Welcome to the Cumulus Support forum.

Latest Cumulus MX V3 release 3.28.6 (build 3283) - 21 March 2024

Cumulus MX V4 beta test release 4.0.0 (build 4018) - 28 March 2024

Legacy Cumulus 1 release v1.9.4 (build 1099) - 28 November 2014 (a patch is available for 1.9.4 build 1099 that extends the date range of drop-down menus to 2030)

Download the Software (Cumulus MX / Cumulus 1 and other related items) from the Wiki

Weather related sensor

For discussion of DIY weather equipment - sensors, accessories, improvements to existing kit etc
Irish Steve
Posts: 15
Joined: Thu 09 Dec 2010 4:01 pm
Weather Station: WH1081
Operating System: XP SP3

Weather related sensor

Post by Irish Steve »

Folks,
It's not for a weather station, but very much weather related, and will end up being used alongside a weather station feeding into a logging PC.

I need a depth sensor, so that I can closely monitor rising water levels in a stream that runs through the garden, for all sorts of reasons, it causes problems, to the extent that we had €120K damage done in 2002 and came closer than I'm comfortable to a repeat performance after 100mm of rain in 24 hrs yesterday! Fortunately, the limited work that the local authority did after last time has helped, but there's not the comfort factor there that I need to have.

The stream is in a walled channel, the channel is about a metre deep, and usually, there's less than 10cm in the bottom of it, and it's fine. If we get abnormal rain, it can end up 25cm or more above it's channel. What I want to acheive is a sensor that gives me the level to the nearest 2.5 or 5 cm, and can then be read back into a unit indoors that will give me an alert if it gets above a (to be determined) specific level.

The level can rise for 2 reasons. One of course is just the total flow, but there can also be a scenario where debris, junk, discarded rubbish or general debris gets hung up on a grid at the entry to a closed pipe at the end of our site, and in that case, the level immediately downstream of the grid is lower than the level upstream.

Either case can result in problems, there's a barrier wall between us and the stream ( after the previous flood) but I need to know if the level has risen for either reason, so that I can take appropriate precautions, which might jsut be clean the grid, or at worst case, might be put barriers up around the house, and start moving vulnerable items out of harms way. Above a certain level, I also need to check that the pumps I have in place to clear water around the house are ready to go, and that the non return valves are not stuck open.

That's the requirement. I've thought of 2 possible ideas for the sensor, the first would be a pressure sensor that would measure the pressure of a trapped air column in a pipe once the water rises above a very low level, the other would be a series resistor chain with a stainless rail that runs parallel to it that would sense the varying resistance as the level rises. I think the second one might need to be AC powered to avoid it becoming contaminated by elecrolysis deposits, and both deposits and ac might make measuring the voltage level using a PIC or similar a bit more tricky. I reckon the sensor will need to be 1.5 mtrs long to allow me to make connections above the worst case scenario water level.

So, that's the challenge, anyone got any other suggestions that might help or give me a better solution? As for the way I get from the sensor to the PC, I'll worry about that in a while. there are several PIC's that will cover that without too much pain.

Thanks in anticipation

Steve
Irish Steve

Shore, if twas easy, we'd all be doin it
AllyCat
Posts: 1124
Joined: Sat 26 Feb 2011 1:58 pm
Weather Station: Fine Offset 1080/1 & 3080
Operating System: Windows XP SP3
Location: SE London

Re: Weather related sensor

Post by AllyCat »

Hi Steve,

Some immediate thoughts:

The resistance of water depends so much on its "purity" that I doubt if a single resistance method can be made to work. But you could make a "ladder" of resistance sensors with "rungs" every 2.5 - 5 cms and code the level from the number of "closed" contacts. Or maybe insulate the vertical rods and measure capacitance and/or loss angle (but I think water is rather a complex dielectric).

I believe that nearly all washing machines use pressure-pipe water level detectors (switches). Maybe you can find a few scrap machines/parts to see if they can be adapted.

Or an acoustic method such as the resonant frequency of an air column in a vertical pipe (with the water level closing the bottom) or ultrasonic "time of flight" using an electronic tape measure or a car reversing sensor (add-on kits can be found quite cheaply on ebay).

Cheers, Alan.
User avatar
beteljuice
Posts: 3292
Joined: Tue 09 Dec 2008 1:37 pm
Weather Station: None !
Operating System: W10 - Threadripper 16core, etc
Location: Dudley, West Midlands, UK

Re: Weather related sensor

Post by beteljuice »

The beteljuice really likes the idea of using a 'reversing sensor'.

They are now cheap and cheerful, and although they don't give the resolution you want, they will work over a realistic range for you.

