Attaching temp sensor in PVC pipe.

[Lyngtoft]

I want to add a temp sensor to the pool pipes and was a bit unsure how to do that.

I ended up with simply drilling a hole in the pipe, putting the sensor in and trying my best to glue it up with PVC glue. But to be honest I'm not feeling very secure about this.. Any one else who have done something similar?



I have seen a few others using some sort of threaded insert, but I have no threading bits to cut that with.

 

[Jimrahbe]

L,

Welcome to TFP... a great place to find the answers to all your "Once in a year and a half" questions...

That looks like a sensor that would be used for a Spa... I too would be worried that over time it could pop off...

The 10K temperatures sensors made for pool automation systems are made to be installed in a hole drilled in a PVC pipe. They are held in place by a large hose clamp..

Did you really join us in 2016 or is that a database error on our part???

Thanks for posting,

Jim R.

 

[Lyngtoft]

Yes I have been here quite a long time, but usually just read Actually in the process of building a chlorinator based on mas985's guide somewhere on here as well.

The sensor is a waterproofed DS18B20 which I connect via a Zwave reciever. Unfortunately the only off the shelf equipment available here in Denmark when it comes to automation is ridiculously expensive.

a hose clamp that would definitely give me more comfort, but not really possible with that probe. I think I'll have to figure out how to get it threaded correctly so I can feel a bit more sure.

Thanks for your input

 

[PoolguyinCT]

That’s going to pop right out.. get a sensor with a saddle clamp. I would leave my house with that rig running.

 

[JoyfulNoise]

So here’s a way to hack this -

Cut into that pipe and install a PVC T union. The two ends of the T will match your pipe size (ex., 2” glue joint) and the center of the T will be a smaller diameter threaded pipe connection, ex., 3/4 female pipe thread.

Next, get a 3/4” male threaded plug and drill a hole into the top of it. Put the thermistor in the hole and pot it with a thick blob of epoxy. Thermistors are cheap as are PVC pipe plugs so just make a bunch of them.

Screw the plug into the T with some pipe tape or liquid Teflon tape (or both ) and you’ve got yourself a solid connection that is easily replaceable in the event the temp sensor dies.

 

[kolcun]

Take a look at what these guys did with a 3D printed bracket, and o-ring setup. Held by hose clamps.

Pool Controller - Hackster.io

 

[Bama Rambler]

You could also install a threaded side port tee like Matt suggested and use a stainless steel cord grip to secure the thermistor in place. Just get a cord grip (CGB) that the bushing range matches the OD of the thermistor.

 

[Lyngtoft]

Thanks for all your inputs, I have a "version 2.0" which I'm feeling a bit better about:

I figured out how to cut a thread with a tap I think it's called. Together with an IP68 coupling of some sort that fits very tight to the probe and finished with some glue:


Although I figured out the T should probably have been pointing down instead, hoping it will work fine regardless.

/Kim

 

[rogerhammond]

Worth noting that if the temp probe is on the suction side of the pump, the danger is the probe getting sucked in to the pipe rather than it popping off, unless there's a possibility of creating back-pressure before the pump? (pump below pool level for instance and the pump being off).

 

[YonnyPiscinas]

Just the thread I was looking for. I'm installing these type of sensors (well sonoff sensors with a TH16) and was thinking of having one on the discharge pipe and one coming out of the solar collectors to monitor the difference in temp. Once the temperature is no longer rising (cloud/rain etc) then it will shut off.

Looks like I need to install a couple of new T sections. Haven't found any T's with a thread on the inside yet though. Tricky to find here.

 

[Rattus Suffocatus]

The OP threading the PVC is likely a good enough solution as long as it doesn't leak. I have worse than that on my beer brewing setup (under less pressure though) and it's held up for close to 5 years even with boiling and cooling cycles. That's a cool idea. I would like to eventually have the pool temperature to be remotely read (and I never plan on buying Factory Automation). That or hacking the Ambient Weather wireless floaty. I suspect that a RPi, an RTL-SDR and rtl_433 could read that protocol (rtl_433 tuned to 915 MHz reads my Ambient WX station just fine). Having a temp sensor in is probably better though. the DS18B20's are pretty good in normal temperature ranges, very easy to use. More pool hacking projects to do.

And I want to experiment with solar heating on the pool. What are you using for a panel design? What I want to try to get working is a photo-voltaic panel running a bilge pump (or multiples). When the sun isn't strong enough to run the pump it will shut off by itself. Sort of like what you are doing with the delta T control. But that way I can claim I am 100% solar!

Thanks for the Sonoff TH16 tip too.. I wasn't aware of that one and they are cheap...and self contained for your project. Neat.

 

[YonnyPiscinas]

The OP threading the PVC is likely a good enough solution as long as it doesn't leak. I have worse than that on my beer brewing setup (under less pressure though) and it's held up for close to 5 years even with boiling and cooling cycles. That's a cool idea. I would like to eventually have the pool temperature to be remotely read (and I never plan on buying Factory Automation). That or hacking the Ambient Weather wireless floaty. I suspect that a RPi, an RTL-SDR and rtl_433 could read that protocol (rtl_433 tuned to 915 MHz reads my Ambient WX station just fine). Having a temp sensor in is probably better though. the DS18B20's are pretty good in normal temperature ranges, very easy to use. More pool hacking projects to do.

And I want to experiment with solar heating on the pool. What are you using for a panel design? What I want to try to get working is a photo-voltaic panel running a bilge pump (or multiples). When the sun isn't strong enough to run the pump it will shut off by itself. Sort of like what you are doing with the delta T control. But that way I can claim I am 100% solar!

Thanks for the Sonoff TH16 tip too.. I wasn't aware of that one and they are cheap...and self contained for your project. Neat.

My TH16 modules arrived just last night. Just 2 for now to test. I now have one sensor dipped into the pool and one on the roof next to the solar panels.
Plan is to cable them in properly next week and hook the roof TH16 to the pump.
I currently have a basic sonoff (R1/R2) switch controlling the pump. Used IFTTT last night to switch on the basic sonoff switch from the TH16 (temp solution). Currently set the sun temp sensor at 35 to activate the pump for the solar panels. Will experiment with that today as the sensor is in the sun - may need to lower that threshold and location. Not sure if the sensor should be resting on a surface to measure the enegry being absorbed or free hanging without contact.

I'll be buying a 3rd sensor/TH16 to plug into the solar panel discharge pipe.

The sonoff temp sensers are short with a small jack. They have 3 internal wires. I had some old cable that was used for the alarm sensors in the house which was about the same gauge and had 3 wires - perfect!
Cut the sensor wire and soldered in an extension. But Sonoff do sell 5m extension cables. 30-60 day delivery time from China!

My panels are made in Brazil - the standard black solar collectors similar to most you see on Florida roofs.
Although I built my own glazing due to constant coastal winds here. We don't get that many days with nice hot winds like Florida. Strong sun but very variable winds and temperatures.

Last night air temp dipped to 16c and was very windy with gusts up to 30 mph. Pool was at 31c at 10pm and now at 8.30am has fallen to 30c. Not bad at all with the cover (raised cover above water surface). May invest in a bubble type cover also to float on the surface to increase insulation properties.

 

[Rattus Suffocatus]

Good deal. I didn't know if you were making your own panels or not. Choices are pretty limited for DIY here, but I suppose since it isn't rocket science that what I can get would work. Supposedly it would give me about 1.5 more months of season if I do it properly with enough panels.

