Temperature sensor for water pump

Hi all! What’s are the best temperature sensor to monitor a water pump?
How can I connect the sensor to the pump, any ideia?

What temperature range are you dealing with? What are you measuring--Case temp? Water temp? Seal temp? How accurate do you need to be?

Assuming the water is liquid, an obvious choice is the DS18B20. It is cheap, adaptable, easy to use, reliable, and well supported on this forum.

You might also get some sensible information from


note that there are two programmes to use, one to sniff the address and the other to use the sensor.

A typical version is in a weatherproof probe that can be fixed to the output pipe under a foam sleeve.

Picture shows one in a home-made thermowell, the other under insulation.

Paulcet, I'd like to measeure pump structure, not water. My intention is to detect when pump is working without water. This is a situation I'd like to avoid and can burn the pump itself.
I need to do a test to know what is normal temperature so, using temp sensor, when temp reaches 10% more than normal temperature, system could stop pump motor.
It is not necessary very accuracy.

Nice pic Nick - I like it, looks just like mine, but less complicated. You've got to where I'm heading too. Temperatures are power.

robert - aren't you tackling this the wrong way? I run a large pump for pumping rainwater - by the time mine got warm enough to detect that it was running dry, it would be too late.

What sort of pump is it? There are thousands of different pump designs, centrifugal, peristaltic, regenerative, positive displacement, submersible, I could go on - you need to give a lot more detail.It's like telling your GP "I've got a pain". Some pumps are actually designed to run dry, peristaltic for instance, but not many.

Most pump motors will have built in thermal protection, but this doesn't protect the seals etc. You need to check whether liquid is running through it, not when it's too late. Look at what a pump does - depending on what is on the outlet side, you will have pressure on the outlet - put a pressure sensor (old washing machine pressure switch) in the line; use a paddle switch for flow; make your own liquid sensor with a 2N3904 transistor or almost any CMOS chip; use an optical device like a liquid level sensor - the refraction of water changes to the refraction of air - there are loads of ways to do it. But not by detecting an overheating pump. Having said that, some air flow meters use a hot wire, i.e. temperature - the more flow, the cooler the hot wire. If you want to monitor the pump body temperature, not a bad idea, strap a DS18B20, or a TMP36, or an LM35, thermistor, thermocouple in close thermal contact. That's the easy bit. Then Arduino can shut the pump down - what voltage by the way? - or is the liquid flow critical?

In a nuclear power station, every aspect of the pump operation - here I'm talking about pumps the size of cars, and critical too - from vibration to temperatures all over it, leaking seals and so on - all on a SCADA system. Not building one?

My intention is to detect when pump is working without water.

Detecting that by waiting for it to get hot sounds like good way to ensure the news is too late. If the pump is electric, I imagine acting on the no-load current would be an easier way and a lot more reliable.

I need to do a test to know what is normal temperature so, using temp sensor, when temp reaches 10% more than normal temperature, system could stop pump motor.

So what happens on a hot summer day or the depths of winter? Normal in winter is going to be different to normal in summer - so you might need to measure ambient as well.

Nick - I did pumps for a living and yes the big ones would be monitoring the load as well on all three phases. These would normally be in a control panel with switch-over to an auxiliary. What we haven’t heard yet is whether the actual delivery of liquid is critical - if it’s a pond full of expensive Koi carp, protecting the pump might be the least of the worries, like coolant round a reactor. Quite often, you would have to add a half a dozen different sensors round a critical pump just to be absolutely certain it was working. I like the old nails and screws in your set-up - but I can see you do take it seriously - mine is a plumber’s nightmare so I don’t let them in the house.

I like the old nails and screws in your set-up -

LoL I have just noticed that, and thought I really need to clean the mess up before I post pictures.

Nice pic Nick, thanks for your hints.
Most of pumps I'd like to monitor don't have thermal protections or other kinds of control.
Basically, I'd like to avoid leaks that causes pump to try to pump air instead of water, and pumping air can damage it.
The way to know about these leaks is monitoring pump temp, and them stop the electric motor and turn alarms on.
Here in Brazil we don't have to much weather temp variation throughout the year, so it is easier to rely on temperature variation. But a system could monitor temp outside and know what's the average temp of a present month to avoid temp variation because of the season of the year.
I'd like to develop a system as simple and cheap as possible, to be viable economically.
I'll try sensors you suggested.

If the pump is a jet pump and is pumping air, the output pressure will drop to nearly nothing. There are inexpensive pressure sensor shutoffs made for just this purpose.

Oh, dear.

Why not just tell us what sort of pump it is, pressures, is it flow critical i.e. does anything die if it fails, voltages etc. makes it a lot easier to help you. At least in Brazil it won't freeze up. You still will need to know what's coming out the end of it. I worked with many different types of pump here in the UK in critical water supply and waste water applications. Not one application relied on the pump going over-temperature to tell if it had failed, in fact if you designed something like that, you would be looking for another job - over-temperature was used to indicate other problems like no ventilation and bearings running hot - every one had either a pressure switch or a flow switch. I think that if you don't get some of the basics set out, you are going to struggle with the rest. I hope it's not someone's drinking water supply. Reliability depends on the right monitoring and some proper engineering.

Nick - I tend to notice stuff like that, usually when one goes through the sole of my shoe. I hope the dangly crimp terminals aren't live.

Ok! Very good ideas and hints!
No, no one will die if pump stops. But, I don't want clients of my system to suffer lack of water in their showers.
Pump is used to pump water for residencial and comercial buildings and small factories too.
Well, my system could monitor pumps temp and ambient temp but now I realize that there are other options to check if there is air, and consequently leak, inside plumbing.

Good luck.

Sounds like a large project, but even if it is only showers (booster pump?) you will need a rock solid control and monitoring system - booster pump - pressure first, use the temperature to catch faults developing before they cause a failure.
I wonder if this is even slightly commercial, whether Arduino is the right course - I would be looking at PLC's (Programmable Logic Controllers) which are made for these types of job - clients generally take in a dim view, in UK at least of anything that uses one-off software, especially without a track record. If it went wrong, you could be on your own, I know - writing some code that measures pressure is going to have to have a lot of error checking to be robust enough. I programmed dozens of PLC's in a large public water utility, but believe me, if you get it wrong, your neck will be on the line.
By all means use an Arduino (which I presume is why you are here) to test some ideas out - there is also the problem of turning a prototype into something that looks the part.
Also bear in mind is that any advice you get here is free and takes up a lot of peoples time.
Get in the habit of framing your requirements clearly and precisely. If this has taken this long to find out that it might be something to do with showers and you do the same the Programming sub-forum with vague requests, you will get your head bitten off - sensor folk are pussy-cats.

You better use a Water Flow Sensor and this one used with Arduino.

Thanks a lot Tigger and other member!
I'll be more precise next time I do questions on this forum.
PLC it not enough for my system. I need to develop lots of other functionalities.
Good, water flow sensor is a good idea too.