Question for the homebrew controller guys about temperature probes

[Lukas060606]

Hi guys,

I am relatively new to the site and if you read my posts so far, you'll see I'm all over the map. That's the way I am when I'm trying to figure out the right path to go down. I'm smack dab in the middle of planning here, and I want my system to be as automated as possible, with every fail safe in place that I can think of. With that said, my speed bump a the moment is temperature probes. My pop is an old time electrician and he taught me about voltage drop over long runs of cable. This is at house current mind you. That's leads me to what I'm trying to understand now. If I end up going with a heat storage tank (or two, as I'm currently thinking...in my furnace room...I'm lucky to have what the home inspector called "a lot of house", in a non-McMansion kind of way) to use with my wood/coal, and I want to measure the top/middle/bottom temperature of both tanks, if I use LM35s or any other probe, by the time I run the wire to my monitoring station (20 feet plus), wouldn't voltage drop be an issue? With such low voltage, the drop has to be significant. I'm thinking the signal will be worthless by the time it gets there. If I'm missing something, please let me know. Maybe it's as simple as manually checking the voltage at the probe and then at the monitor and calibrating from there??? If that is the case I'm thinking I need to do that check at my anticipated maximum and minimum temperature so I have the full calibration range??? This has had me head scratching for about a week now.

Thanks for the help. On a warmth note, I had the house at 85 on Saturday night (which is why I desperately want heat storage!!), it got warm on Sunday (50s) and there's been no furnace activity (except for DHW) since, but the house has been upper 70s the whole time. Tonight the outside temps are dropping fast but the Whitfield is holding up so far.

Thanks also for reading my long-winded post. To anyone that can help, please do.

 

[ewdudley]

...

[If] I use LM35s or any other probe, by the time I run the wire to my monitoring station (20 feet plus), wouldn't voltage drop be an issue? With such low voltage, the drop has to be significant. I'm thinking the signal will be worthless by the time it gets there. If I'm missing something, please let me know.

...

Possibly what you meant to say is 'with such low currents, the voltage drop has to be insignificant'.

Just calculate E=I*R for the sensors you have in mind, using the datasheet and looking-up the resistance for the wire you'd like to use.

Noise is another issue, but if you shield and/or isolate properly you can go a couple hundred feet. Most temperature controllers use an average of multiple samples over time, which helps a lot.

Since you're building a homebrew controller you can go ahead and oversample the signals at some exact multiple of the line frequency and eliminate background mains noise.

--ewd

 

[kabbott]

...

[If] I use LM35 or any other probe, by the time I run the wire to my monitoring station (20 feet plus), wouldn't voltage drop be an issue? With such low voltage, the drop has to be significant. I'm thinking the signal will be worthless by the time it gets there. If I'm missing something, please let me know.

...

Possibly what you meant to say is 'with such low currents, the voltage drop has to be insignificant'.

Just calculate E=I*R for the sensors you have in mind, using the datasheet and looking-up the resistance for the wire you'd like to use.

Noise is another issue, but if you shield and/or isolate properly you can go a couple hundred feet. Most temperature controllers use an average of multiple samples over time, which helps a lot.

Since you're building a homebrew controller you can go ahead and oversample the signals at some exact multiple of the line frequency and eliminate background mains noise.

--ewd

Correct... the current is so low voltage drop is insignificant. Noise on the other hand can be a problem. Study the data sheets and use recommended bypass caps
and resistors. If you are going totally homebrew you can use the lmxx sensors with a 4-20mA output circuit, Again see data sheet. With a system like the "Nofossil control system"
you can compensate for "offset" and "gain" on a per sensor basis although I have not had the need. My longest run is 20 feet or so.


By the way I have a mix of sensors, LM35, LM6x, TMP65... There are some that output 20mV/degree C as opposed to 10mV. may be sightly more accurate over long runs???

 

[Nofossil]

I've run LM61 sensors over 300' of cable with no issues. Some of the other sensor types are a bit more sensitive to induced noise, and it's really important that the sensor power and grounds NOT be shared with anything else.

 

[Lukas060606]

Thanks a lot for the responses. I think my problem is I don't work with these things all that often, so I forget the details. The light bulb should have gone on (with full voltage) that with little current, there's no issue. I really appreciate the help and am sure I'll have more questions along the way.

By the way nofossil, you're a big inspiration for me with what I'm trying to do with my system. I love your burn calculator and can see myself using it on a regular basis once I get my variables dialed in. Thanks to you and everyone else for sharing your vast knowledge.