Page 1 of 3 123 LastLast
Results 1 to 10 of 27

Thread: Water Level Measurement

  1. #1
    Automated Home Legend Paul_B's Avatar
    Join Date
    Jul 2006
    Location
    Eastbourne, UK
    Posts
    605

    Default Water Level Measurement

    I've had a long standing requirement to use Idratek to sense the amount of water in my 600l rainharvesting tank. Ultimately this would form part of an irrigation system as well as turning the guttering diverter to collect or bypass rainwater.

    As a proof of concept I experimented with a Maxbotix Ultrasound sensor which should be able to output a voltage from 0 to 2.55v which feeds a QAI (scaled to 3V as opposed to the normal level). I didn't get very good results out of the Ultrasound device and I am about to revisit this issue and start off with it connected locally to a serial port and hyperterminal.

    Anyone else experimented with this type of sensor?

    Paul

  2. #2
    Automated Home Jr Member
    Join Date
    Nov 2008
    Posts
    25

    Default

    Paul,

    Is it essential that you measure absolute levels all the time? I just wondered if there were some way you could fit a number of cheaper level switches say 1 every 100L and gate these into an ODI.

    I did a project many years ago for a health care company which had multiple tanks of different chemicals and volumes and this actually worked out quite well as we realised that effectively having 'set points' was enough.

    I just know some of the level sensors are very well priced such as Farnell 167-3638.

    Regards,
    Peter

  3. #3
    Automated Home Jr Member RichardC's Avatar
    Join Date
    Mar 2009
    Posts
    29

    Default

    I hope that the following is of use. I am going to put in something similar soon for rainwater and for pond level. So will be very curious as to how you get on.

    In the MaxSonar FAQ they say there is a waterproof sensor option for tanks etc, so doesn't look like you are trying to do anything too unusual.

    Also they also have an comment about cancellation further up the FAQ when there are multiple return paths for the sound.


    When working with commercial ultrasonic sensors (Vega etc) then a couple of rules of thumb I used with water tanks. Even with some of the very pricey sensors it can seem to be black magic. Even when you can tell the sensor about the reflections they can go a bit dicky!

    Keep transducer perpendicular to liquid.
    Also not too close to the tank wall. Sometimes one 1/3 out works best. As putting it in the centre of the tank can cause more issues.

    The sensor you are using says that it has no dead-band, but still give it space at the top. With a short distance Vega then 150mm would generally work. This works well with the tube method below.

    Sometimes to provide the sensor with a way of controlling strange reflections is to put it in a tube, with a breathing hole at the top and a way for liquid to enter at the bottom. The tube width with a Vega ultrasonic is min of 40mm dia. With radar there is a formula, but I normally cheat and ask the manufacturer for the dia of tube.

    Even with a commercial probe used in a control system, I wouldn't trust it completely. We would normally use HH and LL (high high, low low) switch sensors to provide a final stop. Getting a sensor in at the side of the tank wouldn't be good I guess in your application, we have used "bull's bo**ock" style sensors for underground tanks with success.
    like http://uk.rs-online.com/web/search/s...duct&R=3389801

    I hope that's some food for thought.

    Cheers

    Rich

  4. #4
    Automated Home Legend Karam's Avatar
    Join Date
    Mar 2005
    Posts
    839

    Default

    I don't know if I've suggested this before, but an alternative might be to use a pressure sensor attached to a vertical pipe (tube) immersed in the tank.

    The pipe would need to be sealed well to the sensor port so that the air pressure in the pipe does not leak (typical sensors will be designed to fit well to some standard tubing ID) and I should imagine would work if the tube is not vertical so long as it did not allow water to displace air out of the open end and so long as the open end was at the lowest measuring point (which I'd suggest should be a little above totally empty if pipe not vertical).

    The air pressure in the tube is then an indicator of the liquid height above the opening of the tube. For example 1m of water is 10kPa. You can probably find a sensor with suitable kPa range/output which can then feed the QAI.

    BTW you'd need to use a differential or gauge (US gage) type sensor otherwise you'd be measuring atmospheric pressure as well. A quick scan in Farnell for example: product 4197616 is a suitably amplified and temperature compensated 0-15psi sensor (meaning ~ 0-10m). Probably better for you would be a 0-5psi (oh how these units grate on me... ).

  5. #5
    Automated Home Legend Paul_B's Avatar
    Join Date
    Jul 2006
    Location
    Eastbourne, UK
    Posts
    605

    Default

    Thanks for the replies guys. I'm going to continue with the Maxbotix ultrasound device at the moment as I've already bought it. I may have been too confident during the first installation expecting it to just work and see the info on the QAI object in Cortex.

