Sensor for waterpressure
Moskog
Posts: 554
Does anyone have any ideas on where to find a good and cheap sensor for measuring waterpressure?
The idea is to detect water pressure in a tank to be able to calculate the level of water.
The sensor could be placed in the bottom of the water tank, then, lets say a BS2 could read this sensor, digital or analog (through an ADC), and then display the level. (Full, 1/2, 1/4 or so....)
The water pressure will be like let's say in a standard water heater (not connected to supply with pressure) or in other words maybe like on one to two meters level.
Ideas will be appreciated!
KjellO
The idea is to detect water pressure in a tank to be able to calculate the level of water.
The sensor could be placed in the bottom of the water tank, then, lets say a BS2 could read this sensor, digital or analog (through an ADC), and then display the level. (Full, 1/2, 1/4 or so....)
The water pressure will be like let's say in a standard water heater (not connected to supply with pressure) or in other words maybe like on one to two meters level.
Ideas will be appreciated!
KjellO
Comments
-Phil
If by "standard" water heater you're talking about a standard household water heater, like the kind that's common in the USA, then the hot water in the tank IS under pressure, roughly speaking about 40 psi. About 2 meters of elevation would give you, roughly speaking, about 3 psi. Depending on what kind of accuracy you need, I would make sure my pressure gauge is of the "gauge" type and not absolute - because absolute measurements can be affected by changes in air pressure, too, though for water it might be an insignificant difference.
You got me thinking. Then I did a search. It seems like you'd need to spend hundreds of dollars... [noparse]:([/noparse]
Then I thought a bit more.... How about using a Flexiforce sensor underneath a column of water? It's $26, versus hundreds, and you'd have an easy way to hook it up to a ucontroller, and it wouldn't care what kind of pressure the water was at. Not sure how to hold the water right now, but some sort of plexiglass (or metal) tube that sits directly on the Flexiforce, and is connected to the water volume with a flexible tube. Not sure if that came across clearly... The more water in the tube, the heavier it would be, and the more force it would apply to the Flexiforce. You could tune it by changing the diameter of the tube (for more or less weight. You could also cap the tube, so even if the water was under pressure, it would work (I think, didn't think that one through enough).
-Parsko
Which do you really need, pressure or level?· If it's just level, then here's a real cheap and simple solution = Stick a DYI probe in there - and let the conductivity of water do the work for you, no calculations needed.
I'm thinking you're on a farm or so and that this tank is raised to give you·gravity-fed·water pressure? - the height gives you the pressure?
I gather you're just using it as a storage tank and not heating it, right?
If it's an old UPRIGHT electric tank,·there's likely an extra screw plug on top where the ANODE goes. OR·If it doesn't have this, and it won't be under much externally created pressure, you can cut a hole in the top.· The level probe goes in there - once placed, you can rig threaded fittings - or simpler, just EPOXY it closed.· This will work even on pressurized tanks, as long as the seal is good and the glue strong.
1.Take a 1/2 PVC pipe that's·the HEIGHT of the tank - the end·should rest inside on the tank bottom.
2.Figure out where the top of the water level is when full, mark that on the pvc pipe.
3.Determine the level resolution you need (1/4, 1/8/ 1/16), then divide by that many intervals, marking on the pipe from bottom to top water level.
4.Wrap a strip of copper roofing around at each mark, or cut a slightly larger pieace of copper pipe in half --- each section should be ~1cm or so.
5. Take the strip back off and cut each in half - and cut off about 1mm from from each (or use the copper pipe already cut in half, and shear off more of each) --- glue each half on the mark, but make sure the two pieces don't touch... that will be the water's job!
6.Drill two holes at each mark, 180 degrees apart through the copper and into the pipe - or just drill through both sides.
7.Run ONE wire down the inside of the pipe and pull a length through one hole at each mark - that's you sensor common.
8.Run·separate wires for EACH level, so if you have empty, 1/4, 1/2, 3/4, full ... you'd need 5 wires: 1 common 1 for each level, making 4.