We are obviously just coming out of the coldest part of winter here, though in Arizona it is nothing like I grew up with in the Midwest. Water temperature just recovered to 50F (10C) yesterday. I have played years ago with ESP8266 devices (WeMos type) but especially if I could wait for delivery from China those Sonoffs are packaged up really nicely and are dirt cheap. (They are 3x the cost if sourced from a US warehouse.) I should look into it. Though it looks like you have to do a bit of work to get the temperature sensors to talk to Amazon Alexa (the itead software doesn't seem to do anything but switch on and off) which I would want to do.. meaning loading 3rd party firmware and setting up a MQTT connection to a server somewhere in the cloud. It's been a while since I toyed with that. But for panel control they can 100% be run autonomously with several different firmwares for the unit and it would be pretty trivial to write your own from scratch with Arduino as well. Neat. Again thanks for pointing this to me.. I've not done ESP8266 projects since before 2016 apparently so I missed those. One other issue for me is that I plan on running 12V dc pumps so maybe digging out the old WeMos's is a better idea for my setup, especially if I control the pump on or off.

Once you have water flowing over the DS1820's then you shouldn't have any issues with temperature irregularities. I imagine if you wanted to write your own code you could even periodically run the pump for a few seconds and then sample the temperature sensors for turn on.. You wouldn't do much inadvertent cooling if you coded that correctly.

 

[setsailsoon]

Lyn,

You can also get with a screw in enclosure to install like Matt said above but without the epoxy mess like one of these: (USP10981) from a place like here





Or do what most pool controls use with a clamp fitting and an o-ring underneath like this from most pool supply places:



I think the first is better since the plastic ones seem to fail more often than they should.

I hope this helps.

Chris

 

[AUSpool]

I have a temp probe on the suction side that is just a hole and rubber grommet as supplied with by the solar controller. Works on the suction side but I wouldn’t do this on the pressure side.

On the pressure side I retrofitted a 2X2X3/4” threaded snap on but could have used a standard 2X2X3/4” I guess. I’ve used the 3/4”(20mm) threaded outlet for a pressure gauge but could thread anything on, a threaded probe, a probe in a 3/4” probe gland or probe epoxied in a 3/4” threaded plug. There’s really no need to glue a probe gland, cable gland or compression fitting.

A temp probe on the return line may not detect a difference from the inlet probe. I can feel that my water return is warmer than the pool water but the difference is less than 1degC and not detectable on any of my thermometers. My controller turns the pump on when the difference between pool water and roof temp is greater than 10degC.

 

[YonnyPiscinas]

I have a temp probe on the suction side that is just a hole and rubber grommet as supplied with by the solar controller. Works on the suction side but I wouldn’t do this on the pressure side.

On the pressure side I retrofitted a 2X2X3/4” threaded snap on but could have used a standard 2X2X3/4” I guess. I’ve used the 3/4”(20mm) threaded outlet for a pressure gauge but could thread anything on, a threaded probe, a probe in a 3/4” probe gland or probe epoxied in a 3/4” threaded plug. There’s really no need to glue a probe gland, cable gland or compression fitting.

A temp probe on the return line may not detect a difference from the inlet probe. I can feel that my water return is warmer than the pool water but the difference is less than 1degC and not detectable on any of my thermometers. My controller turns the pump on when the difference between pool water and roof temp is greater than 10degC.

Good point there. The sensors I have only show 1c increments. So if my panels are heating 0.5c on a less sunny day only I may not see a difference.

Similar to yours I have set the threshold to turn the pump on when the roof temp reaches a certain number above the pool temp.

 

[jseyfert3]

That or hacking the Ambient Weather wireless floaty. I suspect that a Rasberry Pi, an RTL-SDR and rtl_433 could read that protocol (rtl_433 tuned to 915 MHz reads my Ambient WX station just fine).

Interesting. Makes perfect sense. I'll have to file this away for future reference. I've been meaning to get a RPi anyway (plus several coworkers keep telling me to learn Python). I was actually thinking of getting one of the Ambient WX stations, just the "simple" $70 version with up to 8 remote temp/humidity sensors and data logging. For whatever reason, this was not compatible with the Ambient WX floating pool temp sensor that worked with a cheaper, non-data logging system.

