Recommendations for remote temperature sensors (air)

Posted on
Tue Nov 24, 2009 8:00 pm
wwhigginsjr offline
Posts: 45
Joined: Feb 06, 2007

(No subject)

Matt-

OK, I had not clicked on the Upload All Settings button earlier. Just tried it, and got a message that indicated I needed to set the trigger point OFF to ON greater than the trigger point ON to OFF (these were currently at the default values of 0, 0). Indigo would not upload unless I did that for BOTH analog inputs (even though I'm using just Analog Input 1). Used 600, 200 as per the example in the Indigo message dialog box. Then clicked on the Upload All Settings button. Indigo showed action on the PowerLinc and Event Log (both good, I guess).

Now I click on the Get Input/Output Status button, and I get:

Analog Inputs: 203, 114

These change a bit each time I click on the Get Input/Output Status button. Have not tried the hairdryer yet, BUT 203 doesn't mean anything (20.3F???), and there's nothing connected to Analog Input 2. I'd think that Analog Input 1 should read something near 650, and Analog Input 2 should read 0 as before. How do I fix?

Posted on
Tue Nov 24, 2009 8:16 pm
matt (support) offline
Site Admin
User avatar
Posts: 21417
Joined: Jan 27, 2003
Location: Texas

(No subject)

The value shown is not volts or millivolts. As per the EZIO8SA docs, it is an integer value between 0 and 1023, which represents the voltage value from 0 to 5VDC. So a value of 203 would be 203/1023 * 5VDC = 0.99VDC.

Is your unused analog input grounded?

Image

Posted on
Tue Nov 24, 2009 8:28 pm
wwhigginsjr offline
Posts: 45
Joined: Feb 06, 2007

(No subject)

Matt-

Thanks, but still doesn't make sense. 0.99V is 99 degrees F. Ambient temperature is about 65 degrees F. Multimeter reads 0.649 (= 64.9 degrees F). Somethings still not right in Indigo, I think.

Sorry to ask a stupid question, but I'm pretty new "roll your own" electronic devices scene. What do you mean, "is my unused analog input grounded?"

Nothing's connected to it at all, so I suppose the answer is no. Does the fact that it reads 114 mean that it's picking up stray voltage from somewhere? How do I eliminate?

Thanks again. Primary concern is how to get from Indigo reading of 203 to degrees F.

Posted on
Tue Nov 24, 2009 8:40 pm
matt (support) offline
Site Admin
User avatar
Posts: 21417
Joined: Jan 27, 2003
Location: Texas

(No subject)

Indigo is reporting back the raw integer value that the EZIO8SA is giving it. I'm pretty positive (not 100%, but close) that this isn't an Indigo bug. There isn't much room for the problem to be on the Indigo side of things -- it just gets the raw value (0 to 1023) from the EZIO8SA.

By grounded I mean connecting a wire from that EZIO8SA terminal to the GND terminal. If you haven't done that, then the value will float.

Have you tried your hairdryer test to see how the value climbs/changes in Indigo?

See seanadams post above (on the other page of this thread). He details how to convert the temperature to the raw value. In fact, the value he comes up with based on a temperature of 65 °F is 202, remarkably close to the 203 being reported. It looks like there is, or was, a typo in the original EZIO8SA documentation that said the range value was 0 to 5VDC, when it is actually 0 to 3.3VDC. It sounds like it is working correctly to me.

Image

Posted on
Tue Nov 24, 2009 9:13 pm
wwhigginsjr offline
Posts: 45
Joined: Feb 06, 2007

(No subject)

Matt-

BINGO! I started thinking along the same lines, and started scaling things off the 3.3V as Sean did. Your reply came in just as I was finishing the hairdryer test. So here it is:

(Indigo reported value) / 1023 * 3.3V * 100 = degrees F

203 in Indigo becomes 65.48F and 282 in Indigo (with the hairdryer on the chip; low speed) becomes 90.96F.

You know what my next request is! I wish there were a way for Indigo (not AppleScript, etc.) to take that formula and report volts or millivolts directly (as the LM34DZ chip does). Reporting 0.654V is easily "converted" in the head to 65.4 degrees F. A bit tougher to do X/1023*3.3V*100 in the head (although degrees F works out to be about 1/3 the Indigo reported value).

Thanks for your and Sean's help. Thanks also to those who provided advice/leads on soldering. YouTube seemed to me to be the best source for soldering instruction ("a picture/video is worth a thousand words;" "monkey see, monkey do" and "I can do that" also seem to fit).

I'll ground Analog Input 2, and presume that will zero it out/keep me happy.

Next part: how to waterproof this thing? I saw a post from dalenis in a thread titled waterproof temperature from 2003 (!) that said I could just dip it in epoxy once it was wired properly to waterproof it. What kind/brand/type of epoxy specifically?

I want to put this temperature probe into a hot tub. Idea is to send an alarm to monitoring company (via Indigo, NC EZIO output relay, and existing alarm system) plus an e-mail to self (via Indigo) if temperature of hot tub water falls below X (indicating heater has failed; hot tub in danger of becoming skating rink!).

