I want to make a circuit using the following components:

* 9 V battery
* TL064 op amp (as a Comparator)
* NTC-Thermistor 10 kΩ
* LED, forward biased at 2.7 V
* Resistors (I have lots so can pretty much use anyone)

I want to build the circuit so that heating up the NTC-Thermistor with two fingers (which make it go to about 7.14 kΩ) will activate the TL064 op amp and make it switch and make the light of a LED start to shine.

I need help with the Comparator, i.e. how should I connect to make it behave the way I intend?

Here's a picture of the TL064 op amp, where you can see the connections:

Which connections should I connect into the circut?

Also attached: Data sheet for TL064: http://www.edaboard.com/attachments/98727d1384799087-nrtl061-tl064_e.pdf

Help is appreciated

 

Because of the link and the fact you are a new user on AAC this thread was auto-moderated. I have approved it.

 

Look up "op-amp comparator" in basic op-amp circuits. You'll find plenty of examples.

It seems simple but I always confuse myself with the logic; which pins on the op-amp connect to which thermistor pin and how the LED should behave. You want to set it up so that hot=LED on. Murphy's law almost guarantees you will get cold=ON at least once in your set up!

 

Is this a Homework question?

 

Been there, done that, look up LM723 to see how the connections are made to the comparator inside the chip.

http://forum.allaboutcircuits.com/blog_attachment.php?attachmentid=284&d=1361067695

You can eliminate a lot of the components for your circuit.

 

Here is the way I would do it:

Has some hysteresis, trips at Rt=~8.2K

 

Same idea with a lot less unnecessary complication than my thermostat circuit.
Using a Wheatstone bridge makes good regulation of the voltage reference (provided in the LM723 chip) unnecessary, too. It's just that I had one laying around and posted that because it was easy.

 

Here is the way I would do it:

Me too, with LM339 as the comparator. It needs an external transistor if switching anything more than an LED.

 

Here is the way I would do it:

Has some hysteresis, trips at Rt=~8.2K

OK, I'm stumped. I can't quite see how you got the resistor to be proportional to the voltage. I see the R=V(t) but where is V(t)? Is it related to the little squiggle on top of the V2 output that I can't resolve?

 

OK, I'm stumped. I can't quite see how you got the resistor to be proportional to the voltage. I see the R=V(t) but where is V(t)? Is it related to the little squiggle on top of the V2 output that I can't resolve?

Yep, LTSpice has that nice feature where a resistor can be made a function of lots of things, including a node voltage. I created the artificial voltage source connected to node t (laying on its side) which is a PWL function of two seconds of sim time which is proportional to the desired resistance.

This shows the hysteretic trip points as a function of resistance...

 

Yep, LTSpice has that nice feature where a resistor can be made a function of lots of things, including a node voltage. I created the artificial voltage source connected to node t (laying on its side)............

Thanks. Learned something new today. It was the t laying on its side that I couldn't make out.

It's also neat that dividing the voltage across the resistor by its current automatically gives a vertical scale of ohms on the plot for the resistance of the resistor.

 

Thank you for your replies. If you could help me with another thing, that would be great. I want to calculate how much power is disspated in the comparator. How would I do this?

 

Thank you for your replies. If you could help me with another thing, that would be great. I want to calculate how much power is disspated in the comparator. How would I do this?

You calculate the supply current for all the resistors and op amp when the output is high and multiply that by the supply current to get power.

Then do the same when the output is low.

Then you multiply the first calculation by the percentage of the average time period the output will be high.

Then you multiply the second calculation by the percentage of the average time time period the output will be low.

Then you add those two values together to get the total average power over the total time period.

 

And then you realize it's such a small number that it really doesn't matter and you don't do it again.

Just kidding, sort of.

 

If this wasn't running on a 9V battery, I would completely agree with wayneh, but 9V batteries are rather weak and the energy budget might actually matter in this case.