Can counting sensor counting extra

Karinrinkashi

Member
Join Date
Nov 2017
Location
Atlanta, GA
Posts
40
Hello everyone,


We have can counting sensors on all our lines, i am sure all of us know why. So, one of our lines has started counting more cans than we are producing and our yield has gone way over 100%. We changed the sensor, re-did all the wire connections but did not change the wires. The count goes into a micrologix and from there it goes to our Cloud storage (another company that deals with that side). But before we call them in to investigate, i wanted to make sure that everything on our side is good. What else could be the issue, if anyone has any experience with this. I checked the controller, and it is obviously doing its job because our other lines are also routed to it and are showing reasonable numbers. Could it be the wiring? But the wires look really good and not much wear and tear, obviously who knows what is going on inside the conduits. The microcontroller input light blinks at good intervals and no drag in the blinking or anomalies, but i am assuming going by the controller input lights is not a good way to troubleshoot due to the high speed that it is receiving the inputs at.

Any help would be highly appreciated.

Thanks

Karin
 
Is the sensor photo proximity or inductive metal detecting?

I have had a small piece of paper stuck to a conveyor belt causing extra counts.
(As in check for anything that might be getting in between the cans and the sensor.)

Particularly with photo-detecting - any chance the label has a very dark spot in it?
(This could cause the detector to 'see' the can before the dark spot. Not see the can
during the dark spot then see the can again (and count it twice) after the dark spot.

If the sensing range is too long the counter could conceivably be counting something
behind the cans it is supposed to be counting.

Poet.
 
Optical sensor? Take a closer look at the sensing area. Is there a piece of metal adding reflections?
Check power supply voltage, and verify it doesn't have a lot of AC ripple (noise) from aging caps.
 
With out knowing more info like if the cans are back to back with little gap, or sensor type or speed. Is there a backup sensor close to either upstream or downstream that you could use to put a counter on? That way you could compare counts on both the sensors.
 
Thank you all for your responses.

The sensor is not optical, it is just a object detecting sensor. I am attaching an image of the sensors. I do not think we have a back up counting sensors at all. I will check for any debris.

Karin

Capture.PNG
 
Can you provide model number? It helps if we know what technology is used to detect the object. We can determine that quickly by looking up the specs on the sensor.
 
Have you contacted Sencon tech support? Since this is a sensor designed specifically for the application I would think the manufacturer must have some troubleshooting tips or deployment guidelines, especially since they tout accuracy of 30 counts per million.
 
I second the motion for checking the power supply for AC ripple.
That can cause all manner of problems.

There is a slim possibility that the spare sensor installed is also bad.
I would try swapping the sensor pair from a line that is counting good
with the line that is overcounting. (Both halves of the sensor pair.)

You could also try swapping the input modules the counts run into.

Edit: You could also try putting an oscilloscope on the signals. Check
what the good counting line looks like then look at the problem child.
I would be looking for voltage spikes in the signal - possibly inductive
pickup of extraneous signals?

Poet.
 
I would suggest using a programmed filter on the input. It's possible that (1) you are getting some "bounce" or jitter or other aberrant operation on the input or sensor itself, or (2) you have a VFD lead running parallel to a sensor cable for some distance. Just my 0.14 rmb.
 
Well, it seems that we have fixed it. So following the advice above regarding debris in the belt or in the side rails, i went and took a firm look at it. And i guess, because this time i was really analyzing it minutely, i saw that the one side of the rail was shaking in and out every time the seamer was going into high speed. In low speed the vibration was not so bad, but at high speed the plates holding the sensors were really taking a ride along. So we tightened it up. Those plates are facing the non operator side and are at absolute eye level. So anyone looking at them won't see them vibrating in and out because we are looking into the plane (axis) of the vibration.

That has seemed to fix the problem, the only other thing is did was to rewire the sensor and also rewire all the wire connection from the sensor to the controller.

Thank you all for your help in this, and i hope this thread will save time for someone else in the future.
 
.... i saw that the one side of the rail was shaking in and out every time the seamer was going into high speed. In low speed the vibration was not so bad, but at high speed the plates holding the sensors were really taking a ride along. So we tightened it up. ....


The "computers" always get the blame when the maintenance people have loose screws ... 🔨
 

Similar Topics

Hi, I have a sensor that is FALSE when it is triggered, and a setup with boxes running past the sensor. I need to count how many boxes go past...
Replies
3
Views
1,403
Hey All, I am looking for advice on the best prox sensor or device to use in this application. Customer wants to be able to count potatoes...
Replies
16
Views
5,078
Hi guys, I've an enquiry from our mechanical department to help source out a kind of sensor that is able to detect & count the nuts in a nuts...
Replies
41
Views
8,442
I am looking for sensor to "Count" product..Now before anyone jumps in and says use a photoeye..thats what we have..however its not accurate...
Replies
30
Views
6,008
Hello I am looking for tips on how to count the duration of a given function and then how to display it on the hmi panel in the hh:mm:ss format...
Replies
4
Views
1,687
Back
Top Bottom