Measuring my well depth

[Blue Dog]

So here's my thought- if I drop a pipe into my well (30ft deep), leave the bottom of the tube open, attach a sensor to the top and seal it so as the water level rises and falls the air pressure in the tube will change. Not a new idea but I'd like to make it digital based on a PIC (I almost drained it filling the pool). Does anyone know of a pressure sensor that would interface with a PIC AND has NPT (national pipe threads) so I can easily attach it to standard pipe fittings? Accuracy is another question I need to ponder.

 

[ericgibbs]

hi,
I dont know of a standard thread fitting sensor, but there are a numbers of pressure sensors, small enough to mount within a pipe.

Regarding accuracy, changing ambient atmospheric pressure will have a big effect on the perceived air pressure in a sealed system.

Is your well pump a submersible type?

 

[Blue Dog]

- not submersible pump (in the basement utility room). Good point about ambient air temp- may try a schrader valve to add air until max pressure is read- got any part #'s I commonly used?

 

[upand_at_them]

What about using an ultrasonic sensor?

Mike

 

[ericgibbs]

hi,
As you plan to place a tube inside the well [ plastic downfall pipe about 2-3 inches o/d, at a guess], why not do as Mike suggests, use ultrasonics?

There are many proven projects for ultrasonic water level measurement, just use the pipe as a 'wave guide'.

 

[Blue Dog]

Not too familiar w/the technology- will it read to 30ft?- you've got my interest- any suggested devices/notes? - actually thinking of 1/2" i.d. pipe if I go that route

- wave guide? never heard of that one

 

[bobledoux]

Water Depth Measurement

There are standard measurement tools based on measuring the water depth using a pressure tube.

A tube is run to the bottom of the well. It is pressurized with air until the pressure doesn't increase because air is bubbling out the bottom of the tube. The air pressure required to achieve this state is a function of water depth.

I've seen pressure gauges calibrated to report the water depth in feet or meters.

Here is a paper discussing three measurement methods:

extension.oregonstate.edu/catalog/pdf/ec/ec1368.pdf

 

[Blue Dog]

should be able to 'calibrate' the pic to reflect close to water depth

 

[dch222]

Regarding accuracy, changing ambient atmospheric pressure will have a big effect on the perceived air pressure in a sealed system.

So will temperature

 

[mechie]

Water Depth Measurement

Here is a paper discussing three measurement methods:

extension.oregonstate.edu/catalog/pdf/ec/ec1368.pdf

This paper misses one common, simple technique that could be considered...
If you arrange a series of metal probes at different depths, the longest being a common 'earth', the others staggered by maybe a foot (any convenient distance).
A series of simple conductivity probes like this needs simple electronics to interface, it is inherently digital, so ideal for a microcontroller system and is cheap.
The downside is that it only tells you how many probes are submerged so the distance between probes will be a compromise between complexity (lots of probes) and scant information (only a few probes) - you could set them at uneven distances such that you get good resolution at near-empty and near-high level, ignoring the bit between.

This may sound crude but is used in industry quite a lot !

 

[eblc1388]

My idea:

Use a long single core PVC covered cable, with one end attached to a stone and the end sealed off to prevent contact with water. It is lower vertically into the well till it reaches the bottom, keeping a few inches clearance from the side of the wall. The part of cable in contact with water forms a capacitor between earth & the inside copper core and its capacitance will be proportional to the length it is in water.

Then this "variable capacitor" is connected to an earthed oscillator forming part of the oscillation circuit and a PIC measures the oscillation frequency to determine the portion of wire that is in water.

 

[Blue Dog]

I'm liking the post about the var cap- perhaps try a 'dry' run using the pool. - thanks all for the responses.

 

[ericgibbs]

hi blue dog,
If you go the var-cap route, as the cable is acting as a capacitor referred to ground, you dont need to use a sealed end,
just use a long cable loop, with a weight at centre of the loop, near the well bottom.

That is have both ends of the cable at the well top, avoids the water seal problem and increases the capacitance.

I have never tried cap/well measurement, Lets know how it goes.

 

[3v0]

Sounds interesting. If you do have sufficient resolution with a single loop make several loops. But you knew that.


I would be interested in using a similar setup to measure the water level in ponds.

One should be able to make several runs up and down the probe to increase the sensitivity to water level. For a well a resolution of feet would be good enough. I wonder how many turns would get it down to inches.

 

[eblc1388]

Well the theory is simple but there is the problem of initial oscillation frequency changes which affects capacitance calculation.

The capacitance of the sensing wire is well defined and will not change itself but the oscillator and its component will change with time and temperature.

Therefore, to compensate for this effect, the circuit should first start an oscillation without the "sensing wire" in circuit, and the initial frequency recorded in the PIC. Thereafter, a known value of capacitance is switched into the circuit and the new oscillation frequency measured. This enable a ratio of frequency change upon a fixed capacitance to be calculated.

Then the sensor wire is switched back into the circuit and the new frequency measured. Using the ratio obtained earlier, the water level can be calculated. Using this method, changes in the oscillator frequency will not affect the final result.

 

[ericgibbs]

hi L Chung,

What if a frequency to voltage convertor was used to give a dc voltage
that could be used to drive an analog meter or 'old' panel meter, [opa and offset corrected to suit]

No PIC required??

What do you think?

>> eblc:
The aim is to manually tune the oscillator frequency to a known frequency without the "sensing wire".

EDIT:Yes I agree, I would like to see some of his data, if he decides to use this method.

 

[eblc1388]

It will work if another variable capacitor is connected in parallel with the "sensing wire" and set at the default 50% of its total capacitance. After user disconnect the sensor wire, the meter can be zeroed or calibrated to a certain mark using the VCap. This can be done manually via depressing a "CAL" switch everytime before measurement.

The aim is to manually tune the oscillator frequency to a known frequency without the "sensing wire".

Edit: the frequency change will be small, so a "frequency discriminator" rather than a FV converter would be a better choice in this case.

 

[Pommie]

This is an interesting idea. I'm wondering if a slight variation may give better results. If a coil is wound around a length of say 1" od by 1 or 2 meters long PVC pipe and that was lowered into the water, would the change in inductance be more readable. This would also not require an earth return which I guess would cause a seriously variable ESR due to ground wetness. A recent thread had an LC meter that would be a good starting point.

Mike.

 

[philba]

capacitance is pretty easy to use, beats putting in a bubbler system. You can go the oscillator route or, if you use a microcontroller, you can simply time the charge or discharge rate. the nice thing about capacitance is that it is linear with the depth.

Probe construction is tricky, though. Use teflon insulated wire - you wont get muck and slime growing on it (changes capacitance). You should also make the wire fairly rigid (use tension) so that it retains it's shape after you have calibrated.

 

[mneary]

I guess I've gotten lost in all these notes about calibration, accuracy, stability, etc. iirc, you want to know in advance when the well isn't refilling as fast as you're pumping. A dry well can ruin a pump, and the water isn't so good for a while.

If it's as simple as I just described, choose a 'danger' depth, and run a float down the tube on fishing line. When the float floats no more, a simple microswitch will tell you to give it a chance to recover.