P.A.B.S.T. (PICAXE'd Automatic Bilge/Sump Thing)

jax · August 16, 2012, 12:11pm

aka The Blue Ribbon

Intro:

I have a portable air conditioning unit that sits in the corner of my lab area. Just like all ACs it produces quantities of condensed water which drains into a one galllon jug. When the jug gets full I empty it. The problem is when there are hot and humid days the AC will overfill the jug and I'll come home to a nice squishy mess. Also if I don't empty it at night before going to bed I sometimes wake up to the sounds of an unwanted zen waterfall. No good.

The goal was to make a bilge pump to move the water out of the jug and into the sink behind the AC. The AC purges its condensed water when it's internal tank gets full so I needed intermittent monitoring of the water level in the jug for efficient pump operation.

Materials:

(for final build)

PICAXE 08M2 x 1

DIP8 socket x 1

12V wall wart x 1 (salvaged)

24V diaphram pump x 1 (salvaged)

10K serin pulldown x 1

2N2222A transistor for pump switch x 1

2N2222A transistor for signal ouput x 1

1K base resistors x 2

1M pot for tuning signal output x 1

2.1mm power jack x 1

L7805ACP voltage regulator x 1

0.1uF cap x 2

0.33uF cap x 1

ribbon cable (only three wires necessary) x 1 (salvaged)

Method:

The PICAXE turns on the NPN transistor which sends out a signal through the green wire after running through the 1M potentiometer.

When the water level is up to the orange wire the signal is detected by the PICAXE. It turns on the pump while occasionally monitoring the pin connected to the yellow wire. When the signal is no longer detected by the yellow wire it turns the pump and signal transistor off and goes into power saving mode for thirty minutes. Then it wakes up and repeats the process.

Pic of the sensor ends and intake tube in the collection jug.

Conclusion:

After doing some testing with the water level detection I think I'm going to go a different route. Instead of running the signal through the potentiometer and then picking it up with the uC pins I think I'm going to put the pins on transistors with their bases connected to the orange and yellow level detection wires and skip the signal wire pot.

I'm also having problems with the low level/yellow wire giving false highs from residual moisture. I might try to make another sensor cable/assembly, I don't know.

So far the breadboarded circuit works nicely. No more waking up in the middle of the night wondering if I emptied the water tank or not. That being said I'm planning on making a real PCB and putting it all in an enclosure to tidy everything up. The v0.9 board is already designed.

damo3 · August 17, 2012, 4:26pm

water level sensor

I’m sure there are dozens of ways to detect the water level in a container.

The simplest would be a float switch, othe ideas are using a floating magnet and a reed switch outside the container.

A ping sensor may also give the distance from the top of the container to the water surface, and hence the water level. (not sure if a ping pulse will be reflected by water ??)

If you want the no-fuss approach, use a 12volt bilge pump with a float switch from a boat shop. But I agree it’s more fun to build your own !!

jax · August 17, 2012, 6:03pm

Building is fun! Buying stuff makes me poor and not as fun!

Makers FTW!

I considered float switches but I didn’t have anything readily available in the bits box. I’ve been getting a lot of good ideas for the sensor. The current setup works ok but I’m not against improving The Blue Ribbon. Love hearing other ideas, thanks.

merser · August 17, 2012, 8:28pm

Just an idea to expand on this

I don’t know if your a light sleeper Jax but you could pump this water through some kind of decorative fountain before sending it to the fishes.
If your feeling arty that is. I’m just trying to think though whether the idea makes sense. Maybe it requires a second pump. As the amount of water is too low you would need to recirculate and have an overflow for the excess.