enter image description here

I have tried so many different variations on breadboard of this circuit but the motor seems to just be powered by the power rail on the breadboard. It does not react to any change in the input and seems that the whole input is redundant to the MOSFET. (The MOSFET that I am using is WH50 2N7000.)

Essentially, if the thermistor is increased and exceeds the threshold of 25℃, the motor activates. If below the threshold, the motor will not operate. Theoretically it works but on breadboard it does not. Is there any reason for this?

  • \$\begingroup\$ There is no specific ON/OFF in your circuit. You should read about how MOSFETs work first. \$\endgroup\$ – Long Pham Feb 15 '19 at 18:08
  • 1
    \$\begingroup\$ Theoretically is does not work! \$\endgroup\$ – Andy aka Feb 15 '19 at 18:25
  • \$\begingroup\$ @Andyaka: that depends on what theory you use! \$\endgroup\$ – TimWescott Feb 15 '19 at 18:54

This circuit is really a bad design and much too simple to work properly as it will turn on/off the MOSFET only gradually, that is a bad thing as that will cause current to flow through and a voltage to be present across the MOSFET. As P = V * I that means Power is dissipated in the MOSFET. Not a good thing. Even worse with an 2N7000 which cannot handle much power.

What is needed is an abrupt switch on/off of the MOSFET.

Also the 2N7000 by far isn't powerful enough to drive a motor. You should use a MOSFET which has a maximum current rating that is a lot higher than the 2N7000, it can only handle about 200 mA. Much less than what most motors require. Also see this question.

So probably you have already damaged that poor 2N7000. You really need a better circuit.

Here's an example using some CMOS inverters between the resistors and the gate of the MOSFET. As you also don't want the motor switching on/off constantly around the tripping point you should use schmitt-trigger inverters which will add hysteresis.


For the 2 NOT gates I suggest using the HEF40106 IC (or an equivalent one). Connect the inputs of the 4 unused NOT gates (the HEF40106 has 6 NOT gates) to ground.

  • \$\begingroup\$ Using that gate would make for a very imprecise comparison. A real comparator would be much more precise. \$\endgroup\$ – TimWescott Feb 15 '19 at 19:37
  • \$\begingroup\$ Thank you very much for helping, I will try to find an alternative to the 2N7000 that should be suitable to handle the current and make some changes to the circuit. \$\endgroup\$ – P.Patel Feb 15 '19 at 20:04
  • 1
    \$\begingroup\$ @TimWescott You probably need your comment to better explain what you mean if you want it to be useful. "Precise" can relate to the switching point voltages for a given IC or the data sheet min/typ/max voltages. A given 40106 upper and lower hysteresis points are probably "good enough" in this application and the actual-typical values will usually be closer to the spec sheet typicals than to the spec sheet min/maxs. So ... . \$\endgroup\$ – Russell McMahon Feb 15 '19 at 20:31
  • \$\begingroup\$ @TimWescott Yes, this is a "quick and dirty" circuit. A circuit based on an opamp could have more predictable threshold levels but is also a bit more complex and for a beginner not easy to understand let alone dimension properly (getting the resistor ratios correct I mean). So I chose this one which fixes the main issue of the original circuit. Also for "real precise" I'd use the ADC inside a microcontroller. \$\endgroup\$ – Bimpelrekkie Feb 15 '19 at 21:09
  • \$\begingroup\$ It certainly solves the worst problem with the circuit, and as @RussellMcMahon points out, "precise" is in the eye of the beholder. \$\endgroup\$ – TimWescott Feb 15 '19 at 21:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.