9.Put a screw around the wire and into the copper - right in the holes, and glue (not solder) for contacts... if this is drinking water make sure you use an expoxy the is inert and safe for potable water - many are, just ask the manufacturer.
10. By now, I think you can guess the rest: Take the wires that come out the top and hook them up to your level trip circuit.
\\\\
|:::|
=· =
|::\|
=· =
|:\:|
=· =
|:
=· =~~
|· ·|~~
~~~~~~~~~~~
Howard in Florida
~~~~~~~~~~~
| |
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
No matter how much you push the envelope, it'll still be stationery.
ElectricAye.. yes, the water in a water heater is under pressure usually, from the external supply. Using that example confused up the question a little bit, it was just meant to explain in a simple way what I really want to measure. Its just the level of water in a water tank. Cold water in a tank, maybe 1,5 meter deep, sealed or not, the pressure is just the pressure from the water itself.
Parsko.. you give me a good idea there, on the Flexiforce issue, really something to think about.
And Howard.. you are right guessing what I really talk about, and as mentioned above I want to measure the level of water in the tank. Storage tank, no hot water. Your idea sounds very interesting, as I understand your pipe contains somehing that will act like capacitors that will change value as they get covered or uncovered by the water. The higher resolution, the more of those copper-split-rings I need on the pipe. Do you know if those "sensors" can be read by using RCtime on a BS2? In that case, can all wires be tied together on top of the pipe just to save input-pins on that controller, or will I need 4 input pins to measure four levels of water?
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Searider
Now this is probably the best idea. I went through the product info on the Ping and found out the range is from 2 cm to some 3 meters. I even found out Parallax is making a new, waterproof ultrasonic sensor, in an almost parallell thread in the Basic Stamp forum:
http://forums.parallax.com/showthread.php?p=811162
I guess that's the thread you, Searider speak about.
The Parallax forums are really awesome!
As a follow up to the partial brain fart I was having regarding if the vessel was pressurized...
You could simply tap the top side of the tube (that is holding the cylinder of water that was changing in height) back into the vessel. So, there'd be two (flexible) tubes, one at the top, the other at the bottom, and you simply measure the change in weight. You could optimize it so that the whole assembly was around 23 lbs full of water. If you used PVC, I think you'd get pretty good accuracy (since the water would the a greater percentage of the max weight).
-Parsko
I hope this helps.
http://www.parallax.com/Portals/0/Downloads/docs/cols/nv/vol1/col/nv27.pdf
Stephen Wagner
Interesting about the water-tight ping sensor! Water condensation would probably ruin the old style. ( I thought of using one outside once, but changed my mind after getting one --- thanks for posting about the water-proof version... I need to revist that project! (((An aside to ALL - isn't this great how we can toss ideas around like this! )))
Do you have to do this electronically? I mean, can you just *look* at the tank? You could attach a length of clear, small tubing at the tank bottom and run it up the side to the top ... it will fill to the level of the water inside. If electronics needed, you could put leds and photo whatever's to read them --- even press the tubing inside a sender/receiver unit placed at each level location.
> your pipe contains somehing that will act like capacitors that will change value as they get covered or uncovered by the water.
It's even more simple - water is conductive, each copper ring spot is a water-activated switch.
> The higher resolution, the more of those copper-split-rings I need on the pipe.
Yes, exactly.
> Do you know if those "sensors" can be read by using RCtime on a BS2?
You could probably do that, or make a resistive T-ladder and Cap for RC timing it.
> or will I need 4 input pins to measure four levels of water?
That would be the fastest way ... you'd just have to experiment with the pull-up (OR) pull down resistors to get the tripping right.
cheers,
- H
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
No matter how much you push the envelope, it'll still be stationery.
You could then measure the resistance using an ADC, or RCTIME, depending on your distance and needed resolution.
Hope this helped.
