveyor might be slipping.

Also double check for loose wire connections whether they are at terminals or wire nutted together in a couduit body somewhere.

Good luck

Thanks, I'll do some checking.

I don't know if there is a "reset " in the program. What I do know is there are "go/no go " requirements at the stage where the glue is applied. From what I understand we must have: Drive chain engaged, and box in place. (the PE detecting the box) so I'm now getting suspicious of the eye. I'll find the input and watch that indicator.

As to being able to "burn through ", I doubt it (we are talking about a box full of 1 gal paint cans) but that does give me another avenue to explore. Also the speed is fixed, and I did look at that.

As was noted before it looks like signal loss, for whatever reason.

Thanks for all the ideas, I'll report back what I find.

Mike

Created Date: May 27,2009

Created by: Bob S BN

Troubleshooting method

Because there are many things to look for when troubleshooting intermittent problems, we sometimes set up a video camera focused on important info such as the PLC rack in an effort to catch things in an area you may not have been looking right when it faulted. You can then go back and study the tape over and over without having to be lucky enough to be looking in just the right spot when it faults.

Just a thought,

Created Date: May 27,2009

Created by: milldrone

Both times the signal from the eye (case in place) was always on. It diden't flicker, or go off prematurely.

Not sure what to look for now?

Do you have DS5? If you do, can you post your program here? There are a couple of ways to do it. Zip all the files prj., lcd., scd., scx. ect and attach to your reply. You can also File, export, program, save. This saves the monics to a txt. file you can then either copy and paste to your reply or you can attach the file to your reply. Or you can save as a PDF. file and attach.

Perhaps there is something someone can spot a problem area that might cause a malfunction. Another possibility is that someone can write some "traps " that can help to identify where the problem is coming from.