Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

PWM and Pull up/pull down

  1. Mar 27, 2015 #1
    Hi all,

    I am an electrical engineer and feel I should know this (embarrassing!), but I for the life of me still cant quite get this concept.

    I understand conceptually the idea behind a pull up or a pull down, in that we want the IC to read a high state, by defaultm in the case of a pull up and low by default in case of pull down. But earlier to day an engineer was saying that we needed a lower pull down resistor to drive a pwm from 3 to 5 volts. I still dont get/understand the connection between a pull up/pull down and pwm signal? I get that its ideal to drive a motor using a pwm signal and a resistor could be used as an impedance to drive the pwm, but i still dont get the use of resistor in pwm to pull it "up" or "down"? an example, links or clarifications would be much appreciated. thanks
  2. jcsd
  3. Mar 28, 2015 #2


    User Avatar
    Science Advisor

    Welcome to PF.

    A pull up/down resistor would not be used as part of the H-bridge power circuit that drives a motor, but it may well be used in the PWM signal switching control circuits that drive the H-bridge. http://en.wikipedia.org/wiki/H_bridge

    The digital PWM signal from the controller may not swing through a sufficient voltage to fully turn on or off the H-bridge switches. That will result in significant unnecessary power dissipation in the switch components. A resistor might be used to increase the speed of transition or voltage range.

    Some circuits and ICs have an “open collector/drain” output. They can drive voltages higher than the logic supply, but need some current limited voltage pull up technique to do so. http://en.wikipedia.org/wiki/Open_collector

    Others families such as TTL 74xx, 74LSxx and 74HCTxx outputs have asymmetric drive capabilities. They are happy to sink current but prefer not to source it. Pull up resistors will increase their output voltage range.

    We would need more specific information before we could explain in more detail the particular case you have referred to.
  4. Mar 28, 2015 #3
    Hi Baluncore,

    Thanks for the response.

    The second and third point you made makes sense and seems to apply to the project im working on.. basically I created a level translator circuit boosting 0-5V pwm to 0-12v pwm. While looking at the oscope the senior engineer kept saying we needed to add a lower resistance to the pull down resistor in order to get the voltage we needed. I guess it would make sense that he wanted to get the full voltage range. still why couldnt we get the voltage we want by varying the duty cycle.? and if it was a pull up then I am guessing we would have to increase the resistor value....
    Last edited: Mar 28, 2015
  5. Mar 28, 2015 #4


    User Avatar
    Science Advisor

    So what exactly is the low state impedance of your driver circuit in it's current form? Perhaps your colleague realised that your circuit was providing insufficient low state drive current to give sharp enough on state to off state transitions (of the mosfets).
  6. Mar 28, 2015 #5


    User Avatar
    Science Advisor

    Although PWM sets the average motor voltage over one or more complete cycles, the concern here is with the gate drive voltage during every single cycle. That cannot be averaged.

    The critical thing about power switching is that the switch must either drop a very low voltage or have a very low current flowing. That is to minimise Wswitch = Vswitch * Iswitch. Clean fast transitions become very important in determining Wswitch.

    It is not the average PWM voltage to the motor that is the problem here, it is the voltage to the gates of the switches. Unless that control voltage swings quickly and cleanly through the threshold voltage the switch is being operated for part of the cycle as a "Class A amplifier" rather than a "Class D switch".

    If the PWM period is T then PWM can only be used to control average Vmotor over times greater than T. Times between zero and T are the domain of the gate drive circuits.
  7. Mar 28, 2015 #6

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    uart and baluncore's observations are spot on.
    The 'scope told him something and we don't know what.
    That would've been the time to ask him to point what it was about the 'scope trace telling him that..
    Here's a trace that's in need of some help "pulling down", observe the down transition is slooo-ww, and the up transition is no Barney Oldfield.. .


    Most senior guys appreciate honest questions.

  8. Mar 28, 2015 #7
    thank you baluncore, uart and Jim Hardy.

    Mr. Hardy, the pics really helped! I cant say I completely understand but definitely get some of it. Maybe I should consider taking a more indepth electronics course.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook