We’ve got some friends who have two sump pumps. One is a backup and sounds an alarm when it is switched on. But this only works as long as they’re home to hear it. [Felix Rusu] came up with a solution what will text him if the sump pump fails. This way he can head home, or call someone to check in on the problem if he’s away.
We saw a pretty complicated monitoring system back in January. This one uses a single ultrasonic rangefinder which we think is much simpler. It’s accurate to about 1cm and is simple to use — it’s very popular with the hobby electronics crowd which helps with price and availability of sample code. We hem and haw about the use of a Raspberry Pi board with the project. On the one hand it’s a cheap way to get the sensor on the network and provides the infrastructure you need to send any number of alerts. On the other hand, it’s a lot of power for this particular application. But we figure it can be extended to monitor other utilities in [Felix’s] home, like a sensor to alert him of a leaking water heater. And we think everyone can argue that a monitor like this is well worth the time and effort he spent to develop it.
This seems like it’s going to fail silently while sitting around for 11 months not being needed. Which is the same problem it’s supposed to solve for the sump pump.
that’s easily fixed by a daily text message of the water height.
What if the cell network fails? The easy solution is just to sit by the pump basin with a bucket at the ready, just in case.
if the cell network fails, you won’t get your daily message and will know something is up and check on it. That’s the whole point of a daily message, did you miss that?
If the network fails at the same time as the sump has issues, you probably have other issues (like a big storm or something).
Exactly.
I made a similar system to monitor the level in a silo.
Unfortunately, the ultrasonic range finder failed after about 6 months.
Not sure those transducers are made to last when in continuous use.
They are meant to be mounted on toys and small robots and certainly aren’t rated for working while exposed to moisture or in hazardous environments like every other piece of cheap electronics like the Arduino, RasPi and a billion other boards/sensors.
The point here is the idea being given, which almost always shows an overkill implementation of the most widely known hardware and not of the best one. It’s up to the reader to implement it using the right hardware.
No, it seems like it would fail quickly when the water gets high fast enough to submerge the sensor and short out the power supply, if you don’t get to it in time.
Easy fix. Have the moteino send out a ping to the Raspberry Pi every so often. If the Raspberry Pi doesn’t receive a ping within a specific time period, send out a notice. Basically a watcher for the watcher.
My concern is that this will short the 5v usb power supply once water reaches the sensor. It’s not the sensor I am worried about, it’s possible electrocution or blown fuse situation.
Dude have you ever worked with electricity? 5v is nothing. you aren’t likely to feel anything from 5v, unless maybe you stick it on your tongue.
it’s not likely to blow a breaker because that’d be a ridiculous amount of current on 5v (360amps for a 120v 15A breaker). It might blow a fuse in the power supply, but that’s the whole point of fuses!
And you think a generic 5v usb has fuses or proper mains isolation? That it won’t catch on fire?
cde: Seeing as how this is powered by a USB charger the quality of the charger you choose will have everything to do with how safe it will be. In the pictures I can clearly see he’s using a Motorola branded charger, I’d trust it to not catch fire if it shorted out.
cde: Thankfully if it catches fire it’ll likely be submerged. Problem solved.
How long will the range finder last, exposed to a humid environment like that sump. Look at all the crud on the cover for the sump, won’t that crud foul the sensor?
Can’t understand the need for the Moteino (arduino clone?). I understand that the ultrasonic rangefinder needs 5V but couldn’t this have been done with a level shifter? Can the raspPi measure pulse widths with enough accuracy/precision?
He sells them. The blog is about giving people reasons to buy one.
Yeah you can buy them in my store. Or if you have a jeenode or other similar hardware you can use that. You can also buy all the parts and make them yourself, but you’ll probably spend quite a bit more than $19. Or you can buy other clones that have the same specs, RF capability (if you really prefer spending more).
Because they make them, and it’s wireless. It has a RFM12B transceiver on it. The RPI could do it, but you’d have to have the RPi downstairs, with a wifi device, or run ethernet cable. Why throw a computer at what a microcontroller can do?
Exactly… what if you have 2 sump pumps? Or 5? Buy 5 raspberry Pis? I don’t think so… it makes a lot more sense to use a low cost minimal configuration transmitter that simply reports the data to a central node.
You’re spot on, the Ultrasonic sensor can be attached directly to the Pi, for an example see: http://www.raspberrypi-spy.co.uk/2012/12/ultrasonic-distance-measurement-using-python-part-1/
Yep, you sure can. And BTW you can attach pretty much any sensor to the Pi. You can have as many RaspberryPi’s as you have sensors. But that won’t really help terribly much with scalability I’m afraid :)
On the subject of waterproof range sensors is someone up for a little serial snooping?
I bought a 4 sensor reversing kit from a China eBay seller a few weeks back in the hope of getting the sensors working on the cheap non-waterproof ultrasonic distance sensors but didn’t work as the waterproof ones require more power and are less sensitive, however the car reversing kit outputs a TTL serial signal from the main box to the LED display box, and I can’t determine the baudrate and specific data that’s being sent.
It’s one of these: http://www.ebay.com/itm/290780812939
That page mentions ‘KWP-2000’ which isn’t plain serial data. http://en.wikipedia.org/wiki/Keyword_Protocol_2000 Of course there’s still the matter of if this is used from the sensors or just the main box for possibly interfacing to the car. Though you’ve piqued my interest as I didn’t know those modules could be had so cheaply and our basement is flooded at the moment!
How about a simple float sensor instead of the range finder? Much more durable and is cheaper to boot. If need be, install a couple at different heights for a more fine-grained approach.
I completely agree, in fact that was my first thought.
Or vertical float with an optical encoder.
