Can I use any sensor via an analog input?

[beelzerob]

Sorry, I wasn't trying to make life difficult for you, just giving you the pros and cons of the methodology (SECU16 use) you were trying to use.

Hehe...ya, I know. But life would be a lot easier if I didn't know all those cavaets and failure methods and obscure little facts that means IT WON'T WORK.

PM sent to Jeff, hopefully he can stop by.

Here's the basic idea of what I'm trying to measure with the secu16:
Digital Input1: Flue surface temp (at this point, I can only find snap discs that will allow temp "measurement" up to 400 deg).
Digital Input2: Heat detector (simple open/close connection @ 135 deg)
Analog Input3: Water temp entering boiler (as measured on surface of copper pipe)
Analog Input4: Water temp leaving the boiler (as measured on surface of copper pipe)
Digital Input5: Handle position switch (safety to make sure output is re-engaged.

 

[JeffVolp]

I measure temperature at several places here with just a thermistor and a precision resistor added to a SECU16 analog input. The only issue with such a simple approach is that there is a limited linear range - about 30C - with better than one degree accuracy. At the extremes of a 50C range, the accuracy degrades to about 3 degrees C. The center portion of the curve can be adjusted to the midpoint of the range of interest by proper selection of the precision resistor. For example, 40F to 90F can be covered with high accuracy. It is possible to use this approach for much wider ranges, but some sort of look-up table is needed rather than a simple conversion formula.

There is a thread on the ADI forum on how to do this:

http://www.appdigusers.com/forums/ubbthrea...mp;Number=12047

Should you want to pursue this, just be careful the thermistor leads do not short to the pipe. You need some thermal contact, but you don't want any electrical contact.

Jeff

 

[JeffVolp]

Just a follow-up after reading your previous post. Thermistors can be used to read temperature well above the boiling point. We used wide-range thermistors to track the movement of fire in a ship compartment. That would have no problem with your flue temperatuer measurement.

Since you are interested in wide temperature ranges, you will probably need to use a look-up table approach. Essentially, you calculate several points over the range, and then interpolate between them. The accuracy will be limited by the SECU16 resolution.

Jeff

 

[beelzerob]

Thanks so much for the reply, Jeff...this seems that it might be within my realm to do after all.

I had a lot of trouble finding high-temp capable sensors...do you have a link to some thermistors that can handle, say, 400 deg? (If it goes anywhere near that, the alarms will already be sounding anyway).

And thanks again! I'll have more questions, I'm sure, but I feel more confident in going with just the secu16.

 

[JeffVolp]

Most inexpensive thermistors are good for 125C or 150C.

Allied lists an inexpensive one that is stable up to 400F:

http://www.alliedelec.com/Search/SearchRes...mp;Ntt=837-5195

And another more expensive one that is rated for 300C (572F):

http://www.alliedelec.com/Search/SearchRes...mp;Ntt=254-0006

(NOTE: If those links don't work, just search for the xxx-xxxx part number at the end of each link.)

Note that neither of those are the 1% tolerance units I recommended in the ADI thread, so you would have to do some calibration if you want accurate results.

Here is a typical industrial quality high-temperature thermistor:

http://www.thermometrics.com/htmldocs/hitemp.htm

However, it is apt to be more expensive than what you want to do.

With any of these, it will take some work with a spreadsheet to calculate the corresponding response curves.

Jeff

 

[beelzerob]

Thanks for all that help and links, Jeff, I *really* appreciate it. I'm super excited to find a thermister that can measure the flue temp. That industrial one looks like we could actually mount it directly in the flue! Of course, that would be a source of heat/gas exiting, besides being an impediment to cleaning...but maybe we can work around that.

Weird thing about that allied electronics site...your link worked, and so I saw there was one search result. I clicked on the result, and got an error message. So, I tried searching for the part number like you said....error message. I then went to alliedelec.com (from a google search!)....error message.! I can't seem to access that site at all. Very strange.

I'm very excited about being able to do this....and I'll probably have more questions about how resistors play into this, because that's a black hole of knowledge for me (I decided after my first electronics engineering class that it really was just "magic" after all).

Jeff, do you have a specific thermistor you'd recommend for measuring the temperature of water via the surface of a copper pipe? I know it's a somewhat poor approximation to the water temp, but we'd rather not open up the pipes and put a probe in there. Will pretty much any thermistor work, so long as we try to get the sensor onto the pipe and surround it with insulation?

THANKS!

 

[JeffVolp]

Jeff, do you have a specific thermistor you'd recommend for measuring the temperature of water via the surface of a copper pipe? I know it's a somewhat poor approximation to the water temp, but we'd rather not open up the pipes and put a probe in there. Will pretty much any thermistor work, so long as we try to get the sensor onto the pipe and surround it with insulation?

THANKS!

Copper is an excellent thermal conductor, so the skin temperature will be very close to the water temperature inside. As long as you can get good thermal coupling to the pipe, and insulate the leads, pretty much any thermistor should work. My choice would be one with 1 degree C accuracy to avoid having to do any calibration. One I've used is Mouser: http://mouser.com/Search/Refine.aspx?Keyword=527-0503-10K

It is tiny and relatively fragile, so you may want to mount it in some sort of thermally conductive enclosure first. If you do some searching, you can probably find some designed for exactly what you want to do.

BTW, I've had trouble with the Allied site too. For me the links didn't work, but searching directly for the part number did - but only once. It seemed like once the page was accessed, it wouldn't let me access it again.

Jeff

 

[beelzerob]

Jeff, that's straight awesome help. thanks!!

So if I use the one you linked at Mouser, I'd just connect one pin to the input, one pin to the common, and that's all there is to it? (of course, on the client side I'd have to convert the input voltage to the actual temp somehow....I'm assuming there's a table or some kind of conversion)

 

[JeffVolp]

So if I use the one you linked at Mouser, I'd just connect one pin to the input, one pin to the common, and that's all there is to it? (of course, on the client side I'd have to convert the input voltage to the actual temp somehow....I'm assuming there's a table or some kind of conversion)

It is not quite that simple. A 1% precision resistor must be installed in place of the "pull-up jumper" in the SECU16. The optimum choice for that resistor depends on the range over which you want best linearity. Then there certainly is some math involved converting the SECU16 A/D reading to degrees F.

The details for environmental temperatures are contained in that ADI thread I referenced earlier. I believe you will be running at higher temperatures, so you will probably want to adjust the linear region accordingly. If you contact me directly through my website jvde.us, I will send you the Quattro spreadsheet I used for that thermistor.

Jeff

 

[beelzerob]

It is not quite that simple.

It's NEVER quite that simple.... :angry2: Sometimes it's amazing how much we get accomplished in our HA projects considering all the little details it takes to happen....

Ok, thanks again for the info jeff, and I'll head towards that site and re-read the thread you linked. I'm going to draw up a list of the sensors I'll expect to use and then pass them by you just to make sure they ARE useable, or have been tested and approved. Then I can begin the learning process of what it'll take to use 'em right.

Thanks Jeff.