Let's Make Robots!

Temperature sensor chain

As usual this is a project for my brother, but this time it was actually me that came up with the idea. My plan was to have an array of temperature sensors hanging below a buoy measuring temperature every meter down to 10 meters.
But this was nothing new my brother informed me and what I built is similar to things like these.
(but for a fraction of the cost)


I have worked with the DS18B20 temperature sensor before and it is an 1-wire device that has a laser cut unique address so you can have as many as you like on a single 1-wire bus. My first thought was having a cable with 3 wires, for 5V, GND and data. After goggling a bit it turns out that a regular CAT5 cable is the way to go.

But how to waterproof the sensors?
My first plan was to solder the sensors directly to the wire. Luckily I decided to make a breakout board to make the mounting of the sensors easy.  As the CAT5 cable has 8 wires 4 would be used for the sensors and the remaining 4 would act as tension relief.
My first thought was just to solder the sensors in and paint some epoxy over it to waterproof the sensor. But instead I went for a silicone mold and encapsulate the whole sensor in Epoxy.

My breakout board

First making the master for the mold

 I used FIMO modeling clay and baked to make the master

Then a container to hold the silicone

Then the silicone

The silicone worked great with the Epoxy as it needs now release agent and pops right out

In the end I had my array of sensors

The Arduino data logger is mounted inside a waterproof box. It has my data logger shield on it so this rig should be able to take a measurement of the complete array every 10 minutes for about a month on a regular 9V battery.

PS. I have sent a mail of to Sparkfun to see if they are interested in selling the shields. It might be a long shot, but why not?


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
Duane Degn's picture

Aren't DS18B20 sensors 1-wire devices rather than I2C. I'm pretty sure I2C uses both a data line and a clock line.

I really enjoy reading about your projects. It seems like a lot of them are used to in scientific experiments.

Thanks for posting this.

Geir Andersen's picture

That is absolutely true!
My mistake

Hello Geir,

As I am following your projects with interest, with this I see our projects combine but you were the first :). I am prototyping a datalogger which logs water-temperature per hour up to 20 meters deep. The idea comes from a non-profit organisation which monitors the enviroment of lakes in Holland (Project Baseline) as well as world-wide. Right now, temperature is measured using the divers dive-computer. These measurements aren't really precise, position, depth and time-of-measurement are not the same, thus my idea of having a string of sensors on a wire.

My idea was to use the DS18B20 in a daisy-chain since they're uniquely adressable and require only three wires. This, coupled with a version of your low-power datalogger is just what we need.

I'm trying to build one with a DS3231SN RTC, a micro-SD card and an arduino bare-bone, most likely a 328 running everything on 3V3 and 8MHz with the RTC signal waking the sensors and arduino up every hour. 

Can I ask you for advice and pointers?


By the way, will water @ 10 meters not infiltrate the area where epoxy and cable meet?

Geir Andersen's picture

Hi Leroy
I will gladly help you in any way I can so you should contact me with a PM.
I can also sell you one of my shields that would help you along.

As for your question regarding what happens at high pressures with regards to the Epoxy and cable, the answer would probably be: As the Epoxy encapsulate the sensor and the PCB completely it does not matter if water would enter the cable. The sensor is not inside a hollow epoxy casing but is a massive block of resin with the sensor inside.


ossipee's picture

There is nothing usual about any of your projects Geir, this one reminds me of days at sea in Marine science class. Taking water samples at various depths with nansen bottles. Todays equipment is so much better. As always enjoyed this project and the craftsmanship behind it!