It’s a cheap range finder. Who cares if it shorts out. It’s also 5v. it probably won’t set the house on fire…just make sure it’s plugged into a GFCI.
Now, if the water level reaches the sensor then there are bigger things to worry about because that’s the same level as the basement floor so there’s going to be flooding happening regardless.
The range finder is actually one of the best methods for this type of application because you don’t want a float or anything else touching the mucky water. It’ll get damaged. The range sensor will stay out of the muck and last longer.
Now that said….I wouldn’t mount it to the bottom of the cover. I would make sure it’s higher and a little more protected than that. (You know, add some engineering know how to the project)
But overall I like the application.
“The range finder is actually one of the best methods for this type of application because you don’t want a float or anything else touching the mucky water.”
Yes you do because that is how commercial systems work. all good installs have 3 floats in there 1 for the pump, 1 for the 1st alarm and backup pump activation and the last for the doom is approaching alarm.
Almost all of them have this. from homes to industrial, they use floats.
Float switches that are made for sump pumps are fully enclosed, so I’m not sure how you figure it will get damaged. I’d be interested learn how it could be damaged if you have some insight. :)
Floats do fail; the ones that switch the AC line to the motor directly anyway – it seems the metal ball or contacts inside corrode, possibly from water entering where the cable enters the plastic ball, or from a weak sonic weld of the plastic. I’ve had 6 floats fail in my lifetime.
I happen to like the Basement watchdog dual float controller, although the electronics do add more complexity and therefore more points of failure, the fact that there are two floats wired in parallel (just in-case one of the reed switches that are activated by the float rings) (or if one of the float rings gets caught up on some floating derbis)
I have 4 of these at staggered heights, 2 connected to pumps, 2 connected to a controller that send me email alerts.
I’m still considering this ultrasonic sensor version, maybe if the sensor was potted in epoxy it would hold up against getting splashed?
A friend used something like this 20 years ago on a ‘mud pit’ on a drill ship that drilled oil wells off shore. But a safety engineer had him remove it because there was often natural gas mixed in with the mud that is returned up the drill stem. … Still it worked well. … Great idea, just a different industry.
Man overboard! Throw him a GPS tag? NO a floatation device!
Float with magnet sliding on PVC tube with reed or hall switches inside and dry.
Exactly. This contraption looks like it will last about a week in this environment.
My idea given a task like this would have been a length of cable with the ends held apart a few millimeters by a lump of hot glue, a darlington transistor and a TTL Schmitt trigger to hook up to an arbitrary button on an old cellphone. Program the phone’s “single button dial” (even Nokias from the last millennium have that) to your own number and store the alerting phone’s number in your phone book as “Check Pump”. Costs a fraction, uses an even smaller fraction of the power and can be completely sealed with a glue gun to last many times as long.
Purpose built, sealed float for the win!
http://www.securitysystemspro.com/Sump-pump-alarm-float-switch_p_46.html
a simple floater combined with a venturi ;pipe will drain the water without failure, without electricity, without monitoring.
More complex : I built a monitoring auto drain device made of :
electric reversible styrofoam floater
x10 DS10A window wireless sensor hooked on it
x10 server somewhere with internet access
a weatherproof IR cam (DVR and android access)
Bilge 12v pump with integrated floater
a 12V battery, solar charged
all is functionnal (except x10 warnings and cam) if power is down
Ok guys, this is just 1 way of doing it, not THE way of doing it. You can have a million variations and alternatives to this. For most people it’s easier to find problems and whine. In the meantime I’m enjoying my system which *DOES WORK* and solves a problem, and adds that extra level of protection that I didn’t have, all for about $25 in total. You can add 10 more layers of protection and detection on top. Whoever thinks this simple system sucks is welcome to come up with a much better solution that works for them. The $5 ultrasonic is disposable, I can get another when/if it fails, it also gives me a water level, which I can graph – I like graphs.
How is this for a sophisticated moisture sensor?
I once saw a water sensor in a vintage mechanics’ magazine made from a lever switch mounted near the floor that was held closed with an Aspirin tablet stuck between the lever and the floor. When the Aspirin got wet it crumbled/dissolved and operated the switch…..so the hypothesis went…
Seems a little complicated and untested, what happens if the power goes out? Just installed a system with battery backup so I get messages even if the power is out. No network or phone was needed. Installed in about minutes. You can check it out at http://www.pumpalarm.com. Was a better fit for my application than a homemade device.
Excellent way to share. These devices will continue to keep getting better. I found this PitBoss unit with the battery backup that ended up working best for me. I put one sensor on the sump pump and one on the sewer drain coming in the basement. It saved me when we got 4 inches of rain backing up the system on a Sunday night. It was an easy setup and works better than the old buzzer / light box I had. I got the PitBoss here: http://store.aquatechnologygroup.com/products/pitboss/
You could have purchased direct from the manufacturer by visiting http://www.pumpalarm.com.
Whoa! That’s for sure an innovation! Perfect sensor. I know more techy stuff like this one are gonna come in the near future. Main concern: what if my network fails at the same time my system fails?
I’m interested in trying out this system – http://www.absoluteautomation.com/sumpspot-wifi-sump-pump-alarm-for-texting/ – it sounds like it supervises itself but it uses a basic float rather than a range finder. Plus side is there’s no cell plan, just works on the WiFi, I don’t need another monthly bill haha.
These seem very expensive. I was thinking of using my backup sump pump alarm to trigger the signal. When the backup sump pump float detects a higher water level (caused by main pump failure) it (1) takes over pumping activities and (2) sounds an audible alarm. Surely the current to the alarm can be used to trigger a message? Comments.
Another variegation can be found here
http://m.instructables.com/id/Sump-pump-alarm/
Wifi alarm won’t work during a power outage but cellular will.