    This time around I think I'll start by connecting the Maxbotix device to my serial port, as documented on their site, to get a feel for the output, range and sensitivity. Once I have confidence in this (and my soldering!) I'll move the test kit to the water tank (it is tall and relatively thin in width), first time around I wonder if I was getting echo off the side walls?

    Paul

  6. #6
    Automated Home Legend Karam's Avatar
    Join Date
    Mar 2005
    Posts
    839

    Default

    What is the nature of the problem that you are currently having? I.e what are the symptoms, and have you tried just pointing at a simple flat target such as hard floor or wall? Could it be a QAI setup/connection issue - can you measure the sensor output directly (without connection to QAI) and see if the voltage varies as expected for example?

  7. #7
    Moderator toscal's Avatar
    Join Date
    Oct 2005
    Location
    Near Alicante Spain
    Posts
    2,036

    Default

    Why not use a tube with a floating magnet and then use magnetic door sensors. So you could have a a sensor at the bottom, one at the mid point and one at the top.
    Many of the door sensors are sealed against weather so should work in a submerged environment.
    IF YOU CAN'T FIX IT WITH A HAMMER, YOU'VE GOT AN ELECTRICAL PROBLEM.
    www.casatech.eu Renovation Spain Blog

  8. #8
    Automated Home Legend chris_j_hunter's Avatar
    Join Date
    Dec 2007
    Location
    North Lancashire
    Posts
    1,713

    Default

    >why not use a ...

    SFAIK, that's pretty much what the Farnell sensor (that Peter used as an example) is ... assuming it's OK with being submerged (?) ...

    Karam's thought about using a pressure sensor sound good - and makes me wonder if we use pressure sensors enough, in other situations ... including the more sensitive ones (AKA microphones) ...

    Chris

  9. #9
    Automated Home Legend TimH's Avatar
    Join Date
    Feb 2004
    Location
    UK
    Posts
    503

    Default

    Is your irrigation tank buried or above ground?

    What’s often done in industrial situations is to use the pressure sensor to directly measure the head (height) of the water/liquid, rather than the pressure induced in the air space above it as this latter option requires a pressurized tank, or tube, as Karam points out. When directly measuring the liquid head, and since it is only for irrigation, the error in measured level due to changes in atmospheric pressure can be ignored so instead of needing a differential pressure (or “dP”) transmitter, a straight pressure transmitter is ok (1).
    In the measuring-air-pressure example, the air above the water in the tube is compressible and the compression & expansion of this volume will also introduce errors.
    Re: the need for a differential transmitter here, I think you could avoid this by applying an offset / subtracting atmospheric pressure from the measured pressure, assuming your pressure transmitter reads in absolute pressure, rather than the more-common gauge pressure.

    Liquid height can be determined from: height (m) = (Pressure (bar) * 100,000) / (9.81 * liquid density in kg/m3)
    And 1 bar = 14.5 psi if you prefer colonial units

    So 1 bar = 14.5 psi = 10.2 m of liquid

    As mentioned by another poster, you may want to consider a second, independent “stop” or “trip” for your system. Invariably things fail and it is common in industry to provide a second method of stopping the pump etc. to avoid damaging it, overflowing the tank etc. If you have two ultrasonic detectors then you can also do some clever things like monitor the difference in level reported by the two instruments and generate and alarm if that deviation is too large – a high deviation may be indicative of one of the devices starting to fail, or coming loose from its mounting etc.

    Commercial radar level gauges can be very accurate, measuring levels in large tanks to +/- a few mm. “Guided wave” transmitters are also used commercially and can overcome potential measurement errors from internal features in the vessel, walls etc. Guided wave is uses either a flexible or rigid probe to guide the pulses down and back. Mounting a “through-air” transmitter atop a tube, or stilling-well, can also be used to avoid measurement errors arising from the internal features of the tank.

    There’s some discussion of guided wave measurement here and ultrasonic gets a mention too.
    http://findarticles.com/p/articles/m.../ai_101414561/

    Sorry for the long post, I seem to have rambled a bit, but HTH,

    Tim.

    1. Atmospheric pressure exerts a force on the top surface of the liquid so the same height of liquid can exert different pressures on the bottom of the tank, depending on the atmospheric pressure at the time.

  10. #10
    Automated Home Legend chris_j_hunter's Avatar
    Join Date
    Dec 2007
    Location
    North Lancashire
    Posts
    1,713

    Default

    yet another way might be to use a hall-effect pulse meter to measure outlet flow rate ... higher the head, higher the flow rate ... could be used to measure quantity delivered, too ... using more than one method a good idea if consequences of failure warrant it !

    Chris

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •