Hey All.
I was wondering if someone could help me build a really simple (I think it could be =P) circuit.
Lets say I have a relay (with a com, N/O and a N/C and the relay coil at 12v) and two thermisters, A and B.
I would like to be able to latch over the relay when say the temperature at A is higher than the temp at B.
The application is that A would be installed in a Solar water heater and B in a hot water boiler. I would like the relay to click in only when the water in the Solar water panel measures hotter than the water in the boiler. I'm not worried about how the units will fit into the system as that part I can sort out myself. I possible I would like to make the unit as simple as possible to make it as rugged as it can be. If possible leaving out microprocessors and such. The whole unit I'm running off a 12v battery.
If someone has the heart to maybe make a simple circuit diagram I would love it. I'm rather new to the whole electronics thing but starting to understand it. Thanks much for the help in advance
Gordon

 

Here is a circuit that should basically work for you. But you need to modify it by replacing the top 10k resistor with the other thermistor identical to the first, and make the bottom 10k resistor equal to the nominal thermistor resistance (at 25°C). Which thermistor goes where is determined by whether you use positive or negative temperature coefficient thermistors.

The op amp can be any general purpose op amp.

 

FWIW, this is how I'd handle the comparator and switch. Essentially the same idea, just a different implementation. Same setup of comparing the two thermistors as suggested above.

 

We did one like this on August 4th. Two sensors makes a differential temperature switch.

 

Holy you guys are fast. Thanks for the help. ^^
crutschow I'm not too great with all the values and such at the moment but basically the thermistors will be going anywhere from 0 degrees C to 100 C . would those ones in the diagram be okay for the application?
Another thing I forgot to mention, once the relay has clicked over and lets say the temp in A has now dropped under the temp in B will it click back?

 

A thermistor that has a value of 10kΩ to 20kΩ at 25°C should work fine. Yes, the relay will turn on and off depending upon the relative temperature between the two thermistors.

The temperature difference between ON and OFF is determined by the value of the hysteresis resistor (shown as 270kΩ between the output and the plus op amp input). You may have to change the value to get the difference you want.

 

Okay. here's something.
This is a kit I've got. I think its similar to that frost detection system. besides not been what I'm looking for this little gizmo doesn't switch off when the temp drop or rises above it's set limit either.
But apart from that is it possible to modify this circuit to do what I want?
Behold my powerful paint drawing. I just copied it so its darker and supplied what the circuit looks like on the supplied board. But if possible if theres something I can add to it to do what I want I would love it.
thank again for all your help so far I really appreciate it.
First pic has values and a reference number attached to show where it fits on the second.

 

Use the schematic provided by #12,replacing the potentiometer(Vr1) in the kit w/ the second thermistor.
And as Crutschow pointed out,change the other resistor to provide some hysteresis so the thing don't constantly oscillate,good place to reuse the replaced potentiometer.

 

What do you mean it "doesn't switch off when the temp drop or rises above it's set limit either"? That sounds like it doesn't work at all.

 

..this little gizmo doesn't switch off ...

That 100k base resistor might not be correct. Try a value of 1K-10K. A thermostat that doesn't switch off isn't much of a thermostat.

Are your water pumps AC powered? You need a relay for that, but a small DC pump could be switched directly by the MOSFET circuit I gave you, eliminating the need for a relay.

The circuit from #12 has a nice feature - being a high-side switch, it will fail to "off" if the comparator fails to anything other than a short. The circuit I gave you is a low side switch and will fail to "on" unless the comparator fails to a short. I frankly don't know which comparator failure mode is more common.

 

That 100k base resistor might not be correct. Try a value of 1K-10K.

I did one yesterday that needed 750 ohms. I suggest the 1k approch.

also you'll need a positive feedback resistor to stop it from oscillating. Changing the value of a resistor that is already in the drawing will not fix the oscillating problem.

You should also use a high limit switch to stop pumping water if it is hot enough to boil.
and the LM393 chip is pretty lame. Choose a different amplifier.
While I'm at it, the 741 is pretty lame, too.

ps, I use MSpaint, too.

 

...the LM393 chip is pretty lame. Choose a different amplifier.
While I'm at it, the 741 is pretty lame, too.

No doubt about the 741, but I'm curious what you're getting at regarding the LM393. Low sinking current, maybe? I'm not trying to defend it, but the LM393 is handy because so many folks can run to the local Shack and get one.

 

The 393 chip only promises 6 ma of sink current.
Personally, I have trouble thinking in terms of "sink only" after all the time I've worked with opamps.
"Sink only" is just one more step I have to go through to make sure the logic is correct.

 

The 393 chip only promises 6 ma of sink current.

Yup, that is lame.

 

That 100k base resistor might not be correct. Try a value of 1K-10K. A thermostat that doesn't switch off isn't much of a thermostat.

................

If you lower that base resistor value you may also have to add a resistor (about 1/4 the value of the base resistor) from the transistor base to emitter so the transistor doesn't turn on when the 741 is in it's low output state (about 1.5V above ground).

 

What do you mean it "doesn't switch off when the temp drop or rises above it's set limit either"? That sounds like it doesn't work at all.

Well I set the relay to switch at a certain temp via the var-pot, drop the temp and hear the relay click but raising the temp WAY higher doesnt cause it to go off again. circuit fault or is it just wired that way? Sorry I'm not too bright with this stuff yet. Just trying to do what I can

 

That 100k base resistor might not be correct. Try a value of 1K-10K. A thermostat that doesn't switch off isn't much of a thermostat.

Are your water pumps AC powered? You need a relay for that, but a small DC pump could be switched directly by the MOSFET circuit I gave you, eliminating the need for a relay.

The circuit from #12 has a nice feature - being a high-side switch, it will fail to "off" if the comparator fails to anything other than a short. The circuit I gave you is a low side switch and will fail to "on" unless the comparator fails to a short. I frankly don't know which comparator failure mode is more common.

It's a small 12vDC 10w water pump. The relay is supposed to switch it on via solar power pannels.

 

Well I set the relay to switch at a certain temp via the var-pot, drop the temp and hear the relay click but raising the temp WAY higher doesnt cause it to go off again. circuit fault or is it just wired that way? Sorry I'm not too bright with this stuff yet. Just trying to do what I can

Sounds like the problem I mentioned in my post #15.

As was mentioned, 100kΩ seems much too large for R4. I would change it to about 3kΩ and add a ≈1kΩ resistor from the transistor base to emitter (ground).

 

Sounds like the problem I mentioned in my post #15.

As was mentioned, 100kΩ seems much too large for R4. I would change it to about 3kΩ and add a ≈1kΩ resistor from the transistor base to emitter (ground).

Amusing, considering this is a kit I bought in an electronics shop. nice people supplying sub standard stuff =P

 

It's a small 12vDC 10w water pump. The relay is supposed to switch it on via solar power pannels.

IMHO, I'd ditch the relay altogether and switch the motor directly with a MOSFET. It's less than 1A, so no issue at all for the IRF540N I drew. I use a similar circuit for switching a 5A load and I have no issues. No need for a heat sink on the MOSFET.