Seems like there are two options: either a sensor that translates temperature to a 0 to 5V range, or one that translates temperature to pulses in some way (perhaps varying the time between pulses in proportion to temperature) -- plus, then, changes on the software side to translate the inputs into values.
From what I've read elsewhere on the site, I'm guessing neither the hardware nor the software are available yet.
gebhardm |
No current solution; the FLM's software does not support temperature measurement by now, but with the Jeenode capabilities and the MQTT interfacing it opens up a world of possibilities - feel free to contribute as this is what open source is all about (if you have spare time, of course)
AceNZ |
Sounds like a good reason to get the v2b instead of the v2...
Fluc |
Sorry to throw this topic up, but the latest week i had a idea to made a electronic project to monitor also the temperature of my living room, so i can follow the heating doing its work.
This will be shown on the same graph than my gas consumption. It is not ideal but the °C will be seen as second blue line and the unit who i can't change, stays in L/min or Litre.
Because my gas usage is between 15 and 25 litres that is also ideal to show the same time the temperature between 15 and 25 °C (but in litres of course on the X-axis).
I made a test with 50 Hz pulses that shows exactly 30 litre (30°C) on the graph, the meter-constant i used is 0,01. So 33 Hz is 20 °C and so on ... it stays linear.
I shall use a temperature to voltage chip and a voltage to frequency converter.
The 5 V power supply i shall retrieve from my nearby Vdsl2 modem's usb port.
The project itself is not started yet, because the parts are not arrived yet. I’m waiting on it :-)
I hope it shall working like expected, and some mistakes are always possible...
Fluc |
Update:
Here is the datasheet of the LM35 Precision Centigrade Temperature Sensor : http://www.ece.usu.edu/ece_store/spec/lm35dt-3p.pdf
On page 9, figure 16 is the schema i shall use for the project. One difference is that on the output of the LM35, i must use a divider by 6 to have a rounded meter constant, otherwise is must be 0.0016667, and now i can use 0.01. For this i use 2 or 3 resistors.
Another small difference is the use of a resistance of 820 ohm in the place of 1 kohm on the opto-coupler anode becuase the supply i use is from Usb = 5 Vdc, not 6 Vdc. And the use of a LM331 in the place of the LM131 that is not available any more.
For those who use Fahrenheit as temperature unit, use the LM34.
Fluc |
So, prototyping is ready and working after a trimming on the gain.
I see that the signal is not very clean, the Temp/Freq converter unit is sensitive, so i must put perhaps a filter on it. Also the wires to the FLM are now loose and perhaps the grid frequency is interfering, so i need to use shielded wires.
For the moment, only the liter and liter/minute views are usable in the hour and day tab and i can see the temperature in real °C.
Al the electronics are put temporary on a kind of breadboard. Later i have to use a real pcb board.
See the graph below that’s figure on the same screen than the gas usage. I'm so happy that this is finally working :) http://ubuntuone.com/6kJVxk8DSsW4jDYuF8P2tr
Fluc |
Temperature monitoring working flawlessly for almost two weeks. Last weekend i put al the components to a pcb board and trimming a little the gain adjust. Now it stays where it is, that is nearby the FLM to have short connections.
I am just wondering why nobody placed a reaction to comment this here on the forum ?
It 's not rocket science ...and it is not perfect, but it works ;-)
jgysenbergs |
Most interesting. Would you mind posting a schematic and/or a jpg of your breadboard? Thanks!
Same here, I would like that. Connecting even one sensor - e.g. the one at the top of the boiler - would be very useful. Maybe some temperature sensor like this one?
https://www.tindie.com/products/BBTech/tmp006-contactless-infrared-tempe...
Seems like there are two options: either a sensor that translates temperature to a 0 to 5V range, or one that translates temperature to pulses in some way (perhaps varying the time between pulses in proportion to temperature) -- plus, then, changes on the software side to translate the inputs into values.
From what I've read elsewhere on the site, I'm guessing neither the hardware nor the software are available yet.
No current solution; the FLM's software does not support temperature measurement by now, but with the Jeenode capabilities and the MQTT interfacing it opens up a world of possibilities - feel free to contribute as this is what open source is all about (if you have spare time, of course)
Sounds like a good reason to get the v2b instead of the v2...
