Detecting empty/low fish feed in a hopper

Other Hardware Sensors
1

I have a fish feed hopper which looks something like this picture . The approximate weight of fish feed in the hopper is around 40 kgs. The fish feed pellets can themselves be 1mm to 6 mm in size. The hopper itself weighs around 60kgs. The legs are smaller than the picture. The height of the bin part in the hopper is 2ft, while the funnel part is 1 ft.

I am trying to determine when the hopper is empty or low on fish feed (either works). I have thought about couple of ways to go about this:

a) Use some form of load cell (four half bridge or standard load cells), but have read at quite a few places that it drifts under constant load and might need re-calibration quite frequently, which is not feasible here.

b) Use ultrasonic sensors to get depth, but then again they don't seem to work well with pellets where they don't get a even surface to reflect. (Again, I might be wrong here).

Other sensors like pressure sensor, capacitive sensors or IR sensors also comes to my mind, but haven't used any of them till date so not sure of the effectiveness. I wan't to keep it simple, with far lesser need of recalibration. One thing to keep in mind is that side of the hopper would get covered in dust over passage of time.

Any help or pointers would be highly appreciated.

hopper
hopper349×1024 56.4 KB

2

Use a light barrier at the output

3

Hello amitabhkant

Simply place the fish feed hopper on a scale.

4

I'm surprised ultrasonic didn't work. Have you tried a ToF sensor like this VL6180 Hookup Guide - SparkFun Learn

5

I, too, would have expected an US detector to work. Have you tried this Amitabh Kant?

6

By scale, do you mean using load cells ?

7

Honestly no, but just read at multiple sites that US detector do not work well in such scenarios. But now that you two have said it, will try it on Monday and get back with the results.

8

Will also check this if US detector doesn't work

9

or something comparable

10

If there is no significant drift under constant load with load cells, it would be the ideal solution. I might have to test it out though to see the actual drift.

11

The proof of the pudding is in the eating :grinning:

1 Like
12

You can eat anything, at least once.

13

OMG. I have to say thank you for getting the saying right. All those "the proof is in the pudding" people just put me into "grrr" mode.

14

IIRC, the drift is relative to the weight and the load cell rating. i.e., if you have a 100lb weight on a 100lb capacity load cell, it will exhibit a lot more drift than if it had a constant 20lb load.

I think I read that before.

I built a fluid dispenser that dispensed according to weight and I didn't see any appreciable drift. But the full container was never left on the load cell for more than a day or two and the tare weight of the container was just a few ounces.

15

You can buy these for beehives to measure population loss/gain and honey production. The load cells drift, if you use 1/4 or 1/2 bridge, full bridges are stable - that is, if your carrier does not give in.

1 Like
16

So, US detector works, though because of the funnel shape, the food also forms a funnel shape and varying values are returned, but that I hope to handle it through code. At some point , I would like to test out other options too, but that would be at a later date.

17

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.