o1bigtenor:
> On Fri, Mar 22, 2024 at 4:02 AM Lars Noodén <lars.nooden@???> wrote:
> > On 3/21/24 23:13, o1bigtenor via Dng wrote:
...
> > > (Another project where it would be nice to be able to use a PID control
> > and
> > > be able to read 'accurately' and repeatedly to +/- 0.1C (process
> > > requirement!).)

±1 °C is easy, ±0.1 °C is very hard. Be prepared to set up a
calibration routine if you really want that and I suggest using higher
quality PT100 sensors.

...
> > DS18B20

I have used that one, you can attach a multiple of them to a three wire
cable ("1-wire" bus). Also you can start the temperature conversion with
a single command (like a broadcast), though I had problems with them in
the long term. Some dropped out, didn't investigating it further.
I have some code and designs if you're interested.

///

The -40°C qualification isn't hard to attain, just use suitable grade
of components, high humidity is a bigger problem, water always finds
it's way everywhere long term, but it helps to keep the electronics
warmer than the air. You could also coat the boards.

You could use:
. (long) cables from each sensor to a central warm and dry enclosure
. "die cast" (what is the right word) the sensor electronics and use
water sealed connectors like thoose found in cars and use a CAN or
RS485 bus

///

I have some designs and code at (stm32f1 centric):

https://aspodata.se/git/openhw/

Currently working on
https://aspodata.se/git/openhw/boards_arm_aspo/stm32f100_styrkort/
which contains pt100 inputs, relay outputs, 0-10V outpus and a "1-wire"
interface; communication via rs-232.

Give me a call if you need more help.
I use:
https://graahnul-grom.github.io/ref-manual/lepton/index.html
http://pcb.geda-project.org/

(you could possible replace pcb with pcb-rnd:
http://pcb.geda-project.org/)

Regards,
/Karl Hammar