Shop OBEX P1 Docs P2 Docs Learn Events
Please recommend an H-Bridge I can use with the Propeller — Parallax Forums

Please recommend an H-Bridge I can use with the Propeller

G McMurryG McMurry Posts: 134
edited 2012-03-19 05:58 in Robotics
I am building a circuit that needs to drive a small 12 vdc motor (under 1 amp) full speed in either directions. I would like the lowest parts count so maybe one of those H-Bridge ICs would work great. I just don't know how to pick one.

I need it to operate with 12 vdc for my small motor and 3.3 volt inputs compatible with the Propeller or the PCA95444 IO expander running at 3.3 volts.

I built my prototype using an H-Bridge relay configuration but would like to change to a solid state driver.

Parallax sells the L293D but it doesn't look like it will operate with 3.3 volt inputs.


Any advice would be greatly appreciated.

Greg

Comments

  • Martin_HMartin_H Posts: 4,051
    edited 2012-02-29 17:11
    Parallax sells a Pololu motor driver that says it is compatible with 3v logic levels:

    http://www.parallax.com/Store/Accessories/MotorServos/tabid/163/ProductID/786/List/0/Default.aspx?SortField=ProductName,ProductName

    I bought one but have not yet put it into service. I've used other Pololu motor drivers and they were well made.
  • G McMurryG McMurry Posts: 134
    edited 2012-02-29 17:45
    Looking for something less exotic. An H-Bridge IC of some sort.

    THANKS FOR YOUR REPLY

    Greg
  • Martin_HMartin_H Posts: 4,051
    edited 2012-02-29 17:49
    OK, but that is just two surface mount h-bridge IC's and some support components. Are you building your own board?
  • G McMurryG McMurry Posts: 134
    edited 2012-02-29 17:52
    Yes, I am building my own board and when it works, I will need about 50 of them.

    Greg
  • Martin_HMartin_H Posts: 4,051
    edited 2012-02-29 17:59
    That board uses the Freescale MC33926 full H-bridge IC which is available separately.

    If you want through hole I bet the L298 would be compatible because the propeller's 3.3 volt logic should be able seen as 1 to its 5 volt logic.
  • G McMurryG McMurry Posts: 134
    edited 2012-02-29 21:16
    Great idea... I will take a look at it.

    Greg
  • Duane DegnDuane Degn Posts: 10,588
    edited 2012-02-29 21:54
    Greg,

    Here are my favorite H-bridges. The Texas Intrument's SN754410 and the L298N. I've used both in this project. Both can be directly driven by the Propeller. You'll want to get a bunch of schotty diodes to go with them. Each chip uses eight diodes.

    I read conflicting opinions about the need for external diodes with the SN754410 chip but I'm pretty sure they're a good idea. The SN754410 datasheet schematic has external diodes.

    The DIP SN754410 is easy to use with a breadboard. I thought the L298N wouldn't be breadboad friendly, which it isn't but it's not breadboard mean either. The L298N just needs a little persuasion (bending the leads a bit different than their original positions) when used with a breadboard (as seen in post #46 of my project thread).

    I think the L293D will work with 3.3V logic. I'm 98.3% sure I've use 3.3V logic with a L293D chip. I'm sure I've used 3.3V logic with the SN754410 chip. The recent videos in my project thread show the Prop controlling L298N chips directly (confirming Martin's suspicion).

    I've just recently started using L298N chips and I've been very pleased with them so far. They don't seem to get as hot as the pair of SN754410 chips running in parallel. This shouldn't be too surprising since the L298N is a 4A chip and the SN754410 is a 1A chip. I think the L298N's are worth the extra $0.60. The L298N are less expensive than using two SN754410 chips in parallel.

    While these chips (all three mentioned) will work with 3.3V logic, they need to be powered by 5V.

    BTW, while I have some experience with the chips listed above, that's about the limit of my motor control experience. Don't take my advice as the definitive answer to your motor control options.
  • G McMurryG McMurry Posts: 134
    edited 2012-03-07 17:43
    Thanks for the GREAT advice...

    I went ahead with the SN754410. I am driving a small 2-wire DC motor that pulls about 1/2 amp. I will run very short duty cycles (3 seconds on and 10 minutes off) and at 12vdc. I run the motor alternately fwd and reverse.

    I have been wondering why I don't parallel the two drivers in the chip. Is there any benefit to doing that or is my biggest enemy just heat (watts).

    I included the schematic of how I plan on wiring this. Basically parallel 1A with 4A and 2A with 3A inputs. Then parallel 1Y with 4Y and 2Y with 3Y outputs that would go to the motor.
    093_I2C_h-bridge.jpg

    Then my control lines would be Parallel enable lines plus the little transistor inverter for a direction bit. Set the direction then enable the motor for 2 seconds...

