Hot Tub Temperature Probe

[DavidK]

Was wondering what others have done to measure temperature in a liquid, the water temperature in my hot tub in my case?

Have put a DS1820 series inside a 12” of ¼ plastic tube, and capped it. But eventually water leaks in. Then went to Home Depot and bought some “Scotchkote,” dipped the plastic pipe in that, let it dry 24 hours and that seemed to work for quite awhile, but it eventually it started to leak again.

There are two other options I was considering and that was ¼ copper pipe with a soldered or encasing the DS1820 in epoxy inside a glass test tube. But I really need the 12” length to handle the varying water levels in the tub.

Looking for suggestions

[mackrackit]

I have never had long term luck it the liquid is able to contact the wires. Have tried epoxy, wax, silicone, etc.

Ha the the pipe or tube long enough to come out of the liquid on one end and cap the end in the water.

For our "chemical" tanks we have an extra fitting in the tank where the probe goes, then use a thermo well type plug. http://www.temperatures.com/twells.html

If the is a place in your piping where you could cut in a "T". The thermowell will also work there is you circulating the water.

[DavidK]

mackrackit
That is a great link. Now I know what to ask for, thermowell. The web link sure is good reading plus lots of links to research. Guess I was not to far off when I was thinking of the test tube.

My goal was to get a fairly accurate reading, but maybe that is not as important as I figured. Most people never take a temperature reading before they get in the shower, it is all relative.
Thanks for the tip
David

[duncan303]

Hi David

My goal was to get a fairly accurate reading

With a Hot tub the pipework remains all fairly close within the integrated unit, and could be arkward to find a place to cut a tee into, unless you have a separate Plant room. If you attempt to install an ABS TEE then you might need to make sure you have enough room for a slip coupling, careful if you are going to use a solvent cement, you only get a couple of seconds alignment. Much better to fix a "strap boss" and reduce down to your pocket. Consider mounting the entry to the pipe line from below horizontal thus reducing the requirement for venting(the pocket could fill up with air). Choose a place (any place, you can calibrate to any temperature) that has the most consistent water temperature ie before any heat exchanger or inline heater. So you are really looking at the area at the return drain. Do not worry about setting the pocket "into" the flow as turbluence will take care of uniformity, plus you will find that there will be less wear on the pocket itself, believe it or not but you actually have a water based sand blaster going on. If you have a cartridge or sand filter then you can place the pocket after them.

Then of course the fun starts by integrating the PIC to optimise the heating circuit to reduce overshoot and hysterisis in the heating circuit.


HTH

Duncan

[DavidK]

Duncan
Thanks for the input. Where the water comes into the filter reservoir there is a cover that I can mount the Thermowell to and drop my DS1820 right in. So I really don’t need to tap into the water lines. I think this is what you suggest, keep the sensor it in the return part of the process.

This afternoon I opened up the doors where the controller is and found a thermowell for the existing temperature sensor. Now it looks like I might be able to fit the DS1820 in the same well, but that probably is not the best place. My overall goal of this project is to only monitor the water temperature remotely. We have frequent power bumps in our area and my GFI trips occasionally, so I want to catch the trip by alarming the water temperature. The other reason is that my wife wants to make sure the tub is up to temperature before she runs outside. We have turned into real pansies.

Did a search on EBAY for a “thermowell.” Surprised me, found lots of them.

As far as replacing the pump/heater/light controller it might not be too bad. My tub is vintage so all I have to worry about is time-of-day, minimum temperature, run temperature, pump high/low speed and the light. I would probably use 24 VAC controlled contactors for the 30 amps of 240 VAC current, then use the PIC to control everything else. Not ready to go there yet, but it sounds like a fun concept to design.
David

[duncan303]

Hi David

I responded to your post as a spur of the moment, but have been thinking a little more.

Maybe in your particular situation, the temperature is not critical as in an industrial chemical process; you simply have a fairly large body of water that you anticipate bringing up to and maintaining a temperature within say +/- 2 DegC. Although the thermal characteristics of ABS pipework are very poor you could use this to your advantage, why not simply attach the 1820 directly to the outside of the pipework, AND insulate heavily around the section. You could mount the 1820 directly onto a wide metal band as a type of thermal heat plate.

