31 thoughts on “µM-H-Bridge”

1. Hello. Have you tested this device with motor at 10 Amps? MBRS140LT3 Shottky is ok?

   Reply ↓

   • Hi!

     This is a hard question to answer actually. I in fact have tried it with motors that consume several amps, but the current conducted by these shottky diodes depend a lot on the drive mode, the motor, the mechanical load and other factors. You might want to leave them out and rely on the intrinsic diodes in the MOSFETs for conduction – they at least have better cooling. You probably want to choose a drive mode where these diodes hardly conduct at all no matter what though… See here http://modularcircuits.tantosonline.com/blog/articles/h-bridge-secrets/mosfets-and-catch-diodes/ for more details on what to consider as far as the catch diodes go.

     Reply ↓

2. why is high resistance needed in parallel with zeener in low side mosfet

   Reply ↓

   • It keeps the input-side lock-out FETs (Q1 and Q2) gate at a valid voltage.

     Reply ↓

3. Hello?
   Do not use a 24V motor power?

   Reply ↓

   • No. The drivers on this design can’t survive that much.

     Regards,
     Andras Tantos

     Reply ↓

4. what driver should i go for if i am interested in using 24volts? I dnt want the controller part so dont keep that in mind while suggesting a driver.
   P.S great design

   Reply ↓

   • I like this: http://www.onsemi.com/pub_link/Collateral/NCP5901B-D.PDF

     Reply ↓

     • Thanks a lot. Can you help me with the changes i will have to bring?

       Reply ↓

     • the problem i having is that the application circuit is gven for a buck, what to do if it comes down to a hbridge

       Reply ↓

       • It’s really not that different: you just need a pair of these, one for each side of the bridge. You probably don’t need R3, C6, and definitely won’t need L. What kind of FET and how many in parallel you need, that depends on your application. I would still at least provision for a gate-drive resistor on both the high- and the low-side (missing on the low-side in the application circuit). Just as they show, you can always populate them with 0-ohms if not needed.

         Reply ↓

         • what is confusing me is that why did they use two mosfets on the low side. I would be using the same fets as u used in your design in the same manner as u did. Secondly I was planning to remove jp13 and jp14 alongwith r3, c6 and L. what would you say to it. I am sorry for bugging u, just needed help

         • I don’t claim to understand why they’ve used two low-side FETs. But whatever the reason is, it depends on the application. You should be fine with only one FET. The two jumpers you mention are just for testing, you can leave them out.

5. Another little favour i need is with understanding what bootstrap is for? Can u send me nice link or just explain it a bit

   Reply ↓

   • The piece on drivers (http://www.modularcircuits.com/blog/articles/h-bridge-secrets/h-bridge_drivers/) contains a lot of details about boost circuits

     Reply ↓

6. Thanks for a great deal of knowledge you provide on the blog.

   I was looking at the circuit for this h-bridge driver and I can’t grasp one thing. You’re using 4 n-channel mosfets, which means that you need a higher voltage on the high side. What technique are you using for increasing the voltage? It doesn’t look like you’re bootstrapping, but probably I’m not seeing something.

   Reply ↓

   • Sorry for a dumb question, I haven’t noticed that LTC1155 has an internal charge pump.

     Reply ↓

   • You are correct. There are several ways of dealing with this, please take a look at this article: http://www.modularcircuits.com/blog/articles/h-bridge-secrets/h-bridge_drivers/. In this particular instance, the LTC1155 has a charge pump as you’ve noticed.

     Reply ↓

7. Does it mean that you can use PWM up to 100% duty cycle in your circuit?

   Reply ↓

   • Yes, it does. However the PWM rate cannot be higher then maybe a 1kHz or so.

     Reply ↓

8. I am new to this, so please, can someone explain me the procedure to keep following and maintaining the speed of the motor?
   I have already read h-bridge.cpp, but as i mention before, i am newbie and i could n’t understand it clearly.
   Also, why there are two different lines for measuring the voltage across the motor ends?
   Thank you!

   Reply ↓

   • Vasilis,

     The motor speed control is something I wanted to write about in a follow-up article in the H-Bridge Secrets series. It’s definitely more complex of a subject then to answer it in this format. The just of it, is that it’s a PID control loop. Until I get around writing that article, you google it, you’ll find plenty of good descriptions all over the web.

     As far as the two different voltage measurement points: ideally you want to measure the differential voltage on the motor. Since one end of the motor is at a known voltage (power supply or GND) however you can figure out the differential voltage by just measuring one end. I wanted to build this bridge to be as flexible as possible so I didn’t want to constrain which end of the motor is tied to which power supply. That’s the reason for dual measurement points. Another benefit of this approach is that you can detect faults, when something is not at the expected potential. For example a failed FET or a broken wire would cause detectable changes in the way these voltages change over time. I don’t have such detection implemented in the code though…

     Reply ↓

9. What is the problem if i the q3 and q4 didn’t have an output?

   Reply ↓

   • No, that should not be a problem

     Reply ↓

10. The above article says that the bridge can handle 16A but,
    the total Rdson of the circuit is 0.018(High side)+0.018(diagonal low side) = 0.036ohms
    Power dissipation = (16*16)*0.036 = 9.216W
    Junction to ambient thermal resistance is 62 C/W
    The temperature at the junction would be 9.216*62 + 25(ambient) = 596C
    Correct me if I am wrong

    Reply ↓

    • Great articles on the H bridges BTW

      Reply ↓

    • You have some great points and I’ve corrected the page to clear up the confusion. The 16A current limit referred to the current sensing ability of the circuit and thus is the peak current delivery capability. The continuous current limit is about 10A if the MOSFets are replaced by something better then the IRFZ48 in the schematic, such as the IRL2203. For your calculations, I don’t think it’s fair to add the Rdson of the two MOSFets but then using the junction-to-ambient temperature coefficient of a single package. True, the transistors are packed together very tightly in the current design, so simply using a single package isn’t really realistic either, the truth is somewhere in between. Which is how the 10A current limit came about: 8mOhm of Rdson and 62degC/W temp coefficient (from the IRL2203 datasheet) results in a temp rise of 50degC over ambient. The maximum allowable junction temperature of the device is 175degC, which gives a rather safe margin. Redoing the same calculations with 16A would result in a temperature rise of 126degC, which starts cutting it close.

      I hope this helps and thanks again for the comment,
      Andras

      Reply ↓

      • After commenting I realized that I had considered rdson of two FETs and using temperature co-efficient of a single package. But there was no way to edit the comments.

        Reply ↓

11. You should create another write-up, where you build a bridge from scratch and directly show the calculations and brief reasoning for it. From the output side to the input side.
    1)Motor – its specifications such as Voltage, no-load current, stall current.
    2)MOSFETs for H bridge depending on the Motor specs.
    3)Driver depending on the MOSFET specs and PWM frequency.

    Reply ↓

12. Hello Andras,

    thank you for the amazing work you share with us here 🙂

    Can this be used also to control a TEC in cooling and heating mode? If so, can I control a 180W TEC Module which works at max 12A and max 15V?

    Best regards
    Timo

    Reply ↓

    • Timo,

      I think it can be done, but I have no experience personally. A quick search showed that people do it by adding an inductor in series with the TEC to serve as a voltage-to-current averaging device. Check out the LTC1923: this is a ready-made controller for what you need, I think. Even if not, should provide you with some ideas on how to use an H-bridge for this application.

      Andras

      Reply ↓