I am tasked with designing a TEC controller to run three thermoelectric modules used for heating orcooling some plates to around 70F or so from ambient air temp ranging from 40-100F. The power will come from solar panels connected to some deep cycle batteries. Does anyone have any experience with this sort of thing? I am a mechanical engineer and I don't know much of anything about ECE and controls. I have been doing quite a lot of research the past couple of days but need some help.
I would like to have 4 modes - OFF, Heat only, Cool only, Heat and Cool. The Heat and Cool option would be used when the air temp varies so that cooling might be needed some times and heating at other times. The TEC controller doesnt necessarily need to have bipolar control if a separate resistive heater is used.
Well, I really can't cookbook a design for you, as it would take a day or so to fully engineer something like that.
Depending on your budget, a Arduino plus some RTD's and relays would do everything very cheaply but writing the software would be a pain for a beginner.
The easy button would be to purchase a process controller like Red Lion and program it with the built in software. The Red Lion can account for hysteresis and more or less lock down a stable temperature.
Peltier modules typically consume a lot of current and expect a minimum of five amps each .. I don't think the red Lion can drive the peltiers directly and you would need a relay to switch the current.
Perhaps start off googling PID and download some manuals to get started.
Temperature control is not a trivial task because thermal mass often caused hysteresis (overshoot/undershoot). A PID will make life simpler but they run between $100 - $500 or more. You will also need power supplies, RTDs and relays.
Automation Direct has generic industrial supplies including process controllers and are the cheapest source I have found. The only drawback to cheap stuff is the lack of tech support.
Or a pair of $5 eBay temperature controllers, one set for heat and one set to cool. Add a DPDT relay for one of them, with the common terminals wired to the peltiers, the "a" terminals wired to the battery one polarity and the "b" contact wired through another double pole relay to the battery with the polarity swapped. Control the second relay with the other controller.
With neither controller energized the peltier is deenergized. Either controller will send power, and each with an opposite polarity, but thanks to the dpdt relay they can't both send power at the same time.
http://m.ebay.com/itm/121034195696?nav=SEARCH
plus
http://m.ebay.com/itm/261591716668?nav=SEARCH
Two of each, $20 total plus a couple feet of wire.
Similar results from a home thermostat running the same relays from the heat and ac outputs, if you have the room.
"solar panels connected to some deep cycle batteries"
Rigid installation in a stand alone environment or fed by the grid. Depending on the size of the peltier and the amount of heat you are trying to add or subtract form the environment they can draw a significant amount of power.
Dirt cheap thermostat should do what you are asking in terms of control.
An on/off thermostat was initially considered and would have been super cheap super simple. The problem is that Peltier modules do not like that sort of thermal cycling as it causes large thermal stresses within the module. When doing simple on/off control a cycle time of at least ten minutes is required. I did a quick thermal transient analysis of my system and I think the cycle time will be around 5-7 minutes, so I don't want to go that route. I don't want these failing every several months.
What I think I am going to do is build a simple circuit built around a simple cheapish PWM chip and throw in a filter to help eliminate voltage ripple, which reduces Peltier efficiency. I am thinking I can be clever with relays and a simple on/off thermostat to switch between heating and cooling modes by changing the polarity of the peltier module.
A lot of the cheap PID controllers you can find online do PWM but at too low of a frequency which also can cause premature failure. I probably dont need full PID control so I will do something like just integral control.
A solution was developed using an Arduino but my task is to make the system simpler.
