2
\$\begingroup\$

The two options:

I have been scrolling through forum after forum for a unilateral agreement on what the best motor placement would be within the half-bridge driver circuit: Putting the motor across the high side MOSFET or, across the low side MOSFET. As an example, I have included both options. (Please do not write an essay about my schematic design, these are just aids to describe the issue question!)

Motor across the high side FETS enter image description here

Motor across the low side FETS enter image description here

The average current draw for the proposed example would be around 24 A, hence the parallel arrangement for the FETs.

Also, as seen in the images, I have swapped the 'HO' and 'LO' outputs of the IR2111 driver in order to give the correct output if the 'PWM_In' input was to change. The data sheet diagram for the IR2111 can be found here: IR2111 Datasheet

The research I have conducted so far seems to suggest for applications that are not to drive motors, having a load across the low side FETs is most beneficial as seen in the second image. But for motor driving the first arrangement is more beneficial, at least from what I gather.

Please keep any response casual and respectful, I just want to hear some more experienced opinions on this!

\$\endgroup\$
7
  • 2
    \$\begingroup\$ Is your "DC motor" a BLDC or a true DC motor? You have a misunderstanding of how the motor works either way, but an answer will depend on what type of motor you're using. \$\endgroup\$
    – Darius
    Commented Mar 21, 2022 at 15:17
  • 2
    \$\begingroup\$ Both options are wrong. It might the biggest issue but it is also the easiest to explain and fix. You have second issue almost just as bad which takes a lot more explaining so an essay will be required. \$\endgroup\$
    – DKNguyen
    Commented Mar 21, 2022 at 15:37
  • \$\begingroup\$ @Darius I'm not sure I have a 'misunderstanding' as such more that my motives and spec is very niche. Anyhow, the motor is indeed a permanent magnet dc motor. Any insight would be and I would be greatful. \$\endgroup\$
    – ShashMan
    Commented Mar 21, 2022 at 16:46
  • \$\begingroup\$ @ShashMan what are your motives & spec, then? If you're looking for an answer that conforms to your specifications, you'll need to tell us what those specifications are. \$\endgroup\$
    – Darius
    Commented Mar 21, 2022 at 17:02
  • \$\begingroup\$ @Darius Half-Bridge driver that can power an absolute maximum of 1700 watts, with a continuous rating of 600/700 watts. The maximum current draw is 70 amps (as stipulated by a 70 amp fuse). The 24V coming from two YUASA car batteries that are lead acid. The PWM is coming from an esp32. The circuit have as little losses as possible as it is for an eco racer. hence the arrange of the mosfets. the High side mosfet is there to replace a flyback diode which traditionally wastes heat. \$\endgroup\$
    – ShashMan
    Commented Mar 21, 2022 at 17:20

2 Answers 2

3
\$\begingroup\$

If this is supposed to be a unidirectional motor driver using a half-bridge, then it's rather pointless because that would only allow for unidirectional motor drive. But in a unidirectional motor drive you don't need both high and low-side MOSFETs.

If this is your goal you only need either high-side MOSFETs or low-side MOSFETs, not both. But note that if you did choose high-side MOSFETs you cannot use a boostrap capacitor gate drive since those rely on a low-side MOSFET being present and switching frequently enough to refresh the bootstrap capacitor by allowing current to flow from 12V to ground through the capacitor. This also means 100% duty cycle is not allowed.

You would either go low-side MOSFETs only with low-side gate drive. Or if EMI is a concern and you need the motor to stay connected to ground you would use a continuous high-side gate drive (such as removing the bootstrap diode and connecting a isolated floating voltage supply where the bootstrap capacitor would normally go.


If this is supposed to be an bidirectional motor driver using an H-bridge (which is different than a half-bridge).

  1. Only looking at the right-hand side, that is not an H-bridge. The motor does not go in either of those places. Google schematics of an H-bridge and look at where the motor goes. You placed a permanent dead short where the motor is supposed to go.