Again, what type of epoxy should I be asking for at Home Depot, etc? Thanks to any/all for help.

Posted on
Tue Nov 24, 2009 11:37 pm
berkinet offline
User avatar
Posts: 3290
Joined: Nov 18, 2008
Location: Berkeley, CA, USA & Mougins, France

(No subject)

wwhigginsjr wrote:
Again, what type of epoxy should I be asking for at Home Depot, etc? Thanks to any/all for help.
Congratulation on getting your temperature input working. I think what you want is potting compound. Take a look at this page and click on Potting & Encapsulating Selector Guide. They list a range of products, and importantly for you, provide their thermal conductivity. You might also goggle "potting-compound" Given that your probe will be submersed 100% of the time, it will sooner-or-later reach the same temperature as the water no matter what you pot it in. But, if the temperature starts falling you would want to know as soon as possible to start sharpening your figure skates.

seanadams has mentioned using hot-glue in another thread, so he may have some ideas to offer as well.

Posted on
Wed Nov 25, 2009 12:01 pm
seanadams offline
Posts: 489
Joined: Mar 19, 2008
Location: Saratoga, CA

(No subject)

Glad to hear it's working! So is this an in-ground, or a pre-fabricated/commercial hot tub?

For freeze prevention the most important thing is the temperature at the pump impeller, and a sensor in the vessel is not going to tell you that. In a pre-fab tub, the pump and plumbing is protected from the elements so it will be the last thing to freeze whereas in an in-ground tub, the pump is usually exposed so it will freeze sooner.

For an in-ground tub, you should have a sensor that runs the pump when the air temperature is below 32°F. The deep water in the tub is going to be the last thing to freeze, and even if the heater is not running, just having the pump on is enough to protect the entire system. Unless you live in Alaska.

If you immerse your sensor in the water there are some safety considerations. Commercial pool and spa electronics are designed to ensure a high degree of isolation, and to ensure that any faults are directed to ground, and to ensure that in the event of a fault the power is quickly cut. To address each of those things you need to do all of the following

1) permanently encapsulate the sensor and all electrical connections
2) Bond the "Gnd" wire on your cable by connecting it to the spa's ground reference.
3) make sure the PLM is plugged into a GFI. If the whole spa is on a GFI and you are taking power from inside a junction box in the spa, then you may be OK here already.

Even then you are probably not in compliance with the electrical code because the electronics you're using are not listed for this purpose. If there's a weak link in the system it would the PLM's internal power supply but there's not much you can do about that. Installing a bonding wire is the most important thing as far as that's concerned.

Posted on
Wed Nov 25, 2009 1:04 pm
berkinet offline
User avatar
Posts: 3290
Joined: Nov 18, 2008
Location: Berkeley, CA, USA & Mougins, France

(No subject)

seanadams wrote:
... If there's a weak link in the system it would the PLM's internal power supply but there's not much you can do about that...
To add to that. I am not sure about the PLM, but on most Insteon devices the internal 5 volt power supply has one side common with the 120 volt A/C line.

So, given the points seanadams made, you might be better off strapping the sensor to one of the spa's pipes (like the return to the pump). A little thermal grease between the sensor and the pipe probably wouldn't hurt either, and then maybe insulate around it (to isolate it from the air).

Also, since you seem to have an I/O input free, and there are available conductors in the cat-5 cable, and with your newly learned soldering skills :D you might want to consider adding a second sensor as well for the air temp, or monitor a second point in the spa.

Posted on
Wed Nov 25, 2009 1:31 pm
seanadams offline
Posts: 489
Joined: Mar 19, 2008
Location: Saratoga, CA

(No subject)

berkinet wrote:
seanadams wrote:
... If there's a weak link in the system it would the PLM's internal power supply but there's not much you can do about that...
To add to that. I am not sure about the PLM, but on most Insteon devices the internal 5 volt power supply has one side common with the 120 volt A/C line.


This is true of the Smarthome products that do NOT have low-voltage connections to the outside world (switchlinc, lamplinc, appliancelinc, etc). The internal +5V rail is at 110VAC potential. It's a trick that lets them eliminate a bulky and expensive transformer but you can only do it where the power requirement is very small and the low voltage electronics are not exposed.

PLMs, and 3rd party devices based on the PLM platform, have transformers and are therefore isolated from the mains voltage, because they have to be in order to be able to connect to another device which is at earth potential. However I am talking about even further considerations for wet/outdoor conditions.

Posted on
Wed Nov 25, 2009 2:05 pm
wwhigginsjr offline
Posts: 45
Joined: Feb 06, 2007

(No subject)

Thanks for all the help, and the specifics of your replies. I really couldn't have made any progress on this project without all your help/knowledge, Matt, Sean, berkinet, and anode (on a separate EZIO thread) especially.

