Dear all,
I have recently started to build a small project using an 8 channel multiplexer, on a breadboard, with 8 thermistor inputs (nominally 10kohm @25C).
Ultimately I'll want to extend the wiring so that some of the thermistors will be located in various rooms around the house.
Please excuse my lack of a drawing, I am still trying to get to grips with a circuit drawing package. The technique I have used is to feed 5vdc to one end of each thermistor (all 8 have one leg common to 5vdc), the other leg is fed to the 'top' of a 15kohm resistor which completes the circuit to 0vdc. This forms a voltage divider, the output of which goes to one channel of the multiplexer.
This technique works on the breadboard, but will it translate to 'house' scale?
Can you suggest improvements, new techniques, etc.?
Screened cable?
Twisted pair cable?
Regards, Glorymill
PS, I'll be away for a week from about now, but look forward to your suggestions on my return.
GM
It's customary to include a schematic for design questions.
Nobody is asking you to learn schematic capture software.
All we expect is a handdrawn schematic on a blank sheet of printer paper with a black pen, photographed
and uploaded as an attachment using the "Attachments and other options" dropdown menu below.
This technique works on the breadboard, but will it translate to 'house' scale?
This is too vague. In electronics, the length of the sensor signal wire to the circuit that reads it is
critical. (Ohm's Law)
Please post a list of sensor locations and distances to the arduino.
A quick and dirty calibration method is to take your current circuit to each room and read the temperature
and then move the entire circuit except the thermister to where the circuit will be located and connect the
thermister to the controller with a wire (measure the length) and repeat the test for each of the rooms.
At the end of this process you should have a list of temperatures in each room with the controller in the same room and a list of temperatures of each room when the controller is located wherever it will wind up;
with wire lengths for each thermister. Obviously, this must all be done when the temperature of the
first test is the same as the temperature in the second test. If the first test was done in the morning
when it's cooler and the second test in the afternoon when it's warmer the data would be useless.
If you do all the testing at night when the sun is down , and the temperature is constant in the rooms
you can do all the testing before the sun comes up and the data will be valid.
When you have this data we can discuss whether it is necessary to modifiy the circuit.
This wouldn't be necessary if you had A to D converters converting the temp to digital and sending it
to the controller because it would be digital and not resistance. TTLl 5V logic maximum wire length is
about 25 feet but RS232 is probably good to 100 feet and RS485 is about 300 meters I think.
It may be simpler just to record the voltage drop for each room and add that in your code since the
added wire resistance is a constant and likely not effected by minor temperature changes that would
occur in a room.
What temperature range are you wanting to cover?
Have you considered the 18B20 digital One-Wire sensors. They cover the range -55c to +125C. You can hang many sensors on just 3 wires (Vcc, ground, signal) and they take only 1 Arduino pin. Saves running a lot of wire. Long wires pick up more noise. That can be a problem with analog devices, but with digital not so much. And they do not require calibration.
Wow!
Thank you for your suggestions. The DS18B20 seems to be quite up to the job, and will make future system expansion easy to achieve once I have got to grips with the coding knowledge to get the temperature information out of it.
Answers to previous points: The normal working temperature range will be -10 to +75 C, that is the lowest temperature expected when measuring outside temperature, the highest being the stored hot water temperature, but the majority will be +15 C to +25C (room temperatures).
I'm not used to uploading photos, but clearly it's a skill I must master if I'm to make it easy for you to pass on knowledge on the engineering and software side.
But the biggest shock is that these DS18B20 sophisticated sensors are so cheap, in fact cheaper to use than the extra wire needed to connect thermistors!
Unfortunately it means that, once again, I must abandon my old ideas to take advantage of the newer technologies, and learn more advanced software techniques before I can have a working device.
Thanks.
GM
The one wire library and the Dallas temperature library make using the DS18B20 pretty easy.
Ultimately I'll want to extend the wiring so that some of the thermistors will be located in various rooms around the house.
With that in mind and considering the DS18B20 temperature sensor which is the way to go I would think about simple wireless solutions and maybe consider the ESP8266.
ESP8266 DS18B20 Temperature Sensor with Arduino IDE (Single, Multiple, Web Server). All you need is a few inexpensive DS18B20 modules and a few inexpensive ESP8266 modules and the latter is easily programmed using the Arduino IDE. The link being just a single example and a Google should yield a dozen more.
Additionally if you want to know more than just temperature then consider the DHT11 or DHT22 temperature (C and F) as well as relative humidity sensors. Also single wireand like the DS18B20 easy to mate with an ESP8266 module.
Ron