Homestead water level sensor

Alternative energy sources
Forum rules
The usual. As I have two large solar PV systems here, and my lab assistant just put one in, and others are interested in things like this, here's where that stuff goes. This is mostly for things that work now, not "gee someday a fusor will do this" -- we know that, but it's not someday yet.
The hope is to save anyone embarking on this sort of thing a lot of wasted time and money, as at least I have been off the grid since 1980 and have had a lot of practice (and made mistakes you won't have to).

Homestead water level sensor

Postby Doug Coulter » Wed Dec 24, 2014 11:51 pm

Hopefully, this one will be reliable - I have a couple uses for it, and over the 3+ decades of homesteading, everything else I've tried has failed fairly young. I admit, most of that was for a hot water solar heated system, which just makes things worse, and this is for rain collection/filter tank, and another for the actual basement cistern, so it should be a little easier operating conditions.
Ultrasonic? You gotta be kidding. Sensors in steam...near water at all... forget it. Very short life.
Optical? Condensation and compounds from sometimes hard water - fail.
Thermal? Corrosion of even the supposedly waterproof stuff, complexity (compare to ambient?) and general failure over time.
Floats? Jam, magnet mysteriously gets too weak to work the reed switch and on and on - no matter what I tried, it failed. And these were the "good stuff" used industrially (hopefully not in nuke plants). All sealed and so forth.

So, I got inspired reading an old (1978 or so) National semiconductor Linear Apps book on this topic - they used to make a chip for this, though it was only single channel. The idea is to use the fact that water, a little hard/polluted or not - is a very good capacitor dielectric, around 80 compared to air ~= 1. No DC is applied to or across the water at all, which means if you use good stuff for the electrodes, no metal gets in your water, and nothing electrolyzes - you measure the AC impedance between a drive electrode, and your various (in my case 4) sensing electrodes. I'll be making those out of titanium wire (CP grade, no alloy) since that's about the best there is for this. SS might be OK too, but...I have the Ti wire in various sizes already and it's a lot easier to work with, bend, needs a smaller hole in the tank, and so on.

The circuit I designed (well, kinda a design, more of a copy) looks like this:
WaterSensor.gif
Just a mostly square wave oscillator and a bunch of volt doublers.

Totally old-skool. I even used a 555 for the oscillator, as I had a few new-old-stock laying around, and they seem like being available for another decade or more when I might need a spare. Also, I can run a fairly high voltage on these, off my 13.5v gel cell UPS system associated with my water pumping stuff - I use a 12v spot sprayer pump to pressurize plumbing. Other than being noisy, it's ideal (and I may add a reservoir designed to cut the noise). When you are dependent on rain, or going and hauling water, you want to know when the pump's going - you don't waste one drop on leaky plumbing/faucets and toilet ball-cocks. A drip can fill (or in this case, empty) an ocean. So the noisy pump is kind of useful for knowing if there's an issue, for now. I'm using all qest/pex these days - it can freeze and not break...nice. And it never, ever leaks, along with Kohler fixtures in the sinks etc. For the loo, I get whatever the hardware store guy says is best, and wind up putting in a new one about every two years. Life!

I did add some power supply bypassing to this circuit that isn't shown. .1 uF and 220uF at 25v.

The plan is to hook up a drive electrode, which will be some feet of Ti (fat) wire going to the bottom of the cistern, then horizontally most of the tank length. Then there will be 4 level sensor wires driving those volt doublers (for ac coupling...) at various critical tank levels for that end of things (I will be building a similar system for the rain barrel, with additional turbidity sensor and valves). I need to know - hey, it's full, turn off the intake valve. Hey, it's got enough room for one full rain barrel (the only time you want that water at all, since it has overflowed and is now free of floating junk), water low - conserve, and water almost out - boy, you'd better be getting a backup soon.

This will be wired to an Arudino Uno (hey, they're cheeap) along with some DHT22 temperature/humidity sensors for the basement, the bedroom, the kitchen (opposite ends of the building) and outdoors, just to have something else cool to log from it.
Logging will initially be done with a Raspberry Pi, but since it draws almost the same power as an intel NUC, I may go there when I have the spare change for it - in either case, the display draws more than the computer does - progress!
I used 1n457 diodes. They are what was laying around and are super low leakage. I looked at germanium, as less forward drop might be better, but it looks like the odd 1n34a's I have are too leaky. Schottky might be fine or better. At 100k load resistance, I doubt forward drop is a really big deal anyway...

The built board looks like this.
SensorBoard.JPG
Actual hardware


It bench tests, less than 1/4" of the drive and sense electrodes make this put out cmos logic levels in a test basin on the bench - as in 5+ volts output. In the real application, since I want the other arduino pins for the DHT22's and driving valve solenoids and such - I'm using the a/d inputs for this, which will allow tuning the software vs having to fiddle too much with changing the actual wire lengths and positioning. There will of course be a lot more on this project...homesteading ain't for old guys and I'm getting there. It's like a race with time to see how much I can automate before I get too old to do things by hand.

You might also notice I finally got wise and just buy top quality parts in quantity, at least for the standard values. You save money, hassle, time, and after all, in big lots, they're cheap - even the price of 1% mil spec resistors is pretty low.
Posting as just me, not as the forum owner. Everything I say is "in my opinion" and YMMV -- which should go for everyone without saying.
User avatar
Doug Coulter
 
Posts: 3515
Joined: Wed Jul 14, 2010 7:05 pm
Location: Floyd county, VA, USA

Return to Alt energy

Who is online

Users browsing this forum: No registered users and 7 guests

cron