Better still why not adapt the idea of a bog standard floating pool thermometer. Mount the 1820 at the bulb location; open up the float ball place circuitry, power supply, and a small RF unit inside, even turn the ball into a sexy multicolour “Light glow ball” to reflect the readiness of the tub! Have the RF transmitter send to a LCD or Oled display that you keep in the house, integrate another RF unit into the control panel and Viola. Hook another RF up into a USB dongle and keep an eye on things online, or have it send text updates to your mobile.

What a neat little run of projects all easily achievable and conjunctive, without having to use MPLAB IDE!

Now I wonder if the humidity phase boundary at the surface of the water would excessively interfere with the RF? Would 2.4Ghz be more reliable than 433 etc. I can’t off-hand remember.

I am probably not helping at all here, but am having fun.

Duncan

[b1arrk5]

David,

While it is possible to build a small thermowell using copper like you first suggested you need to know that chemicals like chlorine will eat the metal after an amazingly short period of time. You could purchase brass or stainless steel wells. Our hot tub had a stainless steel tube that held the heating elements, it was an easy place to attach a temperature sensor. As an experiment, you could try gluing the sensor right to the outside of the fiberglass (unfinished side) perhaps midway down, and then cover the sensor with insulation. I'd bet that this would be accurate enough for your needs.

When you get ready to take control of the whole process, I suggest reading the free 'Industrial Control' tutorial available from Parallax in their 'Stamps In Class' series. Good stuff on PID temperature control, etc.

Good Luck!
Jerry.

[DavidK]

Jerry
Your are correct, we don’t need any deteriorating copper. I will think about cutting the insulation away from the fiberglass to paste my sensor. But first I found a plastic hot tub thermowell on EBAY for less that $5. It screws right into a PVC fitting. Just the ticket!! Will give that a try first.

Built up a simple 433 MHz radio link using a 12F683 (TX) & 16F876 (RX), then tested it all weekend. Transmitted simple ASCII code every second at 2400 baud and checked for clear data; 10 errors in a little over 30 hours, that is better than 99% reliability. I can live with that. Coding up the DS1820 is next.

In my junk box I have some low voltage contactors that would work for running the pump, and heater. Your suggestion to review the Parallax Industrial Manual was great. It has some good stuff in it and it will give me a good start if I decide to replace the tub controller. Just might do it anyway just to prove it can be done. Thanks for the tip.
David

[precision]

I always makes to put a DS1820 series inside a 6” of ¼ copper tube. 1mm of ds1820 will remain out from tube to other end, After i seald with Mseal or M-Seal Easimix - 20 Epoxy Putty.( That uneffected withstands temperature up to 180°C ) and avilable in only in 0.37$ in india into tube formate.
one tube is sufficient for about 50 nos. 1820 sensors.
see property of epoxy putty http://www.pidilite.com/productinner...65&subcat2=184

[DavidK]

precision
A quick Google search and there does not appear to be a US distributer for M Seal. However, I did find “Mighty Putty” and “Plumbers Putty.” Maybe they are the same thing as M Seal.

Because of the chemicals in the tub I would probably try plastic first not copper and let it set in the water a week or so to check for leaks. Thanks for the suggestion.

[jtravillian]

DavidK -- I was going to try this:
http://www.embeddeddatasystems.com/p...PPER/STPSR3-12

I haven't ordered it yet, but I was thinking about poking a hole in styrofoam or similar and letting it float on top near the side.

Any thoughts on that?

[DavidK]

Looks OK. I have lots of DS1820's so I hoping to do it myself

[Archangel]

I doubt you would have too many problems with copper, based upon the following: The swimming pool at my mother's house is over 45 years old and it is piped with . . . copper, and no leaks yet.

[Acetronics2]

- 22nd March 2008, 20:15 ( 1st post )

- Today 09:13

Hi Joe,

No need for a probe ... the bath is pretty cold, now ...

Damn !!!!

Alain

[Archangel]

Hi Joe,

No need for a probe ... the bath is pretty cold, now ...

Damn !!!!

Alain

O/P must be a Polar Bear (Disguised as a duck) http://www.polarbearclub.org/ as he posted his reply today too , think hot steamy bath surrounded by snow . . . Oh, and fog rising off the bath . . .
or maybe he lives here on the Left Coast, where in some places it is forecast to be 81 tomorrow. How's the temp. there now?
http://www.usairnet.com/weather/maps...a/temperature/
EDIT: Ahh Haa he is from here! Seriously you should have been here yesterday, beautiful fall day, warm . . . Like Nice in the summertime, but without the topless sunbathers, Damn.