Boosting a square wave from 12 to 24 volts

Discussion in 'General Electronics Chat' started by birdsnestbreadboard, May 20, 2016.

  1. birdsnestbreadboard

    Thread Starter New Member

    May 20, 2016
    6
    0
    Hello all,
    I have a 12 volt square wave output that let's say comes from a logic-driven black box (I have no control of this as it comes from a component from an outside entity). The output square wave has a varied duty cycle but cycles at a frequency around 31kHz. I am using it to fire a linear actuator which requires 24 volts.

    I need to boost this signal to 24 volts while maintaining the integrity of the signal (minimal distortion in the duty cycles and frequency). Can this be accomplished with a high-frequency op amp? I was thinking about using the TLE2301 amp. So I'm pretty much trying to boost a PWM from 12 to 24 volts.

    Thanks for your time
     
  2. #12

    Expert

    Nov 30, 2010
    16,298
    6,810
    Simpler. Just have the square wave drive the gate of a mosfet. Connect the 24 volts to the linear actuator and connect the actuator to the drain terminal of the mosfet.
     
  3. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,540
    1,251
    Output source current?
    Output sink current?

    ak
     
  4. birdsnestbreadboard

    Thread Starter New Member

    May 20, 2016
    6
    0
    That's a good idea. Will the fact that the gate signal is 12v and possibly near 500mA matter? I've only ever driven MOSFETs with low power microcontroller signals. Thanks a lot for the quick reply.
     
  5. birdsnestbreadboard

    Thread Starter New Member

    May 20, 2016
    6
    0
    I should have specified this. The 12v signal that I'm trying to boost is a source current.
     
  6. BobTPH

    Active Member

    Jun 5, 2013
    782
    114
    The MOSFET will be quite happy with 12V on the gate. In fact, most require 10V for maximum performance, and can handle up to 20V.

    The current rating of the output does not matter. The MOSFET gate is essentially a capacitor, and once charged, take no current at all.

    Bob
     
  7. BobTPH

    Active Member

    Jun 5, 2013
    782
    114
    Do you mean a current source, as in, it will provide the same current no matter what load you put on it (within some range?) If so, why are you calling it 12V. You cannot have a current source at a specific voltage.

    Bob
     
  8. birdsnestbreadboard

    Thread Starter New Member

    May 20, 2016
    6
    0
    Sorry- The current is not infinite.
     
  9. birdsnestbreadboard

    Thread Starter New Member

    May 20, 2016
    6
    0
    Good to know. I think this is a good solution.
     
  10. hp1729

    Well-Known Member

    Nov 23, 2015
    1,954
    219
    Are high voltage op amps like the LM343 still available?
    OPA445 Available from DigiKey, about $12. Good to + and - 45 V
    MC1436 / MC1536?
     
    Last edited: May 21, 2016
  11. #12

    Expert

    Nov 30, 2010
    16,298
    6,810
    Not relevant to the Thread Starter.
    Besides, you know how to answer your own question by using vendor websites, Google, and eBay.
     
  12. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,540
    1,251
    It sounds like the 24 V output has to source current into the load, but not sink current from the load when the output is 0 V. yes/no

    If yes, I'm surprised. I would think that when driving something with PWM, a cleanly defined off time is important.
    If yes, how much current does the output have to source at 24 V?
    How close to exactly 24 V must the output be? Is 23 V good enough?
    At 31 kHz, what are the maximum and minimum pulse widths?

    Since the output sounds like PWM, logic polarity is important. The 24 V output must be non-inverted with respect to the incoming 12 V waveform. We're heading toward an NPN driver followed by a P-channel MOSFET or PNP BJT. Everyone loves power MOSFETs, but in industrial applications I've found bipolars to be better for long term reliability. 2N4401 + TIPxx (depending on output current), plus 3 resistors.

    ak
     
  13. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,540
    1,251
    He means the circuit must source current at 24 V, not that it is a constant current source circuit with 24 V compliance. It pumps current into the external load based on the load impedance.

    ak
     
  14. birdsnestbreadboard

    Thread Starter New Member

    May 20, 2016
    6
    0
    Hey Ak,
    Your assumption is correct- I need to source current at 24 V. I apologize for all the confusion, I am still learning. Are you saying that if I don't sink current as well at 0 v, the cutoff is going to be sluggish?

    At 24 V, the load is drawing 1.09 A. The voltage needs to be at 24 v or possibly even a little higher. The minimum pulse widths at 33kHz (I know that I initially said 31) are around 12 microseconds.

    I am running a linear actuator that I have fired at a voltage as high as 31 v. In this case, the load was pulling around 1.3 amps.

    It sounds like you are suggesting a "Darlington pair" of transistors for keeping the signal non-inverted. Is this a correct assumption? I have been drawing off of the circuit concept shown here for ideas (I realize that this would invert it though and that I need another transistor) upload_2016-5-23_16-35-17.png
     
  15. BobTPH

    Active Member

    Jun 5, 2013
    782
    114
    The circuit you show does not invert the signal. The power is on to the solenoid when your 12V signal is high and off when it is low.

    Bob
     
  16. ian field

    Distinguished Member

    Oct 27, 2012
    4,415
    784
    If the output is open collector/drain; you can simply strap on the grounded gate section of a casc-ode circuit. (its easier with a LL MOSFET than a bipolar transistor). The grounded gate section is non inverting and Vdd can be as high as the MOSFET VDS rating as long as the load isn't reactive.

    The current gain of the grounded gate section is unity, so the load current cannot be more than the original output can handle. You can increase the output current by adding a PNP/P-channel follower in the manner of a Szicklai pair.
     
Loading...