Help with thermal fan controller

Thread Starter

oran0007

Joined Sep 1, 2013
4
Firstly, I did do a search (http://forum.allaboutcircuits.com/showthread.php?t=69524&highlight=thermal+fan+controller) and my question was not addressed in the results.

I am building a cooling unit for my entertainment center and have run into a problem (potentially).

http://www.sentex.ca/~mec1995/circ/auto-fan.html is the source of my circuit and I have followed it as closely as possible (the one amp relays I got from Radio Shack do not fit the breadboard properly). I am using 5A relays instead. My plan is to have two sets of two fans controlled by separate thermistors (and a single pot if possible) to cool two zones. I will make a probe out of each thermistor to ensure that each zone is only cooled when necessary.

I have the circuit working to turn on a single fan wired directly to the breadboard, but it never turns off. I don't know if there is a problem or if the threshold to turn on is simply too low. My house is still over 84° F at 11 p.m. so it may be that, but is there a way I can know for sure without simply waiting for cooler weather?

I am a computer science student, not an engineer and this is my first attempt at making an electronic device, so I really don't know this stuff. I would imagine there is a formula I can apply as long as I can get the information about my pot and thermistor.

If needed, I was thinking of adding the requirement that our googletv blu-ray player be powered on. Is adding a third relay which would affect the presence of voltage to the rest of the circuit the best way to accomplish this?

By the way, I am going analog for now, but eventually I will probably build this into a digital device.

Also, I have a couple of questions about best practices.
1. I set up an excel file to plan my breadboard; is this okay?
2. Should I find a way to test the 1A relays before using them? There is continuity between COM and NC and with power applied across the coil it does audibly switch, but I have not devised a way so far to check the NO pin with power applied (haven't spent much time worrying about that).
 

MaxHeadRoom

Joined Jul 18, 2013
22,010
I have built similar circuits and would have tended to use a comparator such as LM311.
Obviously applying some cooling agent to the thermistor should allow the circuit to be tested for switching action at least?
You can buy a can of 'Arctic Air' If necessary.
Max.
 

Thread Starter

oran0007

Joined Sep 1, 2013
4
I have built similar circuits and would have tended to use a comparator such as LM311.
Obviously applying some cooling agent to the thermistor should allow the circuit to be tested for switching action at least?
You can buy a can of 'Arctic Air' If necessary.
Max.
Wow. I feel like an idiot. I have air dusters and they should work, of course they wont help me know whether I have the pot set properly, but I should be able to address that later, especially if I just leave it on the breadboard for now. Then again, maybe I can borrow a lab at school if they have temperature controlled cases or something. I'll have to talk to someone in the engineering dept.

Thanks for the answer.
 

wayneh

Joined Sep 9, 2010
16,513
I think we've seen this before. That old op-amp (there are none older) cannot pull the transistor base low enough to deactivate the relay.

A comparator would be better, and I'd also just use a MOSFET switch. That would eliminate the transistor with its biasing resistors, and the relay. Much simpler and cheaper.
 

Thread Starter

oran0007

Joined Sep 1, 2013
4
I think we've seen this before. That old op-amp (there are none older) cannot pull the transistor base low enough to deactivate the relay.

A comparator would be better, and I'd also just use a MOSFET switch. That would eliminate the transistor with its biasing resistors, and the relay. Much simpler and cheaper.
So I should switch out the lm741 for an lm311. I can do that I guess. Now on switching out the transistor and relay for the MOSFET, is it necessary? I do already have the other items, so it wouldn't be cheaper to switch. ;)
 

wayneh

Joined Sep 9, 2010
16,513
It's not necessary, unless you have continued problems turning off the relay. I think you'll be fine once you switch to using a comparator. The comparator might be able to switch the relay directly without needing the transistor. You'll need to read the data sheets.

Note that you may want to look at the LM339 quad comparator, if you need multiple comparators. It has a lower current capability, though. You'll definitely need a transistor for the output.
 

anotheruser1

Joined Dec 6, 2011
30
I have done several of these and i used a comparator and a voltage divider so i could use a trim pot to offset the temperatures for on and off to wherever i need them. I used lm393 for my experiment
 

wayneh

Joined Sep 9, 2010
16,513
We're talking about thermostatic control here, on or off. One thing you might like to consider is a proportional controller, so the fans come on sooner but at a low level and only rev up if they need to. It's more complicated, but there are a LOT of example circuits out there.
 

Thread Starter

oran0007

Joined Sep 1, 2013
4
I have done several of these and i used a comparator and a voltage divider so i could use a trim pot to offset the temperatures for on and off to wherever i need them. I used lm393 for my experiment
I considered that, but it seemed a little ambitious for a first project. As I will likely replace with a digital one, I don't know that I will try this, but I'll keep it in mind.

We're talking about thermostatic control here, on or off. One thing you might like to consider is a proportional controller, so the fans come on sooner but at a low level and only rev up if they need to. It's more complicated, but there are a LOT of example circuits out there.
This I thought about also, but I bought four 120mm 3-pin cpu fans, so maybe next time.
 

wayneh

Joined Sep 9, 2010
16,513
I considered that, but it seemed a little ambitious for a first project.
Give it some more thought. It may be a bit intimidating but I don't think you will regret this strategy if you choose it. Comparators are very useful in many circuits and easy to use once you get past the brief learning curve.
 

THE_RB

Joined Feb 11, 2008
5,438
We're talking about thermostatic control here, on or off. One thing you might like to consider is a proportional controller, so the fans come on sooner but at a low level and only rev up if they need to. It's more complicated, but there are a LOT of example circuits out there.
I like that idea. Maybe it doesn't need to be more complicated?

Just using a darlington transistor, it will not turn on at all until its Vbe is >1v, then a small "linear" or "ramp" zone can be added by using an emitter resistor.

Total parts could be as low as thermistor:trimpot divider (to base), a cap between base to gnd, an emitter resistor to gnd, and the fan connected between +12v and collector.
 
Top