I'm trying to gain a deeper understanding of DC stepper motor because I'm using them for an arduino project.

I was wondering about why the common lead of unipolar stepper motors is usually (always?) connected to the high level voltage instead of ground.

Connecting it to ground should be the most logical thing to do, however, it is not the case. Of course, both cases should work fine for the motor, it just changes the controlling sequence, so that's not a big thing. But is there some advantages, some hidden motives to connect it to the high level voltage?

For reference: I'm using a 28byj-48 stepper motor (5 wires)


2 Answers 2


Silicon switching devices intended for low side switching (ie, the ground side) such as N-FETs and NPN transistors utilize bucket brigade displacement of free electrons in the silicon as charge carriers. Conversely, the simple form of high side switches use positive charge carriers - the so called "electron holes" which are less mobile than actual electrons.

N channel devices work so much better as switches, that in bipolar systems which need to switch both sides, above moderate power levels, it's often chosen to build more complicated level translation circuitry to boost the control signal and allow using N channel switches on the high side as well, rather than lower performing P channel devices.

But for a simple unipolar setup, you only need to switch one side, so you chose to switch the low side where high-performing N channel devices can be used directly.

  • 2
    \$\begingroup\$ Additionally, switching 'to ground' requires no extra to switch a higher voltage than the controlling (logic) voltage. \$\endgroup\$ Sep 26, 2018 at 18:00
  • \$\begingroup\$ Does this apply when I use a ULN2003AN to drive the motor as well? I mean, whether the coil is connected to +5V or to ground, the buffer (which is actually a darlington) is still the same and can switch either to ground or to +5V. So I don't see high side or low side here \$\endgroup\$
    – Tripola
    Sep 29, 2018 at 20:42
  • \$\begingroup\$ There's a low side as soon as you consider the polarity of the control signal in relation to the controlled power. Using a low side switch on the high side requires floating level translation, something that would not really be achievable for a multi-channel part, unless it were built inside for each channel independently. \$\endgroup\$ Sep 29, 2018 at 21:16

R_on and switching time (for the same silicon area) is for N-MOS lower (better). Hence, for the same performance N-MOS are cheaper.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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