Hi everyone!
I was new to these forum so I do apologize if there are rules of some sort that i have not followed.
I was planning to count the number of small fish pellets passing on a transparent tube that was being sucked by a water pump. Is there an existing sensor for arduino that is capable of counting such small objects considering that it was suspended in water inside a transparent tube at a relatively fast velocity?
I was trying my best not to resort in camera detection.
Hellonoob02
Did you try to use an IR light barrier?
Hi,
give more information about the "little ball",
such as diameter, color, and if possible estimate the speed at which it falls.
Does it only pass one ball at a time through the tube?
Is it possible for two or more balls to pass together without space between them?
I was still on the planning stage so I really don't know what sensors are the best for this application. I was worried about the fact that it was inside a tube with water which might not be applicable for some sensors.
me too.
The pellets would be around a milimeter, the color would vary from green and red. Just a generic fish food pellet. The speed would be estimated around 2 - 4 feet/second
I suspect water bubbles may be an issue?
I'd use a BeagleBone AI and a camera.
It would also be possible for 2 or more pellets to flow at the same time since the tube is bigger compared to the pellets
There's also that. Thanks for the suggestion. If there's no other choice, I might use the camera detection as a last resort.
Hi,
You can use a small laser emitter on one side of the tube and a detector on the other side.
You will have to make these devices waterproof, perhaps using hot glue.
I used an Optek OPB70DWZ to detect a 0.5mm black line on a yellow background as documented here so it should work for a 1mm pellet.
This was the reflective version of the sensor. If that doesn't work, you can shine an LED from the other side of the tube.
There seems to be many resources on the 'net:
detect particulates in water flow at DuckDuckGo
I suspect the issue will be aggravated by increased flow rates and flow volumes. That is, the ratio of particle diameter to flow pipe diameter is of primary concern: pretty easy to count bowling balls going down an alley, but toss a jar of jelly beans and discrimination is nearly impossible.
A possible solution would be to craft (3D print or injection mold) a diverter that flattens the flow and have multiple detectors per channel passage.
Concept:
"flatten" the liquid flow into a rectangular area of shallow depth the channels being of a machined size to cause particles to be linear flow, one behind the other.
The total numbers of channels required would be based on flow rate through the conventional piping: channels will divert flow, thus the faster the flow, the more channels required to maintain a minimum flow and reduce channel pressure.
A photo-transistor or pixel-like matrix underneath the flat panel would serve as a one-one detector. Electrical processing of the photoelectric elements would feed a microprocessor- realizing some pre-processing may need to be accomplished.
There are numerous concerns in taking a laminar flow of liquid through a flat multi-channel diverter. Ensuring that particulates are uniformingly being distributed and that channel machining is large enough to avoid restrictive flow but small enough to prevent adjacent particle binding.
Needless to say, such an approach is complex but successful manufacturing should yield accurate results.
If accuracy can be sacrificed, then a calibrated light source passing through a transparent tube would have light absorbed based on the density of the particulate and an average could be derived for any given flowrate.
Lasers, cameras and software algorithms are currently used in calculating particulates in the atmosphere.
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.
