I need to put together a basic heater with the simplest feedback, fridge-like: switch on when the temp is below setpoint by 0.1C, switch off when the temp above setpoint by 0.1C. No PID, ramp control, Peltiers etc., I need to maintai temp 10-15C above ambient, with 0.1-0.2C resolution, and, if it develops oscillations, I would like for the oscillations to stay within 0.2C. To give some numbers, to heat the object 10C above ambient, at equilibrium, I apply 5V to the heater (an Omega kapton pad 1/2 by 2 in), drawing about 0.1 amps. I have looked through the forum... and I was hoping to avoid using MOSFET and to avoid this turning into an epic project. May be there is even an Arduino kit for this? The object is relatively well insulated, and there is an independent (water based) system controlling the temperature from the other side of the object, there are no major ambient temperature fluctuations short term (hour time scale). I could even avoid using feedback, but since ambient varies by 3-4 degrees long-term (day to day), so I was hoping to implement a basic feedback.
If its ambient varies from 3 to 4 ° C and its expectation is 10 to 15 ° C,
it would mean that its tolerance would be 5 ° C, (15 - 10 = 5)
which is greater than the variation of the ambient temperature.
If its ambient varies from 3 to 4 ° C and its expectation is 10 to 15 ° C,
it would mean that its tolerance would be 5 ° C, (15 - 10 = 5)
which is greater than the variation of the ambient temperature.
Thank you for writing! I would need to adjust and watch the voltage each time I use the unit to compensate for changing initial baseline conditions.
BanditDave:
Rather than a microcontroller I suggest a self contained temperature sensor such as
Cost ~$5.00
Thank you for the suggestion! This self contained unit does sound attractive. So, this acts as a relay, and I use it to interrupt my 5V line, correct? The temperature probe is a bit large, by the look of it, the mass of the probe is about 10% of the mass of my object, but it should work, I think. I can also see if they make units with smaller probes.
You don't wanna know, to me anything about digital electronics that's a step way from plug and play turns epic. The self-contained switch with a temperature display, is based around the same type of transistor, correct? I am tempted to try the self-contained one, b.c. I also get the temperature display. I just do not like the "enormous" size of the probe (ideally I want a tiny thermistor, say 1/8").
There's the important specification. 0.1A is 100mA. Each Arduino is specified to output 20mA although you can usually use 40mA with no problem. So an Arduino pin cannot drive the heater directly. You must have a MOSFET or other transistor.
"Aha!" you say, "I'll use five outputs and add them up!" No you can't do that either. There is a maximum current for the entire chip. You cannot run 5 Arduino pins at 20mA for every pin. You must have a MOSFET or other transistor.
0.1A is considered to be small for a power transistor. A surface-mount transistor for this current would be about the size of a breadcrumb. Or just get a bigger one so you have something to hold onto.
You may well find that an on/off thermostat ( switch ) type approach will not give the control you want .
For example if your heater is ON and the setpoint is reached and the heater then turns off. The heater parts will still be hot and the temperature will continue to rise , until it cools. When it then cools the heater will turn on , but will take a while to heat up , meanwhile the temperature continues to fall .
Something to think about is that some of the older fridges had small heaters inside , some room thermostats have heaters too -why ?
You may well need PID to control this .
Ponder also that the rate of temperture rise is dependant upon ambient temperature- so the overshoot and hence the average controlled temperature depends on the outside temperature . Keeping tight temperature control is not easy