Measure feed level in a hoper

I have a Arduino controller feeding my Koi. Has been working well for a year now. I would like to and some means to measure the feed in the hopper.

The hopper is a large funnel. The feed is dry pellets, about 2 mm in diameter. 10% steps would be nice.

Any ideas how to sense the level?

Surely depends on how your hopper looks like. If possible, It think hacking a digital kitchen scale could work (if the funnel could somehow sit on the scale).

Some sort of range sensor at the top of the hopper? IR or ultrasonic, not sure which would be better. You'd of course have to apply some formula derived from the hopper geometry to translate height of feed into volumetric fill percentage.

Otacon2k - Weighing the hopper is a good idea until you have to deal with practicalities, such as the coupling to the feed motor etc. Still looking for a simpler method...

ajb - A range sensor may be okay. I have looked at them and they seem tuned for longer distances, and harder more reflective return surfaces.

I have thought that a ball, 'floating' on top of the feed would be a good reflective target or house a magnet? A ball would not collect any feed and could be slid to the top when the hopper is filled. The ball could follow a wire down the side of the (copper) hopper (funnel) and activate sensors (reed relays)? If I used 10 reeds that would take 10 digital inputs... Maybe I could form a resistive ladder that presents a stepped voltage, as each reed is activated, to an analog pin instead... I am liking this concept...

So far that is about as close as I get to a workable solution.

Thanks for your thoughts... Any other ideas?

I'm unclear about the dimensions of the hopper, but how about a simple float switch (ball-cock), but with the arm attached to the shaft of a pot?

a large funnel

The last funnel I saw was attached to a cruise liner :D

You could use IR, but instead of using it for measuring distance by reflection, you could have a series of break beam sensors, either on the outside of a clear hopper, or on the inside.

how about ldrs? gona be darker with food covering it, and lighter wiv no food covering it!

or if you felin adventureous you could use a piezo system.

AWOL -

but with the arm attached to the shaft of a pot?

From my experience pots are prone to get dirty/noisy.

phil - IR beams may seem okay, but with time the sides of the hopper become covered with a course "flour" of ground pelets. That would lead to false detection. What is an "ldr" by the way?

PS The feed 'funnel' would be about 60 cm top and 60 cm deep with a 1 cm opening at the bottom. Copper is the material of choice... so far. The unit I have in operation is made of rubber, the same rubber used for pond liners.

http://en.wikipedia.org/wiki/Light_Dependent_Resistor

Not a bad idea - if there were sufficient contrast between the colour of the feed and colour of the hopper, with sufficiently constant illumination, you could use the amount of light reflected to determine the fullness.

Copper? You're not keeping cephalopods then.

ACtually already use an LDR (Light Dependent Resistor) to determine sunrise a sunset to turn on/of other water feature pumps and UV filter.

Not sure the light reflected, even if illuminated would be stable long term. The feed is much like dry cat or dog food, so little to no reflection. The funnel sides would change with time, as the feed flour coats the sides.

All good ideas... and thanks for your suggestions. Any more ideas?

How about some version of this sensor?

http://www.sparkfun.com/commerce/product_info.php?products_id=9091

http://spectrasymbol.com/typo3/site/en/softpotsplash.html

I'm not sure if the weight of the feed would give enough variation but it sounds like it might. Just run the strip down the slope of the funnel.

By the way, is there a cover on the hopper?

Or, how about whacking the hopper with something, solenoid, etc., and measure the resonant frequency. My quick test with a pot full of Grape Nuts yielded a noticeable frequency change. Some sort of variation of Helmholtz resonance.

Emily Thanks for those links.

The magnetoPot seems like it would work very well. http://www.spectrasymbol.com/typo3/site/en/magnetopots.html

I have requested samples. Thanks!

You're welcome. That's an interesting looking sensor but it's not clear to me how you would use it for your level sensor. Would you arrange the magnet to "roll" down the side of the hopper, over the strip?

Would you arrange the magnet to "roll" down the side of the hopper, over the strip?

The hopper is copper. The sensor tape would be applied to the outside. A ball on a guide wire, with a magnet inside would slide down the inside on the opposite side of the sensor as the feed level fell.

The company has kindly offered to send a sample. I'll let you know how it worked out! Thanks.

Oh, I get it. That's pretty cool.

I did a little experiment earlier in an oatmeal box with a little speaker and an electret microphone and you can see the resonant point if you sweep the speaker up the frequency scale. Maybe a low pass filter and a little code and the Arduino could figure it out.

I'll have to play with it a little more later.

Get some acrylic tubing, and make a zig-zag array like a marble run.

/
\
/
\
/

(Not so steep, of course)

The you could use LEDs and photodiodes or LDRs to tell which tube have food in them.

It could even be decorative: light up the tubes with RGB LEDs at night, and vary the color :slight_smile:

Ran

I like the idea of making it decorative as well as functional.