This model has distance lights and meter, different range 'beep' alarms and differentiates between two sets of sensors (left and and right). So you could put one above the other to extend your range or have one at the inlet and one at the outlet, then the colour bars would indicate any great difference.

OK placement and bleeps from one sensor and not the other might take some thought, but for the price plus a 12v power supply and extending the sensor leads ??????

http://www.ebay.co.uk/itm/LED-CAR-PARKI ... 45fdf53d5d
Image
......................Imagine, what you will KNOW tomorrow !
User avatar
mermaidbeachweather
Posts: 71
Joined: Mon 05 Sep 2011 5:19 am
Weather Station: Vantage Vue
Operating System: Windows 7
Location: Mermaid Bch QLD Australia
Contact:

Re: Weather related sensor

Post by mermaidbeachweather »

beteljuice wrote:The beteljuice really likes the idea of using a 'reversing sensor'.

They are now cheap and cheerful, and although they don't give the resolution you want, they will work over a realistic range for you.

This model has distance lights and meter, different range 'beep' alarms and differentiates between two sets of sensors (left and and right). So you could put one above the other to extend your range or have one at the inlet and one at the outlet, then the colour bars would indicate any great difference.

OK placement and bleeps from one sensor and not the other might take some thought, but for the price plus a 12v power supply and extending the sensor leads ??????

http://www.ebay.co.uk/itm/LED-CAR-PARKI ... 45fdf53d5d
what about a simple float valve connected to a micro switch which sets off an alarm 8-)
apenwith
Posts: 418
Joined: Sun 26 Apr 2009 7:30 pm
Weather Station: w8681 BlakeLarsen Sun Recorder
Operating System: Windows 7 SP1
Location: West Cornwall UK
Contact:

Re: Weather related sensor

Post by apenwith »

Hi
Not a solution I'm afraid just a suggestion which might influence the design. It would be useful to know the rate of change of level and possibly the rate of flow.

But a suggestional also - a rheostat with a long arm attached to a float (though there would be a lot of 'bounce' in turbulent water) or a magnet attached to a float inside a plastic tube (open to the water at the bottom) with reed switches fixed to the outside all enclosed in another tube. Each reed switch could vary the resistance of the whole circuit much as the FO wind direction sensor works. By restricting the air outlet of the inner tube some degree of damping could be achieved.

Just a thought

Regards
Alan
Irish Steve
Posts: 15
Joined: Thu 09 Dec 2010 4:01 pm
Weather Station: WH1081
Operating System: XP SP3

Re: Weather related sensor

Post by Irish Steve »

apenwith wrote:Hi
Not a solution I'm afraid just a suggestion which might influence the design. It would be useful to know the rate of change of level and possibly the rate of flow.

But a suggestional also - a rheostat with a long arm attached to a float (though there would be a lot of 'bounce' in turbulent water) or a magnet attached to a float inside a plastic tube (open to the water at the bottom) with reed switches fixed to the outside all enclosed in another tube. Each reed switch could vary the resistance of the whole circuit much as the FO wind direction sensor works. By restricting the air outlet of the inner tube some degree of damping could be achieved.

Just a thought

Regards
Alan
Alan, that last idea's got possibilities. Under "normal" circumstances, there's less than 100mm in the bottom of the channel, and the rate of flow is probably 1 Mtr in about 5 to 10 seconds. Once it starts raining significantly, the rate of change is not massive, unless a large item blocks the grid in which case, it can come up to the top of the channel, within probably a minute or two, depending on the depth at the time and the flow rate, which at peak is close to 2 Mtrs/sec, so some force behind it also.

For that reason, the float on an arm isn't too easy, as I need to cover a movement range of up to 1500mm, the channel itself is 1000mm, anything above 1000mm is out of the channel as such, but there's a retaining wall set back from the edge of the channel that gives me another 700 mm before it comes over the wall, and if it does, we're in DEEP trouble, figuratively and literally, as the floor level of the house is only 75 mm above path level.

The concept of reed switches in a sealed tube with a magnet moving over the tube has possibilities, as that can be shielded from the turbulence and debris in the stream by an outer pipe. the flow rate is sufficiently high at peaks that anything moving in the stream, and we've had items as large as railway sleepers (yeah) and they have the potential to do serious damage to anything they hit in passing.

I've never used them, the other thought that triggers is hall effect sensors, as they also may be operated in a non contact manner, which is what I ideally want, and that then gives me another thought, use a floating ball in a pipe with an ultrasonic sensor on the top, that way, the ball should give a reliable return to the sensor, and an accurate level from it.

I will throw a thought about that last idea into one of the Robotics forums, they will have experience of using ultrasonics for range finding, and interfacing to micro boards and the like.

