Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Water level sensor materials

Status
Not open for further replies.

R_C

Member
Several years ago I made a simple water level sensor for the fresh water tank on my boat. I used a series of stainless steel machine screws protruding from a PVC tube immersed in the tank to light up to six LEDs to approximate the level of the water. A screw under water activates a transistor to light the LED.

It worked well for some time but slowly the sealant I used around the machine screws failed and the PVC tube began to leak. Next, I filled the PVC tube with polyester resin. Although the heat of the curing resin bent the tube it worked for some months but eventually the tube still developed leaks.

I'm looking for ideas for alternate materials or construction. The rigid PVC tube and the polycarbonate lid I created hold up well to the sloshing in the tank. I just need a better way to keep the wires waterproof so only the conducting sensors see the water. Thanks in advance for any ideas.
 
You might consider using a rig not unlike an auto brake (or coolant, etc) fluid level detector.

These totally isolate the level sensor(s) and are as leak proof as you can get.

A PVC (Shedule 20) tube with hall-effect transistor(s) on the inside and a donut like magnetic float ring on the outside (that can float up or down as the water level changes) would suit your needs.

Your current PCB and components could still be used. It would just require some biasing resistors to accomodate the on/off switching of the hall-effects.

You could probably salvage an engine coolant magnetic float ring from a junk yard.
 
Last edited:
How about stripping back some 2.5mm twin and earth and using the insulated wire down your tube instead of the screws.
 
Since this is our potable water supply (we still filter it) I am concerned about what materials are in the water. I haven't used hall-effect transistors but I'll read up. Are there any insulated wires that are waterproof? If a length of wire with a bit of insulation removed at the end served as the sensor how would the conductor stand up over time? It will be in and out of the water.
 
What you can do is get a piece of PVC and something that can act as a float, like strofoam. Put that inside the tube with say two small magnets to balance it.

Noe you can detect the position from outside with a reed switch or hall effect sensor, Your gauge probably has to be a bit smarter because it needs to keep track of where things are. It would get a blip when it passed a point, but the level would not be continuous as you have now.

The float can't rotate, so a swuare rod may have to be used.

Another idea that comes to mind is stainless rivets, but not sure how that would work out.

To avod corrosion, sensors like yours should be exicted by an AC voltage rather than DC.
 
Thanks for all the ideas. Sensor corrosion from the DC current isn't a problem since we only turn the gauge on for a few seconds to check the water level. I'm inclined to avoid any moving parts such as floats since rough seas could lead to failure. I think my simplest solution would be to use the same design, perhaps with the suggested stainless rivets instead of machine screws, and fill the PVC tube with some material that would prevent any water penetration. Maybe silicon-based electronic potting material? I'll have to see what will stick to PVC or some other plastic tubing.
 
I had to buy a box of 500 of 1/8 countersunk aluminum rivets becuse I needed 3. Countersunk would work better, I think. Use a backing washer too.

How about using an O-ring on the backside of the rivet sandwhiched between a backing washer?

It would be fun putting it together, but you can always tap a hole behind the rivet and insert a plug with teflon tape or PVC glue. Thenuse a stick with a flat bottomes hole (pilot point bit) to place the washer and backing washer there.

Here are some stainless ones on ebay **broken link removed**

Not sure whay you care that it gets wet on the. How about a large stainless screw or all-thread thru the pipe and then use ring terminal on the outside of the pipe. Use the heat shrink with the sticky seal on the inside to cover the crimp and adhere to the insulation.
 
Last edited:
If the bottom of the PVC tube is not plugged, I found that filling the tank sends water right up the tube through the RJ-45 connector and out the top of the tank even though the tank is vented. So, my goal is to keep the inside contents of the PVC tube completely dry and only the sensors on the outside of the tube come in contact with water. I originally countersunk the holes and put a rubber washer behind a washer and nut on the outside. But with a threaded machine screw, that does not work and water still penetrated the tube. Pop-riveting through the PVC to the ring terminal on the inside should be simpler than the machine screw but I would still like to fill the PVC tube somehow afterwards to guarantee no water can penetrate. Water on the inside of the tube leads to inaccurate readings on the gauge.
 