    I also thought I would add two LEDs back to back across the motor lines as an indicator of motor operation.

    Any thoughts?

    Greg
    093_I2C_h-bridge.jpg
    800 x 536 - 68K
  • Martin_HMartin_H Posts: 4,051
    edited 2012-03-08 06:38
    Driving one motor by doubling the outputs of the SN75441 should increase the current handling. I've seen people stack one SN75441 on top of the other and solder their pins together. This doubles the current capacity of both channels, but increases current consumption of the logic portions of the chips. There are heat sinks that can wrap around the top chip and draw heat from the bottom one too. It looks sort of funny but works.

    Gadget Gangster used to sell a module called the OctoDriver (http://gadgetgangster.com/find-a-project/56?projectnum=255) which was four SN75441 used to control two motor channels. While the logic current consumption was high it was a beast of a motor driver board.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2012-03-08 07:39
    G McMurry wrote: »
    plus the little transistor inverter for a direction bit.

    I'm not sure what advantage the tansistor gives you? The same pin that controls the transistor could control the second input pin. The various motor control objects assume the Prop is connected to both control pins.


    Most applications I've seen leave the enable pins pulled-high. I think controlling the enable pin lets one coast to a stop instead of having the motor act as a break so depending on the application, it might be good to be able to control it.

    Any other comments I've already make in my earlier post (diodes etc.).
  • G McMurryG McMurry Posts: 134
    edited 2012-03-08 08:21
    I am using the Transistor for an inverter to give me two line control of the device. One for Direction and one for Enable. When disabled, the outputs go tri-state so the motor wires are not shorted allowing it to freewheel and reducing power when not in use. The other line (with the transistor) then just becomes a direction line.

    I found a little heat sink for the IC but I bet with my expected duty cycle, it will be unnecessary. On my PCB, I have loads of ground area that will take away a good amount of heat.

    BTW -- I am no longer connecting the device directly to a Propeller. I am using a Spinneret and have a PCA9554 connected via I2C to create my IO. I am just running out of pins.

    Greg
  • Duane DegnDuane Degn Posts: 10,588
    edited 2012-03-08 08:46
    G McMurry wrote: »
    I am using the Transistor for an inverter to give me two line control of the device. One for Direction and one for Enable. When disabled, the outputs go tri-state so the motor wires are not shorted allowing it to freewheel and reducing power when not in use. The other line (with the transistor) then just becomes a direction line.

    I found a little heat sink for the IC but I bet with my expected duty cycle, it will be unnecessary. On my PCB, I have loads of ground area that will take away a good amount of heat.

    BTW -- I am no longer connecting the device directly to a Propeller. I am using a Spinneret and have a PCA9554 connected via I2C to create my IO. I am just running out of pins.

    Greg

    OK, I think I understand. You're not controlling the speed of the motor. You just want to turn it on and off and control the direction or rotation.

    Thanks for the explaination.
  • G McMurryG McMurry Posts: 134
    edited 2012-03-18 23:07
    OK -- I have had much success with using the SN75441 on my project. However, I am also having HUGE noise problems. Can anyone help me pick the proper diodes to use on the output? I got my circuit to work using some Varistors I had laying around but I need to incorporate the proper diodes now. All the literature mentions diodes but I have no idea what kind of diode or how to pick one.

    I am running a 12 volt 400ma motor. Duty cycle is extremely low and I have paralleled the two outputs on my SN75441.

    THANKS

    Greg
  • Martin_HMartin_H Posts: 4,051
    edited 2012-03-19 05:46
    The diodes can't hurt, but for noise some 0.1 uF capacitors across the motors leads should help as well.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2012-03-19 05:58
    G McMurry wrote: »
    OK -- I have had much success with using the SN75441 on my project. However, I am also having HUGE noise problems. Can anyone help me pick the proper diodes to use on the output? I got my circuit to work using some Varistors I had laying around but I need to incorporate the proper diodes now. All the literature mentions diodes but I have no idea what kind of diode or how to pick one.

    I am running a 12 volt 400ma motor. Duty cycle is extremely low and I have paralleled the two outputs on my SN75441.

    THANKS

    Greg

    These are the diodes I'm using. I think you want four diodes per motor.

    I used to use the SN754410 a lot. I like the L298N chips a lot more. They don't cost much more (less than two SN754410) and don't get as hot in my experience.

    Besides the caps between the leads that Martin mentioned, I've also read suggestions (which I believe to be good advice) to use two more caps. One from each lead to the motor's casing.
Sign In or Register to comment.