Ideas please

Hi I'm faced with this problem ,our can seamer keeps jamming up but dose not stop as it just rips the cans to bits till some one notices.

Now My theory is I'll count cans in and cans out and look for a difference now this is ok till they get a jam up and clear out the damaged cans how do I reset the count as I won't know how many cans are left in the seamer..

I'm Now thinking maybe a bit shift..

Any Ideas would be appreciated

The PLC is a SLC 5/03...

Try search for "part tracking" as there has been a number of threads allready.

But maybe the solution is relatively simple.
When the seamer jams, does that mean that no cans come out of the machine at all ?
If so, you could program a simple timeout alarm. If cans enter the machine, a timer is started. Every time a can leaves the machine, the timer must be reset. When the time expires, generate an alarm and stop the machine.

If on the other hand, the cans can pass thru the machine somehow, even when some of the cans are jammed, then you must make a more intelligent alarm.
I do something like this to counter a 5 minutes dead time:
I have a FIFO register with 10 memory positions that sums up parts within 30 seconds.
You can use the output of a similar FIFO register to generate what should come out of the machine. You have to compensate for the difference between the FIFO output and the real count of cans on the output (there will be a difference even when everything is OK). I would do this by converting the FIFO output to a filtered "cans per minute - delayed input", and compare this to the filtered signal "cans per minute - real ouput". Activate the alarm if the deviation gets too high.

edit: And to the question you were really asking:
You have to prompt the operator on the HMI:
"please remove all cans from the seamer". And when he starts the seamer again, clear the FIFO buffer.

Thanks Jasper The timer Idea might work as we are only looking for the really big jams where we end up with 30 + cans in there .....

I've been thinking too deep (I must remember K.I.S.S.)

P.S There is no HMI (production manager runs away when I aproach as he always ends up spending money if I corner him.)

Prox addition?

Maybe you can rig up a proximity sensor, some where near the line where "can bits / shrapnel" should NOT be, and if the Prox trips an audible alert bleeps.

Another idea would be to put a current monitor on your motor line and monitor for an "out of nominal current range" situation. I've never done this but I'm sure there is some comparative sensor out there that could do the job.

Also talk to the Field Service Guru who is employeed by the Equipment Manufacturer and works on this equipment all the time. Perhaps a solution has already been fabricated - and you just don't know about it.

We have tried using a current monitor and there is no change between seaming the can and ripping it to bits ,well not untill the seamer is full i.e 30 + cans and by then its getting realy messy...

Jezz, in an application I set up on a high speed filler/rejector process, we counted bottles in and out to the same counter register. When the counter value went too high, indicating more bottles in than out, we shut the filler down. Every time the start button on the filler was pushed it cleared (reset) the counter. To avoid accumulation errors over a long period, we had a timer running that reset the counter every five minutes. This was so that one rejected bottle every so often wouldn't accumulate to the point of being out of range and shutting the filler down unnecessarily.

I'll go with Rube's idea.
But rather than clearing the count value completely, just subtract ONE can from the count value every 5 minutes (or something that fits the approximate number of missed cans per timeunit).
Then, calculate the rate-of-change in the count value. If the rate-of-change rises steeply, then there is a jam.
In that way you will also cover the moment right after the 5 minute period.

Jesper may have a good additional point, depending on the distance between the in and out sensor (our sensors were 26" apart) and line speed in PPM (in our case 650 Bottles per minute). We looked at a difference of 10, in other words 10 more went in than came out. The speed of the bottles made it highly unlikely to worry about the period right after resetting the timer. A slower line or greater distance between sensors might make the subtraction aspect beneficial.

My initial thought was in-line with Hesters' thinking. That is, configure a new design to detect the occurance. However, is that not similar to a doctor treating the symptom rather than the cause? It's quite evident that the equipment has a serious repetetive malfunction. Look a little deeper and get to the root cause of the jam. Try not to settle for acceptance. Too many people will say that it has always done that and will always do that. BS! Fix the problem.

i would use some detectors as close in the machine as possible.
one on station number 1 for detecting going in.
1 for detecting going out again.
one for station detector
have it with bit shift and stop if any difference.
on the input have a plate on top of cans to see if a can is too high.
have a plate on the side just from top to bottom seam and see if a can is misformed.
slow down machine if input is low so you will have only full stations.
on the empty station check for can or rest of can with a plate.

We use a Spring loaded gate at discharge of seamer which will trip a prox and cans fall out of gate into trash bin. Our seamer are all quite old American can and canco seamer and run about 450 can per minute all relay logic. But gate work quite well we get maybe 5 or 10 cans jammed in machine and only takes a minute to clear out.

Is there anywhere you can put a through beam photo or laser detector that will be just above or next to a can being seamed, but if the can gets shreadded the flailing metal will break this sensor.

And if it is next to the can, only monitor it during the actual seaming process, not while the can is moving afterwards.

Another option (taken from the stamping presses) would be to place a metal spring probe on the head so it would be grounded out if it touched any part of a shreadded can, when touching it could ground out one side of a relay coil to signal the PLC to stop - or you could buy a Ground Detector to monitor the probe.