Ultra Sonic HC-SR04 give wrong measures in a well.

Hello everybody, a nice day from Belgium.

I'm using an arduino uno, a led display (2*16) a flat cable (2m) and a HC-SR04 to make some measures.
After some calibrations on the livingroom's table, I had good accuracy between 0.20 and 3.0 meters.

If I install all the things in the well (above water level :slight_smile: ) I have always the same (changing from 110 to 130 cm) measures, even if I lift the sensor up and down.

It's an old well digged in 1870, actualy the water level underground is 180 cm, walls are made of corroded bricks and the diameter is 80 cm.

Well, do you think it is possible that the walls are reflecting the sound ?
Did you experience this problem already ?
What is the angle of utra sonic sound diffusion ?
Putting the SR04 in a 80 mm pvc pipe reaching the water is an idea.
What do you think about?

Can you, please, have look at this.

Many thanks in advance. I will be happy to read you.

Bernard

macrid:
Well, do you think it is possible that the walls are reflecting the sound ?
What is the angle of utra sonic sound diffusion ?

Very well possible - iirc it’s a cone of about 30°. Check the spec sheet for details. This would explain the fairly constant distance you measure.
Does it change when you move the sensor horizontally across the well opening?

I use this sensor to do water level measurements in a stainless steel container - wider than deep, I get very accurate results but must build in a delay between measurements to let echoes die out. Measuring multiple times with too little time in between gives all kinds of weird results.

"Does it change when you move the sensor horizontally across the well opening?"
I will try this immediately, just to reinstall all stuffs in the well.

Thanks

Yessss,

indeed the measurements are changing when I'm moving the sensor horizontaly, giving very erratics readings.
So I'm facing reflections on those good old bricks.....

Now I will test with a PVC pipe. The more or less polished inside of the pipe wil, maybe, eliminate refections.

Cross fingers.

Thank you for giving interrest at my little problem.

Bernard

May also give rise to even more reflections.

Another solution for your water level problem: PVC pipe that sticks in the water, sealed on one end, with pressure sensor. The air pressure inside the pipe is a direct measure of the water level.

Hello,

some usefull informations ---> http://www.maxbotix.com/articles/124.htm

But I will surely NOT buy a 8 inches pvc pipe for that application.

I probably will try the solution you propose. Pessure.
Have you, please, some already tryed hardware reference for me ?

Thanks in advance.

Bernard

You can also have it ping a few times and take a MEDIAN of the readings so that it would be more accurate? Also ignore values that are "super off"?

I may be wrong, this is all new to me as well, but I've had really good luck with my SR04 measurements being correct

if you give up with that sensor: i remember my electronicsteacher telling a story about meat being sold for more than it was worth (waterabsorbtie), so to know how much meat & how much water there was did they measure the capacity. So in your case: measure the capacity between the bottem and the top of the water (a floating wire or something)


might not be helpfull, sorry:

-the walls will mostlikely not send the signal back, why: for a schoolproject we had to make such a robot that avoids objects (i made the best one at my university, hoera me), i found out that when the sensor is almost not "loodrecht" (sorry forgot the english term) that the sensor wouldn't detect it

-couldn't it be the water having trouble sending the signal back?

Succes! please place an update when your project is finished, i'm interested

bieboebap:
so to know how much meat & how much water there was did they measure the capacity.

I think you mean density here.

-the walls will mostlikely not send the signal back, why: for a schoolproject we had to make such a robot that avoids objects (i made the best one at my university, hoera me), i found out that when the sensor is almost not "loodrecht" (sorry forgot the english term) that the sensor wouldn't detect it

That's perpendicular.

This is not entirely true: the sensor has a cone within which it detects objects. A rather narrow cone, but enough to start bouncing off of walls.

It is indeed pretty direction sensitive but far from as exact as say a laser range finder. Those can actually scan the surroundings and create a much more detailed picture of what's going on around them.

-couldn't it be the water having trouble sending the signal back?

No, water is a fantastic reflector. The HC-SR04 is a great detector for water surfaces, especially when you place it right above the water surface, and many people use it for just that purpose. Rough walls like in an old well may cause problems, though.

Hello,

I’ve made some extra tests. If I approach the sensor near to the water, less than 1 meter, the reading seems ok.
Then, I progressively lift the sensor and the measures are becoming erratics from 130~150 cm till the top of the well.

So, they are really reflections on the bricks.

I will try the pressure way.

Can you give me, please, some advice for a captor ?

Thanks in advance.

Bernard.

wvmarle:
May also give rise to even more reflections.

Another solution for your water level problem: PVC pipe that sticks in the water, sealed on one end, with pressure sensor. The air pressure inside the pipe is a direct measure of the water level.

Assuming no long-term gas-exchange with the water, otherwise it would need regular maintenance.

Perhaps a PVC tube full of water, sealed both ends, hanging from a load-cell?

MarkT:
Assuming no long-term gas-exchange with the water, otherwise it would need regular maintenance.

Perhaps a PVC tube full of water, sealed both ends, hanging from a load-cell?

That's maybe even worse from a maintenance point of view with algae and so growing on the pipe, changing its weight.