Hi, I constructed a circuit similar to that in Forrest Mims' thermistor opamp trigger on page 37 of science and communications circuits (vol 2) in case someone has it lying around. basically it's non-inverting comparator with a 100k pot between +in and ground and a 10k thermistor between V+ and -in in a voltage divider with a 10k resistor to ground.

I'm getting quite a bit of output current when i would think it should be off. Anyone have any insight? The output current is like 1/2-2/3 when it's 'off'. I can post a picture if that helps.

I wanted to use this to switch a MOSFET to turn a fan on in an external hard drive enclosure when the temperature passes a certain point.

 

Can you post a schematic? A hand-drawn copy would not infringe on Mr. Mim's intellectual property rights.

 

Here is a copy of the circuit. I have tried it in simulation but I can't get the LED to light at all.

 

Have you looked at the output signal on a scope? You may be getting oscillation near the temperature switching threshold. Adding some hysteresis to the circuit should help considerably.

Additionally, what you are trying to do is really a job for a comparator, rather than on op-amp. Try replacing the 741 with an LM311. You may (probably will) need to add some hysteresis to prevent oscillation.

 

Oh! VR1 is connected improperly...

The way you have it wired, the (-) input will never exceed the (+) input. Sorry, I should have caught that first!

 

A few more hints:

1. Consider grounding your thermistor and swapping the +/- inputs to the amp. This will allow you to use shielded wire to your thermistor if you intend to make a long temperature probe. It will also help to prevent letting the smoke out if your probe accidentally touches the electrical ground.

2. Rather than VR1 swinging from rail-to-rail (which is what you should have when you hook it up right), consider the addition of two fixed resistors, one on each end of the pot. This will *tremendously* improve the setability of the pot, and limit drift due to temperature and mechanical shock. If you can tell me the *exact* thermistor part number you are using, and the desired range of set points (min temp and max temp), I can tell you what resistors to use (in either 5% or 1% values -- your choice).

Have fun!

 

I have a very similar circuit made with a 723 chip. Basically an op amp with a zener voltage reference. My circuit demonstrates the grounded shield sensor and the padded potentiometer.

 

I have a very similar circuit made with a 723 chip. Basically an op amp with a zener voltage reference. My circuit demonstrates the grounded shield sensor and the padded potentiometer.

Is that hysteresis you are adding with the 4001 and 680K?

 

Yes.......

 

Edward's circuit is like mine. Why would grounding the thermistor and swapping the inputs allow me to use shielded wire to the thermistor?
I'm using a TC503 thermistor and would like the fan to trigger at 115-120F.

Just realized that it's a 100K thermistor. I couldn't find a data sheet and someone posted incorrectly in a forum that's 10K. I'll rebuild the circuit for the 100K and see what happens now.

 

Why would grounding the thermistor and swapping the inputs allow me to use shielded wire to the thermistor?

Because that's the whole point of arranging the circuit to use a grounded shield. If you arrange the circuit to have the thermistor in the non-grounded position, you end up using a twisted, shielded pair to keep noise pickup to a minimum. Two wires and a shield is 3 wires. Cheaper and less noisy to use 1 wire and a shield.