The hot tub is a 2008 Grandee (model name) from Hot Springs (manufacturer/brand). According to the local dealer, Hot Springs is working on a add-on module to interface with an alarm/monitoring company for low temps (etc?), but it is not finished/available yet. Pretty sure that if I set one of you guys loose on the existing controller/circuit board, you'd find the terminals I need to read the temp off the Hot Springs temperature sensor directly! Any takers???

May be I could locate the Hot Springs temperature sensor, and get a multimeter voltage off of it. Anyone think this is a good idea? Or a waste of time (e.g., already tried by others, sensor is inaccessible, sensor works much differently than LM34DZ chip, etc. Advantage is that the Hot Springs temperature sensor is undoubtedly grounded, UL listed, etc. to meet electrical code.

Set the idea above aside for the moment. Answers to specific questions:

1) I will encapsulate my "home-grown" sensor in potting compound (see earlier post).
2) I will bond the home-grown sensor's GND wire to spa's internal ground, which is connected both to house electrical system ground and to the rebar and bonding wire in the "pit" that the hot tub sits in (top of tub is at grade; hot tub was dropped into pit to appear as if it was/is "built-in").
3) PLM will be in basement of the house and connected to my home-grown temperature sensor via spare/empty conduit that runs between hot tub pit and basement. By code, all power outlets in basement are GFCI.
4) Strapping to pipe might be better from aesthetic point of view (sensor not seen dangling in the tub) as well as safety. How reliable is the thermal transfer from water to pipe to sensor surrounded by colder air (e.g., "a reading of 95 F on pipe sensor indicates true water temperature of 104 F" or similar)?
5) Was thinking of other places I need to monitor for potential freezing of pipes. Two other problem areas come to mind, but thought I might monitor outside air temperature and the temperature in spa equipment compartment (curious as to differential; undoubtedly becomes less interesting over time!).

Thanks again. Keep the suggestions coming![/list]

Posted on
Wed Nov 25, 2009 2:26 pm
berkinet offline
User avatar
Posts: 3290
Joined: Nov 18, 2008
Location: Berkeley, CA, USA & Mougins, France

(No subject)

wwhigginsjr wrote:
...4) Strapping to pipe might be better from aesthetic point of view (sensor not seen dangling in the tub) as well as safety. How reliable is the thermal transfer from water to pipe to sensor surrounded by colder air (e.g., "a reading of 95 F on pipe sensor indicates true water temperature of 104 F" or similar)?
You should probably insulate the sensor from the ambient air as much as possible. But, that should be relatively easy. Or, maybe there is a location in a protected area or an already insulated enclosure where you could place the sensor.

If you do a good job of coupling the sensor to the pipe and insulate it well, the impact of air temp should be minimal.

Posted on
Wed Nov 25, 2009 3:06 pm
seanadams offline
Posts: 489
Joined: Mar 19, 2008
Location: Saratoga, CA

(No subject)

Strapping to the outside of a PVC pipe will not work well. It would have to be metal but you're not going to find much of that in there.

If you really want to do this the right way (no wires hanging over the side of the tub) this is what you need:

http://www.pentairpool.com/pdfs/Linear- ... sor-IG.pdf

You drill a hole in the PVC and put this in the hole, then strap a hose clamp around it to hold it against the pipe. This is what a professional installer would use for a pool, for example.

There's more math involved in reading a thermistor but it can be done with Applescript and the EZIO.

Posted on
Fri Dec 04, 2009 3:03 pm
artpics offline
Posts: 232
Joined: Feb 24, 2009
Location: Calabasas CA

(No subject)

I have done it, i bought a weather station it can use 3 sensors, that measures temp and humidity all are wireless :)

1. is in the snake enclosure
it triggers the snakes light via temp

2. is in any room i want which will control the T1900 thermostat.

thanks to Berkinet for help with the scripts that pulls the variables from the weather software into indigo

Posted on
Thu Jan 28, 2010 1:13 am
Snakesitter offline
Posts: 218
Joined: Feb 25, 2008
Location: Los Angeles, CA

Re:

I wrote:
...tAlso, at the present time the AppleTalk for the Phidgets controller is broken in Snow Leopard (10.6)... ...But, it is being worked on and should be available soon.

We're now one big step closer: http://www.perceptiveautomation.com/userforum/viewtopic.php?f=4&t=5454

Posted on
Wed Dec 22, 2010 5:51 pm
gmusser offline
Posts: 290
Joined: Feb 12, 2005
Location: New Jersey

Re: Recommendations for remote temperature sensors (air)

kpfriedberg wrote:
if you hook up the duct sensor (http://www.smarthome.com/30409/Venstar-Remote-Duct-Digital-Temperature-Sensor-Wire/p.aspx) to S1 within your Venstar, you can put the probe wherever you like (with enough wire), and then you can, on Program 26, change which sensor is the 'default' sensor, out of the box its the ambient air temp, but you can make it S1, i do this for my radiant floor heat where i have the sensor reading the floor temp, and that drives when heat goes on/off. It works quite well actually.


Can the default sensor be selected dynamically, via Insteon commands?

George

Who is online

Users browsing this forum: No registered users and 22 guests