We're missing the power supply here. Based on discussions with Hao, we need a power supply capable of delivering at least 7A for a 100W 12V heating element such as this one. The calculation is: 7A ~ 100W/12V.

Regarding the power needed to warm up the air for a bit (please correct this if I made any mistake), here's a back of the envelope calculation, based on https://www.e-education.psu.edu/earth103/node/1005

Air heat capacity is 700 (j kg^-1 °K^-1).
Volume inside the box: 0.03 (m^3)
Mass of the air inside the box: 0.04 (Kg) based on this: https://www.aqua-calc.com/calculate/volume-to-weight
Starting air temperature: 20 °C == 293 °K
Starting air energy: 0.04 Kg 293 °K 700 j kg^-1 °K^-1 ~ 8200 J
Ending air temperature: 300 °K (about 30°C)
Ending air energy: 0.04 Kg 300 °K 700 j kg^-1 °K^-1 ~ 8400 J
Energy needed: 200J
Heating power is 100W == 100 J s^-1
So it seems we would warm up the air in the box by 10 °C in (200 J) / (100 J s^-1)=2 seconds!

This is assuming that there's no temperature loss to the environment and that we wouldn't be warming up the box itself. In practice, it might take a few minutes to get to that uniform temperature, but still seems quick enough, no?

PS - Let's keep these calculations in a model as well... I'd like to see how close they get to the real thing. They are a very abstract model too :P

clagms commented 4 years ago

Regarding the fan. Calculations based on https://www.cuidevices.com/blog/understanding-airflow-fundamentals-for-proper-dc-fan-selection

Fan airflow: 48 CFM ~ 0.02 m^3 s^-1
Box volume: 0.03 m^3

So it takes (0.03 m^3) / (0.02 m^3 s^-1) ~~ 2 seconds to "theoretically" move all the air within the box. And probably create a tornado as well :P I don't know how uniform the temperature would become with this airflow rate... We will probably have to try out with different fan placements to make sure we get uniform temperature readings (and understands how long it takes for the uniformity to be observed). Or we can just use 4 or 5 temperature sensors around the box. With this much equipament, there won't be space for yogurt :D

The fan I used as ref was this one: https://www.banggood.com/90x90x25mm-12V-4Pin-Computer-PC-CPU-Silent-Cooling-Cooler-Case-Fan-p-975995.html?rmmds=search&cur_warehouse=CN

clagms commented 4 years ago

Assuming that we don't need analog ports (do we?), then I think the raspberry pi will be the best here. We can use python to code it, and it can easily be used to transmit the readings over the web. It can also be coded to received new parameters and the like from the web.

Alternatively, we can use an arduino mega. It has plenty of ports for temperature sensors and all that. https://store.arduino.cc/arduino-mega-2560-rev3 But coding it can get messy if we have to record data over the wire.

clagms commented 4 years ago

For temperature sensors, I think the LM35 is a good one. It operates through the digital ports. https://www.banggood.com/LM35DZ-TO-92-LM35-Precision-Centigrade-Temperature-Sensor-p-917706.html?rmmds=search&cur_warehouse=CN

We have to be careful with electrical noise (like running it close to the fan and/or heater or having the wires along the higher voltage wires) and/or connecting the sensor through long wires (because long wires introduce resistance, so you need some extra circuitry to compensate for this... i don't know what that extra circuitry is)

EDIT: The one Casper proposed is also good! Let's stick to that one then. I hadn't seen it :P

clagms commented 4 years ago

We also need relays, to turn on the heater. I assume we can just let the fan always run. https://www.banggood.com/3Pcs-5V-Relay-5-12V-TTL-Signal-1-Channel-Module-High-Level-Expansion-Board-p-1178211.html?rmmds=myorder&cur_warehouse=CN

haof-github commented 4 years ago

We're missing the power supply here. Based on discussions with Hao, we need a power supply capable of delivering at least 7A for a 100W 12V heating element such as this one. The calculation is: 7A ~ 100W/12V.

Regarding the power needed to warm up the air for a bit (please correct this if I made any mistake), here's a back of the envelope calculation, based on https://www.e-education.psu.edu/earth103/node/1005

Air heat capacity is 700 (j kg^-1 °K^-1).

Volume inside the box: 0.03 (m^3)

Mass of the air inside the box: 0.04 (Kg) based on this: https://www.aqua-calc.com/calculate/volume-to-weight

Starting air temperature: 20 °C == 293 °K

Starting air energy: 0.04 Kg 293 °K 700 j kg^-1 °K^-1 ~ 8200 J

Ending air temperature: 300 °K (about 30°C)

Ending air energy: 0.04 Kg 300 °K 700 j kg^-1 °K^-1 ~ 8400 J

Energy needed: 200J

Heating power is 100W == 100 J s^-1

So it seems we would warm up the air in the box by 10 °C in (200 J) / (100 J s^-1)=2 seconds!

This is assuming that there's no temperature loss to the environment and that we wouldn't be warming up the box itself. In practice, it might take a few minutes to get to that uniform temperature, but still seems quick enough, no?

PS - Let's keep these calculations in a model as well... I'd like to see how close they get to the real thing. They are a very abstract model too :P

The difference in celsius is the same in kelvin, So it should be Q=(0.04 Kg) (700 j kg^-1 °K^-1) (30-20) = 280 J. Here we use the air heat capacity as 700, however it changes according to pressure and temperature.

Conclusion: I think 100W power supply is good enough for the incubator.

INTO-CPS-Association / example_digital-twin_incubator

Hardware #6

Box:

Fan:

Heater:

Controller:

Temperature Sensor:

Power Supply

Relay

Optocouplers

Mosfet