Thanks, this will be ongoing, I've been distracted for a few days with mundane things like changing the bearings on our washing machine, which was a much bigger and more time consuming job than I'd expected, and had to be done if we wanted to use the machine and sleep at the same time.

Thanks to everyone for the ideas, I will come back with the results when I have some,

Cheers

Steve
Irish Steve

Shore, if twas easy, we'd all be doin it
philcdav
Posts: 244
Joined: Tue 24 Jun 2008 9:43 pm
Weather Station: MyDEL WX2008 Mk2 Fine Offset
Operating System: XP and W7
Location: Maghull, nr Liverpool, UK
Contact:

Re: Weather related sensor

Post by philcdav »

Hi guys.

KISS - keep it simple.

How about a plastic tube, mounted vertically, containing a 'ping pong' ball.

You could use the 'reversing sensor' to measure the depth OR

fit a resistance rod to the ball and calibrate.

This could be a piece of light plastic rod, gluedd to ball, with a wire wound resistance (NOT copper) sliding thru a wheatstone bridge system to 'measue' resistance

The tube would have to be solidly mounted (concrete?) and open top and bottom to allow water in and air out.

I would 'set' a minimum level for the ball to stop the tube bottom silting up, a nail thru tube woould do.



Cheap, easy to maintain
Phil - G0DOR
daedalusminos
Posts: 15
Joined: Sun 23 Jan 2011 10:24 pm
Weather Station: TBD
Operating System: Windows XP SP2
Location: Oby, Norfolk

Re: Weather related sensor

Post by daedalusminos »

This can be done with an LED, photo transistor and clear acrylic rod. Essentially it relies on the difference in refractive index between air and water and the total internal reflection down the acrylic tube (in air).
An LED and phototransistor are mounted at the top of the tube, the LED is pulsed and the back scatter is measured by the phototransistor.
Down the rod at 5cm (or 2.5cm) intervals are cut and polished 'ports' which allow light to escape when under water but retain total internal reflection in air. As the water level rises you'll see a drop in the back scatter and photo transistor O/P as each port is covered.
Drawbacks:cutting polishing the ports is time consuming. As it's an absolute measurement, you need to periodically recalibrate 'dry' or allow for LED degradation over time (split off some light aand use as reference) and aslo allow for rod surface contamination or periodically clean.

There are other variants e.g. LEDs placed down the rod at the measuring intervals shining into the tube.

Jonathan
sanramonrover
Posts: 96
Joined: Mon 02 Aug 2010 6:51 pm
Weather Station: 265NC
Operating System: dreaded Win 7
Location: to be San Ramón, Costa Rica

Re: Weather related sensor

Post by sanramonrover »

Maybe one more non electronic solution?

To measure the depth in my 20,000 liter water tank in Australia, I made the following:
a weighted float,
very fine nylon line (fishing line),
old spring driven alarm clock,
glass jar large enough for the clock,
plastic (PVC) tube large enough for the weighted float, long enough for the water depth.

Remove the balance wheel and perhaps the next shaft and gear from the clock,
Attach the weighted float to the nylon line,
Attach the other end to one of the shafts of the clock (which one depends on depth being measured)
Wind the clock up so it is just strong enough to take up the slack in the nylon line as the weighted float rises, but weak enough to allow the weighted float to drop freely.

Mount the clock in an inverted glass jar on top of a tube submersed in the water,
punch a small hole in the lid for the nylon line to exit unhindered (after careful calculation or measuring),
re-attach the weighted float.

Calibrate a blank dial, using either the hour or minute hand (depending on water depth) to indicate MAX and MIN, and any intermediate marks desired.

The system worked for many years for me. Readable from a considerable distance.
Gina
Posts: 1885
Joined: Sat 21 Feb 2009 12:41 pm
Weather Station: Nothing working ATM - making one
Operating System: OS X, Linux Mint, Win7 & XP
Location: Devon UK

Re: Weather related sensor

Post by Gina »

I was going to suggest Alan's reed switch and floating magnet method :) He beat me to it :lol:

We have a related issue (not an actual problem). We have a duck pond which fills from a little stream and I'd like a level sensor for that. It's the opposite case - during dry weather the stream level drops too low to fill the pond and it dries up. When that happens we refill the pond manually. So we want to know when the pond water level gets too low.
Gina

Sorry, no banner - weather station out of action. Hoping to be up and running with a new home-made one soon.
gemini06720
Posts: 1700
Joined: Mon 10 Aug 2009 10:16 pm
Weather Station: No weather station
Operating System: No operating system
Location: World...