Given I'd like to set up something to track pool temp vs ambient sunlight/wind/temp, so I have a baseline to compare if I eventually add an air conditioner pool heater (like the Hotspot FPH) and/or solar panels so I can track how much heat those sources are adding and how much of an affect they have. Plus it would allow me to compare various insulating or windblocking on heat loss (above ground Intex pool) . An Pi would probably be the perfect tool for the job. Could even throw on some current sensors and track overall pool and spa power usage. I already wanted to track spa power usage as a function of temp and ambient conditions as I had a crazy idea to use a high quality domestic heating panel to heat my spa while in standby to offset as much standby power loss as possible, but I want to make sure it could actually make financial sense, or at least break even, before I would start that project...

 

[Katodude]

Go with the idea from Setsailsoon. Get one of these.

Amazon.com: 1/2 inch thermowell Stainless Steel 304 with Plastic Cap for Beer fermenter ds18b20 pt100 Homebrew Boiler 30 50 100 150 200 300 400 500mm (30mm): Industrial & Scientific

Amazon.com: 1/2 inch thermowell Stainless Steel 304 with Plastic Cap for Beer fermenter ds18b20 pt100 Homebrew Boiler 30 50 100 150 200 300 400 500mm (30mm): Industrial & Scientific

www.amazon.com

put it in a T connector that you can get from any of the big box stores and you are all set.

 

[Rattus Suffocatus]

I also brew beer and I suspect those thermistor type thermowells will have too much thermal mass to do what he wants... especially if he isn't getting 1C difference in and out. I have both in my electric brew kettle... I use a PT100 resistor which is more sensitive and the DS1820 I use to also monitor (that I had to rig because I've never seen one properly packed up with threads on it) the system is far quicker in settling and more accurate, even at only a 0.5C accuracy. A K-thermocouple is less accurate than the PT100.

The Thermowell only might work in this case, but they do slow down the reading rate a lot. The good part is that minutes even probably don't matter here. I still think a better way to approach this is with light monitoring to make sure it doesn't run at night. That and a probe in the pool itself if you want to make sure the water doesn't get too warm.

Hopefully I will be able to figure this out a little. My back yard faces north so to do any solar I have to put the panels at ground level anyway.. the bilge pump idea may very well work because if I do it correctly I shouldn't have that much head on the pump. The problem with the roof solar systems that is you need a real pump to pump water up 15 feet... and that negates a lot of the "free" solar energy. I have a small $16 pump to play with on my 100W panel and battery that I have to also play with. I might get something to work. I'll let you all know if so...

 

[Rattus Suffocatus]

Interesting. Makes perfect sense. I'll have to file this away for future reference. I've been meaning to get a RPi anyway (plus several coworkers keep telling me to learn Python). I was actually thinking of getting one of the Ambient WX stations, just the "simple" $70 version with up to 8 remote temp/humidity sensors and data logging. For whatever reason, this was not compatible with the Ambient WX floating pool temp sensor that worked with a cheaper, non-data logging system.

Given I'd like to set up something to track pool temp vs ambient sunlight/wind/temp, so I have a baseline to compare if I eventually add an air conditioner pool heater (like the Hotspot FPH) and/or solar panels so I can track how much heat those sources are adding and how much of an affect they have. Plus it would allow me to compare various insulating or windblocking on heat loss (above ground Intex pool) . An Pi would probably be the perfect tool for the job. Could even throw on some current sensors and track overall pool and spa power usage. I already wanted to track spa power usage as a function of temp and ambient conditions as I had a crazy idea to use a high quality domestic heating panel to heat my spa while in standby to offset as much standby power loss as possible, but I want to make sure it could actually make financial sense, or at least break even, before I would start that project...

Honestly.. it'a about a 15 minute project if the protocol is supported. I suspect it is. I do know that the floaties are on 433 MHz and not 915 MHz like my wireless weather station is. But I suspect that all of the data protocols are supported. It turns out that the Ambient is just a relabeled Chinese station ... I am sorry I forget the name... and it instantly identified all of my sensors on the 915 MHz station. I could have just piped the data into a script and then done anything I wanted with it.

I am tempted to try it.. maybe in the spring when the pool is warm enough to use again. I guess for $30, if not having the display inside my the door is still better than what I am currently using...