  2. You connected the low-side gate drive outputs to the high-side MOSFETs and you connected the high-side gate drive outputs to the low-side MOSFETs.

  3. Looking at the way the left-hand side is wired into the right-hand side, Your drive signals are also wrong in more than just #2. I think the entire reason your circuit is mangled is because you looked at the gate driver schematics for a half-bridge and then tried merging it with an H-bridge. The 'H' in H-bridge does not stand for half-bridge. An H-bridge is four MOSFET arranged to shape like an H. So the left-half of an H-bridge is one half bridge, and the right-side of an H-bridge is another half-bridge.

    Your H-bridge needs two gate-driver ICs so that you have two high-side gate drives and two low-side gate drives, such that you have one for each of the four MOSFETs present.

\$\endgroup\$
8
  • \$\begingroup\$ The OP's schematic looks pretty much like that in the datasheet, other than the OP has parallel mosfets for more current - pdf1.alldatasheet.com/datasheet-pdf/view/68060/IRF/IR2111.html \$\endgroup\$
    – HandyHowie
    Commented Mar 21, 2022 at 15:49
  • \$\begingroup\$ @HandyHowie That's schematic for a half-bridge, true, but you almost never drive motors through just one half-bridge. I interpreted that it should be an H-bridge. If it's not supposed to be an H-bridge but a push-pull driver then it's kind of pointless for two reasons: (1) it's not bipolar (not that the IC could support it) so it could only spin in one direction (2) If it's supposed to be unidirectional then half-bridge and associated gate drive is pointless since you don't need both high and low-side transistors in that case. \$\endgroup\$
    – DKNguyen
    Commented Mar 21, 2022 at 15:52
  • 1
    \$\begingroup\$ @HandyHowie Edited to address that possibility. \$\endgroup\$
    – DKNguyen
    Commented Mar 21, 2022 at 15:56
  • \$\begingroup\$ Thanks for the response @DKNguyen however, it is my intention to use a half-bridge in this way as the high side is used in the same way a diode might be used to deal with back emf. The system I am trying to design is unidirectional by specification and it also most be as efficient as possible so losses over a diode would not be acceptable. While your comments about a H bridge being more widely used, yes, but for the reasons already stated, I cannot use one! \$\endgroup\$
    – ShashMan
    Commented Mar 21, 2022 at 16:19
  • \$\begingroup\$ @ShashMan Understood. Then note the bootstrap requirement if you are expecting 100% duty. Also doubt the need for parallel low-side MOSFETs if that's the case. \$\endgroup\$
    – DKNguyen
    Commented Mar 21, 2022 at 17:08
2
\$\begingroup\$

Your best option is number 3, this circuit:

schematic

simulate this circuit – Schematic created using CircuitLab

Choose your freewheeling diode to take at the very lease the full supply voltage and max inrush current, preferably more. Your PWM source may be too weak to drive a 24 A capable MOSFET, so you may need a buffer in between.

C1 is your local decoupling, unless you really know what you are doing.

\$\endgroup\$
3
  • \$\begingroup\$ This circuit was the first I considered for a motor driver as I'm sure most would have. However, for the reason I have a high and low mosfet in my half bridge drawings I don't want to use a diode. As i understand for the loads I was to put through the driver there would be too much waste heat produced by the back emf. Correct me if I am wrong. \$\endgroup\$
    – ShashMan
    Commented Mar 21, 2022 at 16:36
  • 1
    \$\begingroup\$ Unless we are talking about 5 V or less, very little is wasted in the diode. If you are ok with <200 V rating, feel free to use Shottky to lower it even further. Do you need to actively break the motor? \$\endgroup\$
    – winny
    Commented Mar 21, 2022 at 18:31
  • 1
    \$\begingroup\$ Motor back EMF is mostly motional capacitance, as winny says the diode will only see small currents. but if you want to have a electirc brake then yes, adding a MOFET that shorts the motor could do that. \$\endgroup\$ Commented Mar 22, 2022 at 4:08

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.