Re: Weather related sensor

Post by gemini06720 »

Irish Steve wrote:...I need a depth sensor, so that I can closely monitor rising water levels in a stream that runs through the garden...
Steve, here is another idea. Although designed to calculate the snow depth, this this 1 wire sensor project could possibly/probably be adapted to calculate/measure liquid water levels... ;)
AllyCat
Posts: 1124
Joined: Sat 26 Feb 2011 1:58 pm
Weather Station: Fine Offset 1080/1 & 3080
Operating System: Windows XP SP3
Location: SE London

Re: Weather related sensor

Post by AllyCat »

Hi,

Having been designing electronic "gadgets" for almost 50 years, I'm all for a simple "mechanical" solution. ;)

A weighted float on a cord turning a pulley (e.g. a bicycle wheel) and potentiometer seems the simplest. To make the pulley a more manageable diameter could use a simple geartrain, or a "ten turn", 1/4 inch shafted pot might still be obtainable. Certainly pre-sets with the wiper on a "lead screw" (typically 40 turns) are available, but might not be sufficiently rugged.
apenwith wrote:Each reed switch could vary the resistance of the whole circuit much as the FO wind direction sensor works.
Yes, I also like the idea of reed switches. But beware that the FO concept requires a magnetic field large enough that there is an overlap where two adjacent reeds operate together (and thus two resistances are connected in parallel). Otherwise there will be positions where NO reed is operated and thus no measurement. This could of course be solved by "remembering" (storing) the last operated reed.

But, IMHO the FO technique is a good starting point. For example a "ladder" of reeds with 10 cm spacing and a magnetic field 15 cms high could give 5 cms resolution. The magnetic field could be created with a few button-sized "super" magnets (typically from ebay in packs of 20). However, I think it would be difficult/impossible to design a chain of "weighted resitors" for much more than 8 reeds, so I would use two ladders with interleaved "rungs" so that at most only one reed and resistor value is active in each. Or you might even increase the resolution and/or reduce the number of reeds by using two magnetic fields separated by for example 2.25 rung-spacings.

Cheers, Alan.
Gina
Posts: 1885
Joined: Sat 21 Feb 2009 12:41 pm
Weather Station: Nothing working ATM - making one
Operating System: OS X, Linux Mint, Win7 & XP
Location: Devon UK

Re: Weather related sensor

Post by Gina »

Trim-pots are not likely to last long as they're not designed for continuous change - like volume controls on cheap radios wear out around your normal listening level as you turn them up and down.
Gina

Sorry, no banner - weather station out of action. Hoping to be up and running with a new home-made one soon.
User avatar
yv1hx
Posts: 223
Joined: Mon 05 Apr 2010 10:40 pm
Weather Station: No station yet ...
Operating System: Win XP Professional
Location: Some point in the Earth

Re: Weather related sensor

Post by yv1hx »

As a Telecoms technologist, I'm always tempted by the electronic, automatic, then no-human solutions.

Maybe some of this ultrasonics sensors:

http://www.olimex.cl/pdf/Sensors/Proxim ... asheet.pdf

http://www.parallax.com/tabid/768/Produ ... fault.aspx

http://www.jameco.com/Jameco/Products/ProdDS/136654.pdf

Controlled by a pic like this:

http://www.sparkfun.com/products/9195

Combined with some short range compatible data radio, like this:

http://ftp1.digi.com/support/documentat ... 0976_G.pdf

Can be a solution.

Just my 3 cents :roll:
Marco
BigDave
Posts: 31
Joined: Fri 11 Dec 2009 6:46 pm
Weather Station: Fine Offset
Operating System: W2K Server
Location: London

Re: Weather related sensor

Post by BigDave »

No posts on this thread for some time, so a solution has probably already been found, but just in case here's my tuppence worth...

Personally, I think I'd be looking to use this type of sensor or something like it:

http://www.grisk.com/specialty/2600water_sensor.html

It's an encapsulated sensor which contains a reed relay, and a small circuit to determine whether there's a conductive liquid shorting out two contacts or not. It's a normally closed contact (going open when liquid is detected).

I think I'd be looking to mount an array of these at say, 10 cm intervals onto a backplane of some kind, probably aluminium sheet, and then enclose the whole thing in a sturdy drainpipe (open top and bottom of course!), bolted to the sidewalls of the channel.

As the water level rises, the sensors would go open circuit as the water reached the next level, and the top level sensor could be used to trigger a "flooding imminent" alert.

Being normally closed, I could then be sure that if the wire got cut, I would know about it. In fact, I'd probably use a dual end-of-line arrangement such that the circuit normally reads resistance X, when water is detected, Resistance X+Y, and a short or open circuit represents a cut or shorted cable.

In terms of what could be used to monitor such an array of sensors, I guess that would depend upon exactly what information you wanted from it, or perhaps put more accurately, what condition(s) constitute an alert.
Post Reply