Hi, I´m looking for a circuit to control a ventilation fan. Basically my problem is the following : I have a storeroom which is hot in the summer and cold in the winter. The enviroment could be improved if a circuit compared the outdoor and indoor temperature and activated a fan everytime there was a diference in temperature.With the change in the seasons it would only be a question of using a selector to change the indoor and outdoor temperature sensors.The fan is a AC window pane type at one end of the room, and a grill vent at the other end.
Any sugestions ?Thanks.

 

I think your answer lies somewhere here
http://pcbheaven.com/handbook/archive/desc.php

 

Thanks for the tip Kammenos, but all I managed to find were circuits which monitored only one temperature, some had the interesting characteristic of varying the fan speed but I found none that suit my need.

 

Seems to me that what you need is simply a detection of ranges; if the air temperature outside is more desireable than the temperature inside, then run the fan.

Let's say the inside target range was 68°F to 78°F, with the ideal being 73°F.

If it were 65°F inside and 75°F outside, the fan would run until it was 68°F inside, and cut off.

If it were 80°F inside and 65°F outside, the fan would run until it was 78°F inside and cut off.

You could help it to feel cooler in the summertime by having ceiling fans that would switch on if the temperature was higher than 76°F during working hours.

 

Odd coincidence: I was looking at a similar problem this afternoon while contemplating vivaria for exotic frogs...

Variable A: Temp inside too hot, yes =1 no =0
Variable B: Temp inside too cold, yes =1 no =0
Variable C: Temp outside warmer than temp inside, yes =1, no =0
Variable D: Temp outside cooler than temp inside, yes =1, no =0
Variable E: Vent open (or, in this case, fan running), yes =1, no =0

E = A*D + B*C

In addition to two temperature sensors, I was thinking this could be achieved with either a quad comparator and some logic gates, or with a microprocessor featuring at least two ADC inputs. I'd rather solder than program (personal preference) so used the former solution when doodling on the napkin.

I left the napkin at work. I'll bring it home tomorrow and draw something up for presentation here.

 

Not odd that you mentioned quad comparators and logic gates; I was heading that direction myself.

A uC would certainly be the low parts count option. An 8-pin PIC12F675 would have more than enough horsepower to read a couple of voltages from a couple of thermistor/resistor junctions, and decide whether a fan should be switched on or off via a relay. You'd even have some pins left over in case you wanted to send the temperatures to a serial LCD.

 

Make a window comparator, or maybe more than one, use the outputs to drive a speed control.

 

Here's a simple back-of-an-envelope idea that should work with a bit of tweaking. Replace one of the lower 10K resistors with a fixed 5k and multi-turn 10k in series so that the balance can be adjusted:

 

Here's the one I had in mind:



Temperature sensors would have to produce an increasing positive voltage with increasing temperature. Output would be positive when the fan needed to turn on.

"A" comparator o/p is high when inside temp exceeds upper setpoint. "C" comparator o/p is high when inside temp exceeds outside temp.

"B" comparator o/p is high when inside temp is below lower setpoint. "D" comparator o/p is high when outside temp exceeds inside temp.

If inside temp is too high AND inside temp exceeds outside temp, OR if inside temp is too low AND inside temp is below outside temp, then output is high.

 

I was expecting a simple “off the shelf” circuit with simple characteristics but I guess the end product will be a more refined design, variable fan speed is not essential, as the fan I have is not very noisy, but for the sake of evolution I am open to suggestions. I do share thingmaker3`s preference of soldering over programming, something that is not in my domain.
If SgtWookie`s specifications are taken into account we have the makings of a commercial product adequate for the vivaria, a situation ripe for royalty`s dispute(joking)

 

I think your eventual solution will be somewhere between Blocco's and Thingmaker's design. Blocco's design will drive a relay, but has quite a few discrete components. Thingmaker's design uses comparators (which reduces component count while increasing stability and reliability), but has a MOSFET output, which is not suitable for driving an AC fan.

It is an interesting and worthwhile problem though.

 

I'm also wondering if component count can be reduced by using NOR gates and reversing the inputs on all four of the comparators in my schematic.