Need Help - Change sensor to accurately detect box jams

Hi guys....first post here. It's been about 6 years since I worked heavily with PLC's and automation, but I'm getting back into it. I was hoping you might be able to help me with an issue I've noticed at work.


Problem:

The sensor we use just reads if there's something there, or not. A make or break if you will. When boxes are pushed together and there is no gap between them it thinks there's a jam (after 6 seconds) even when the boxes are moving past (as they should).

I'm trying to find a sensor that will see the distance changes from box to box (mounted on the side of the conveyor). Only when there is a true jam and the distance isn't changing will it send a jam signal.





I've tried researching sensors and have been overwhelmed by the endless possibilities and combinations (not all of which I understand).

We are currently using an Allen-Bradley 42EF-P2RJB-A2H

It would be ideal to have a similar style plug, but it isn't a deal breaker.




Here's an idea of what I'm dealing with. (The actual conveyor the jams are occurring on is a decline belt, not a roller accumulator.) There are varying box sixes, but they are all packed together. Creating a gap between boxes or using metering belt is not an option here.







Thank you for any help/guidance you can offer!

https://www.automationdirect.com/adc/Shopping/Catalog/Sensors_-z-_Encoders/Ultrasonic_Proximity_Sensors/18mm_Round_2200mm_Sensing_Distance/UK1F-E7-0A

What about using a ultrasonic sensor with analog output. It looks like the boxes are different distances away from the side of the conveyor. If you see the same distance for a period of time then the conveyor is jammed.

Just a thought.
Regards,

gclshortt said:

https://www.automationdirect.com/adc/Shopping/Catalog/Sensors_-z-_Encoders/Ultrasonic_Proximity_Sensors/18mm_Round_2200mm_Sensing_Distance/UK1F-E7-0A

What about using a ultrasonic sensor with analog output. It looks like the boxes are different distances away from the side of the conveyor. If you see the same distance for a period of time then the conveyor is jammed.

Just a thought.
Regards,

Click to expand...

That's just the thing I was brainstorming.... just didn't know what to use. Thank you!

Is there a product that acts like analog, but outputs digital? If I changed to analog output, then that means running new wire and an analog card

Is there a way to measure belt travel? Information from belt movement (distance traveled) will allow you to measure box length with the photo (on/off) sensor.

i guess u need to improve the mechanical system movement

Tough to do without an Analog sensor. You can get distance sensors that are digital, but around a fixed point, so <50mm =off, >50mm = on. Your boxes might be at 20mm and 45mm and not switch.
I suppose if someone made a rate-of-change distance sensor you could use that, but I haven't seen one before.

google found this sensor

https://www.baumer.com/de/en/produc...easuring-object-edges-widths-and-gaps/c/13559

I can't imagine there is a simple replacement sensor to do what you need, but in addition to what you're already using maybe a simple vision sensor like the keyence iv could help. I would think you'd be able to detect box edges in conjunction with a few height checks (not sure how much variation you have between boxes).

As someone else mentioned a mechanical fix may be best. I would think a slower conveyor feeding a faster conveyor would create separation

Good luck

The boxes on those gravity conveyors are discharging to something, correct? How about placing your existing presence sensor(s) at the discharge point(s). Absence of a box at the discharge point indicates either no boxes being fed or a jam condition upstream.

Is this a classic example of trying to fix a mechanical issue by programming or electrical hardware?

Is there a way to install a solenoid actuated “meter” to insert a gap between boxes?

I agree with Steve Bailey, but I would suggest two sensors. The upstream sensor detects boxes arriving and the downstream sensor detects boxes leaving. If you want a more elaborate jam detector then install several sensors at fixed intervals.
When the upstream sensor is blocked and the next downstream sensor is not start a timer which is preset to the travel time between sensors. If that timer times out you have a jam.

The sure fire solution would be a laser doppler velocimeter with a very wide distance range. It would work by producing a signal proportional to the velocity of the surface passing in front of it or by producing an encoder type signal related to the surface passing by.

If such a sensor exists, it would be very expensive. Those that I have dealt with in the past cost in the $10,000 range to implement and have a quite tight distance tolerance. They are extremely accurate, and you don't need that level of precision.

I said it then (a decade ago) and I will say it again, how does my $15 optical mouse work and why can't I buy an industrial sensor with similar technology for under $500?

Another thing to look at might be a low end vision system.

Get your local reps for IFM, Keyence, Banner, Baumer and others into your facility and let them compete with each other to provide you a solution. Often the sensor people will have an idea or a product that you might not run across from your own experience. Often they will propose things that simply don't work, so ask for a free trial, PO pending success.

dogleg43 said:

Is this a classic example of trying to fix a mechanical issue by programming or electrical hardware?

Is there a way to install a solenoid actuated “meter” to insert a gap between boxes?

Click to expand...

I agree. Or a section of powered conveyor that is faster than the section before it thus creating a gap.

Seems simple to me with a distance device, like a laser point measuring device. Should eliminate any non-gap boxes. Programming might be quite fun, or not.

Are there any markers on the box? A sensor that distinguishes between black/white and box brown?

The picture you posted is it your application? Than this would be feasible?