It seems to me that any tube flexible (elastic) enough to allow you to weigh its contents would also be flexible (elastic) enough to stiffen up once it is under pressure, and this stiffness vs. pressure phenomenon would mess up your weighing system. Under pressure, your flexible tube would suddenly become somewhat springy, the springiness of which would then depend on temperature, aging of the material, etc. The whole thing of compensating for various system interactions then gets very complicated very fast.
my 2 cents worth.
Just thinking out loud here:
·
I worked for a Ward Packing Co long time ago·as Machinist apprentice. Their main drinking water tank controller was three electrodes in the water tank (over the main offices). For some reason the tank would run dry and the head machinist was trying to figure out why. I told him I had some EE background and would be happy to look at the circuit. Turns out the circuit was a CMOS RS flip-flop on a wire rap board in an area that had a lot of sea air in the control room. I explain to him how it worked and ask him if he wanted me to fix it. Well the “Head Master” electrician had a cow and no way in heck was I suppose to have anything to do with fixing the problem! And I warned the Head machinist that if it failed it could be either lock the pumps on, or off, depending on how if failed. I was told by the electrician I did not know what I was talking about. Two days later, I came out of the bunk house and could plainly see that the system failed. I asked the Head machinist where the cloud was that was producing all this rain that was flooding the offices with 4 inches of rain? He was not happy (that’s because he had 6 inches of water in his bed room.) Turns out the wire going to the connections to turn off the pumps had corroded and could not shut off the pumps.
·
So what we ended up doing is getting three float switches. One of High shut off (and when on, would start pump 1), One for medium level to turn on pump 2, and One for low to turn on pump 3. The cool things about these float switches was there was no way for corrosion to mess up the control lines because switch and wiring were all hermitically sealed. Control was fix by if the·switch ball was tilted up or down compared to the cord going into the ball.·I asked one of the other electricians if there was a possibility for the lines to get tangled up, He didn’t think so. And as for the operation of keeping the tank full, tanglement seems not to be a problem.
·
For our 100 thousand gallon Boiler water supply tank that feed two 125 horse power boilers, and an really cool old boiler that dated back to the early 1900's, I made up for the Company a sliding scale on the outside of the Boiler water supply tank·with a float on the inside, and the indicator balanced up and down according to the level. It would not surprise me if you could find a way to encode that type of system to give you some type of feed back as Desy2820 described. If I remember right, the fun part is most variable resistors only have 14 turns, so you’re going to have to do some gearing to get the number of turn per top and bottom of the tank. I wanted to do this for measuring our lake level, and was looking for a wheel encoder that when turned would give an optical binary·output according to the level of the lake, never got the chance to do it, but had it all worked out. The way we kept·this Boiler water supply tank full was an overgrown 4 inch toilet tank style off and on valve- but I don’t know if that would help in your application. I would still think of using a float ball switch to shut off the pumps to prevent flooding in this type of application would be best.
-D9W·
http://www.parallax.com/StoreSearchResults/tabid/768/List/0/SortField/4/ProductID/590/Default.aspx?txtSearch=water+level
This may not give you the depth sensing range you're looking for, but I thought I'd throw it out there. I believe the manufacturer also makes other ranges available.
The water-tight ultrasonic ranger sounds like a great idea too!
Tim
www.summitracing.com/parts/ATM-3262/
-dan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Arguing with myself--sometimes me, myself, and I don't always agree.
(Former) Caterpillar product support technician
http://golto.com/i-1721171-fuel-oil-pressure-sensor-0-100-psi-brass-body-w-1-8-27-npt-thread-3-wire-connector-5-vdc-input-0-5-4-5-volt-output.html
it is a little expensive on this site at $93.86, but im sure if you look around, you might find one cheaper. most pressure transducers are pretty expensive, especially with higher pressures, $93 isn't relativly expensive.
it's 0.5 to 4.5vdc output for 0-100psi,whitch you could interface into a basic stamp with an adc, like the ltc1298. then just some simple math for scaling.