Sorry to throw this topic up, but the latest week i had a idea to made a electronic project to monitor also the temperature of my living room, so i can follow the heating doing its work.
This will be shown on the same graph than my gas consumption. It is not ideal but the °C will be seen as second blue line and the unit who i can't change, stays in L/min or Litre.
Because my gas usage is between 15 and 25 litres that is also ideal to show the same time the temperature between 15 and 25 °C (but in litres of course on the X-axis).
I made a test with 50 Hz pulses that shows exactly 30 litre (30°C) on the graph, the meter-constant i used is 0,01. So 33 Hz is 20 °C and so on ... it stays linear.
I shall use a temperature to voltage chip and a voltage to frequency converter.
The 5 V power supply i shall retrieve from my nearby Vdsl2 modem's usb port.
The project itself is not started yet, because the parts are not arrived yet. I’m waiting on it :-)
I hope it shall working like expected, and some mistakes are always possible...
Update:
Here is the datasheet of the LM35 Precision Centigrade Temperature Sensor :
http://www.ece.usu.edu/ece_store/spec/lm35dt-3p.pdf
On page 9, figure 16 is the schema i shall use for the project. One difference is that on the output of the LM35, i must use a divider by 6 to have a rounded meter constant, otherwise is must be 0.0016667, and now i can use 0.01. For this i use 2 or 3 resistors.
Another small difference is the use of a resistance of 820 ohm in the place of 1 kohm on the opto-coupler anode becuase the supply i use is from Usb = 5 Vdc, not 6 Vdc. And the use of a LM331 in the place of the LM131 that is not available any more.
For those who use Fahrenheit as temperature unit, use the LM34.
So, prototyping is ready and working after a trimming on the gain.
I see that the signal is not very clean, the Temp/Freq converter unit is sensitive, so i must put perhaps a filter on it. Also the wires to the FLM are now loose and perhaps the grid frequency is interfering, so i need to use shielded wires.
For the moment, only the liter and liter/minute views are usable in the hour and day tab and i can see the temperature in real °C.
Al the electronics are put temporary on a kind of breadboard. Later i have to use a real pcb board.
See the graph below that’s figure on the same screen than the gas usage. I'm so happy that this is finally working :)
http://ubuntuone.com/6kJVxk8DSsW4jDYuF8P2tr
Temperature monitoring working flawlessly for almost two weeks. Last weekend i put al the components to a pcb board and trimming a little the gain adjust. Now it stays where it is, that is nearby the FLM to have short connections.
I am just wondering why nobody placed a reaction to comment this here on the forum ?
It 's not rocket science ...and it is not perfect, but it works ;-)
Most interesting. Would you mind posting a schematic and/or a jpg of your breadboard? Thanks!
Ok, here comes a picture of the pcb board with comments on it:
http://ubuntuone.com/65KvR8MbPhwhFmoSiXvzbr
The schematic of the pcb board at scale 4/1:
http://ubuntuone.com/4OILFH2lvB2iRqTl04FYcp
and the LM35:
http://ubuntuone.com/3OwuzP8kf9jqhEPIctzqtH
The datasheet of the LM35 where you can find the schematic is used on page 6:
http://www.alldatasheet.com/datasheet-pdf/pdf/8866/NSC/LM35.html
The datasheet of the LM331 where you find everything to know to build well this Temp to Freq converter:
http://www.alldatasheet.com/datasheet-pdf/pdf/8641/NSC/LM331.html
Use metalfilm resistors of 1% if possible. The Vout of the LM35 must be divided by 6 to obtain 33,33mV in the place of 200mV at 20°C room temperature. So you can measure at pin 3 of the LM331 the 33,33Hz pulses. Only measurable with a scope-meter or good multimeter like the Fluke 85 or newer.
On the board there is also place to feed the optical watersensor with 3,6V and a 3mm red led who runs at 4mA.
Success for those who will build this device :)
Sorry, on the LM35 datasheet the schematic is figure 16 on page 9, not page 6 !
I write this also a few comments higher...
Sorry, some links to the pictures do not work anymore.
Ask me for it and i shall provide these for you.