Stirling marine tank guage senders go in from the top of the tank, a long insulated rod with stainless electrodes along it, couldnt you use some 1/4 stainless rod from the top rather than the side?
 
If the bottom of the PVC tube is not plugged, I found that filling the tank sends water right up the tube through the RJ-45 connector and out the top of the tank even though the tank is vented.

If you drilled a vent hole in the tube near the top, it wouldn't build pressure when filling.
With light fixtures etc, I have found that it's better to let water out than to try to keep it from getting in.
 
If you drilled a vent hole in the tube near the top, it wouldn't build pressure when filling.

Aha. So, maybe the solution is to not cap the bottom of the PVC tube, add the vent holes near the top, so the tube is simply a holder for the sensor wires. The sensor wires could be fastened to the interior or exterior of the tube and I just waterproof the RJ-45 breakout board and connector at the top. With the vents in the sides of the tube near the top the tube is vented into the tank not out the top and the water level inside and outside the tube are the same so the sensor wires can be inside the tube. That should do the job.
 
Bare with me as I try to explain this. It may take a few responses.

You tried a PVC tube with screws at discreet locations placed in from the side. What about putting different length screws (threaded rod, etc.) in from the top? It only matters when the water contacts the end of the probe not how much of the probe is covered. That will also make them adjustable and you will be able to use the present circuit.

Please excuse the crude picture. Hopefully the concept is clear.

probelayout.jpg
 
You tried a PVC tube with screws at discreet locations placed in from the side. What about putting different length screws (threaded rod, etc.) in from the top? It only matters when the water contacts the end of the probe not how much of the probe is covered. That will also make them adjustable and you will be able to use the present circuit.

Both solutions should be equally effective. I just find the screws on the side easier to build. As you can see from the attached images it's pretty simple. Attempting to waterproof the entire tube may have been my real problem. Waterproofing just the RJ-45 connector in the PVC coupling at the top could be the solution.

Sensor1.jpg

Sensor2.jpg
 
Thats how commercial ones look.
Seeing your schematic you will have dc on those electrodes, fresh water isnt too bad however you probably will get some electrolosys going on with the electrodes, if you had a push to make switch to make a tank reading that would sort out the issue as the system would only operate for a couple of seconds.
 
There are waterproof RJ45 connectors.

There is also a BIG difference in trying to seal a thread vs. trying to seal a cylinder.

In your design, why bother sealing the inside.

Adhesive lined heat shrink should be all that you need on the inside. https://www.electro-tech-online.com/custompdfs/2013/08/76650-0026.pdf Crimp a connector and seal.

Wanna be fancy with the wire insulation, use teflon insulated wire.
 
Last edited:
Thats how commercial ones look.
Seeing your schematic you will have dc on those electrodes, fresh water isnt too bad however you probably will get some electrolosys going on with the electrodes, if you had a push to make switch to make a tank reading that would sort out the issue as the system would only operate for a couple of seconds.

We do only flip the toggle switch for a few seconds to check the water level but I like your suggestion. We could inadvertently leave the gauge on or something could bump the toggle switch so a momentary switch is a better solution. Thanks.
 
I appreciate all the suggestions. I now see the solution. I'll just make a new PVC tube sensor leaving off the bottom end cap so water rises inside the tube. I'll add new vent holes in the side of the tube near the top and waterproof the RJ45 connector and breakout board. The wiring will be identical to what I already have. Now the ring terminals on the inside become the sensors and I just need to secure them to the PVC. I could use pop rivets but I think I'll stick with machine screws for easy disassembling. I'll add adhesive heat shrink at the ring terminal connections to protect the wires and replace the toggle with a momentary switch. That should do it.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top