But Inodor Pereya seems to not understand the concept of pressure.
Inodoro.
Well, thank you so very much Klaus, for your little pressure lesson. I would say that, after being in love with physics for the last 40+ years, I have a pretty good idea how pressure works in fluids. You, on the other hand, seem to be a little confused.
"Barometric pressure" is defined as "the weight of the air column on top of a body", here's, from wikipedia:
Atmospheric pressure is the force per unit area exerted against a surface by the weight of air above that surface in the Earth's atmosphere. In most circumstances atmospheric pressure is closely approximated by the hydrostatic pressure caused by the weight of air above the measurement point.
http://en.wikipedia.org/wiki/Atmospheric_pressure
And here's what they have to say about "weight" (among other things (it's a long article)):
The definitions of the physical concept of weight given above define it as a vector quantity, having both magnitude and direction. For an object at rest on the surface of the Earth, its weight is a force that points down, approximately towards the centre of the Earth.
http://en.wikipedia.org/wiki/Weight#Vector_or_scalar
So, being that a load cell (or any other kind of spring loaded scale) measures "weight", it's gonna be specifically sensitive only to the vertical component of the atmospheric hydrostatic pressure, meaning that, while it's true that pressure in fluids is exerted in all directions, the only component of the barometric pressure that's gonna act on the scale is the vertical, downward component, the "weight", which is the only component the load cell measures.
So to me this says that they don't calibrate at the factory becuase the calibration would be off due to gravitational changes caused by your distance from the center of the earth. This also affects air. Which makes total sense.
Agree?
Really? you mean, like I said here?:
I don't have the formal training to give you a scientific answer for that. All I can tell you is that that's the main difference between a spring scale (like the digital scale we're talking about) and a balance. It's not that the spring scale is exclusively affected by barometric pressure. It's said it's affected by gravity, meaning the combination of the Earth's pull on the object being weighed, and the weight of the air column over it.
But that's about the extent of my knowledge on it. Maybe one of the forum's engineers can give you a better answer.
Either way, so far, I think the idea of measuring the impedance between 2 rods is by far the best one, for this application. It's accurate, affordable, and as reliable (in the long run) as you want to make it.
If you do agree than can we tie it into Klaus's statement, that after calibration barometric pressure should have no effect on the scale?
No, we can't. Barometric pressure is constantly changing, due to changes in the thickness of the atmosphere, air temperature, etc, so, even when the scale in site calibration will make it a lot more accurate than the same scale calibrated at the factory, it will still never have the accuracy of a balance, which is the reason why balances are still in use.
Two metal rods may work, however, I am unfimilair with that solution. I understand the prinicipals, but in practice are there some things that would have to be taken into account? I would assume temperature would have an effect on the reading?
Two metal rods do work. The only disadvantages are that you need to keep the water quality consistent (which is just a matter of filtering it), as impurity levels can affect the water impedance, and the eventual oxidation/calcium deposits on the rods (which can be taken care of by using 2 SS rods, and keeping them clean). Temperature does affect the reading, but water temperature is normally fairly stable, so the differences are negligible.
I love to watch these "Flat Earth" discussions as they wander around the original post issue. The original question was about the use of a pressure sensor and PID for level control, which is the method of choice in industrial applications, load cells and scale transmitter second. The resistive approach and rotary encoders are virtually non existent because of inherent difficulty in setup, non standard outputs, and need for stilling well for float apparatus.
Well, I'm glad you're enjoying yourself.
The fact that a solution is not used (here) in an industrial application, doesn't mean that solution is not good. There are different reasons why that may happen (like your standardized output comment) that may not apply here.
Both the pressure and load cell approach share two major drawbacks: they lose calibration over time, and they need dedicated electronics to work.
After bouncing ideas for days, I still consider the resistive setup to be the best option, for several different reasons:
1. "Calibration" is only affected (providing the water quality is consistent) by the electrodes' shape, distance to each other, and cleanliness.
2. As long as the electrodes are straight, output is 100% linear.
3. There are no special electronics needed, to the point that, if the user wants to use a microcontroller to automate the whole process, he can connect the rods directly to an analog input. From then on, it's all programming.
4. The system is extremely rugged. There's no mobile parts, no deformable parts (like the spring in a load cell, or the diaphragm on a pressure sender), nothing. Just 2 metal rods and a length of wire.
5. You can easily adjust the sensitivity of the system, just by changing the position of the electrodes.
6. The system is, literally, dirt cheap.
None of the other systems can match that.