The ones on the left go to 1K platinum RTDs. The ones on the right are relay contacts to pretend to be a thermostat to our gas furnace. Top are digital and analog interfaces to the Integrity board. All low current, powered by a 12-volt wart and an LM7805 on the upper board.
Not to connect field wiring.
I'm getting a couple tenths of a degree C accuracy from the RTDs. The measurement and math are ratiometric against some 1K 0.1% resistors.
Incidentally, I recommend this gadget
for projects like this. It works as advertised and is easy to use. I expected to have to do some signal averaging on my temperature acquisition, but it's dead stable without. Of course, the braid+foil shielded RG174 runs to the RTDs may have helped.
expander),
out.
I got 100% spares for this project, and a bunch of bare boards for free, so it should last.
Absolutely *everything* about this project worked as expected, first time. Except the software, of course. The hardware I planned and checked. The software, I just sat down and started typing. I had a little trouble shelling out to the FTP stuff while running the realtime part, which I fixed by writing a separate program to do the ftp transfers.
The hard part was pulling the wires.
back to the LM135.
Yeah, I had to consider all sorts of failure modes so that the heat wouldn't stay on full-blast for weeks.
multi-tasking OS either.
100 C.The semiconductor sensors usually aren't very accurate. The RTDs are fantastic. I can't imagine how they calibrate them in production.
I considered the Z-wave home automation stuff, but their temp sensors only go down to 40F and are of unknown accuracy. Plus, I wouldn't have long-term maintainability, and remote access could be a problem. And I wouldn't have had a project for this week.
Here it is installed in the ski-boot closet. All I have to do is wire up the furnace control and clean things up.
ftp://jjlarkin.lmi.net/Auto_wired.jpg
Needs more tie-wraps.
John