Loading ...
Sorry, an error occurred while loading the content.
 

Re: [beam] Protecting an H-bridge

Expand Messages
  • Gary Tolley
    So Wilf, What do you think of it? Gary T
    Message 1 of 7 , Nov 30, 2005
      So Wilf,
      What do you think of it?

      Gary T
    • wilf_nv
      Hi Gary, I m not sure if I am looking at the right drawing but here are a few comments: 1) The transistors (Q1-4) should each have a resistor in series with
      Message 2 of 7 , Dec 1, 2005
        Hi Gary,

        I'm not sure if I am looking at the right drawing but here are a few
        comments:

        1) The transistors (Q1-4) should each have a resistor in series with
        the base.
        2) I understand you would control an external h-bridge with the outputs
        of this circuit but the L293 already is a fullfledged 1A motor driver.
        3) The L293 output levels are not compatible with most H bridge inputs
        4) 2K2 Resistor (R1-4) are not recommended in series with the L293
        inputs.
        5) U2 pin 11 connected to pin 12 will cause bad oscillations.



        wilf

        --- In beam@yahoogroups.com, Gary Tolley <grogyan@w...> wrote:
        >
        > So Wilf,
        > What do you think of it?
        >
        > Gary T
        >
      • Gary Tolley
        At 06:38 p.m. 2/12/2005, you wrote: Hi Gary, I m not sure if I am looking at the right drawing but here are a few comments: 1) The transistors (Q1-4) should
        Message 3 of 7 , Dec 1, 2005
          At 06:38 p.m. 2/12/2005, you wrote:
          Hi Gary, 

          I'm not sure if I am looking at the right drawing but here are a few
          comments:

          1) The transistors (Q1-4) should each have a resistor in series with
          the base.

          This is a work in progress, and any ideas and suggestions are welcome.
          I havn't yet breadboarded it and probably won't, but I cannot see why I need a bias resistor, which also destabilzes the signal delay circuitry.

          2) I understand you would control an external h-bridge with the outputs
          of this circuit but the L293 already is a fullfledged 1A motor driver.

          The L293 is indeed a complete driver, and so it has been selected to run further testing before implementing my circuit to custom driver circuits.

          3) The L293 output levels are not compatible with most H bridge inputs

          Who, what?
          I don't understand what your saying.

          4) 2K2 Resistor (R1-4) are not recommended in series with the L293
          inputs.

          It is recomended as its part of the signal delay circuit (CR), how do you suggest I implement CR?

          5) U2 pin 11 connected to pin 12 will cause bad oscillations.
          Funny story, as I was working on it, I ran out of space, moving whole parts was the only soloution, and as result many connections appeared, I tried to find them all before posting, obviously I missed one.  Thanks.



          wilf

          --- In beam@yahoogroups.com, Gary Tolley <grogyan@w...> wrote:
          >
          > So Wilf,
          > What do you think of it?
          >
          > Gary T
          >






          YAHOO! GROUPS LINKS



        • Wilf Rigter
          ... OK, so the external h-bridges would be used in place of the L293! That s the bit that I was not clear on. I thought that the L293 was part of the logic or
          Message 4 of 7 , Dec 2, 2005
            On December 01, 2005 9:56 PM, you wrote:
             
            >The L293 is indeed a complete driver, and so it has been selected to run
            > further testing before implementing my circuit to custom driver circuits.
            OK, so the external h-bridges would be used in place of the L293! That's the bit that I was not clear on. I thought that the L293 was part of the logic or some kind of pre-driver for a much larger h-bridge. 
             
            >I cannot see why I need a bias resistor (in series with the base
            of Q1-4)  , which also destabilzes the signal delay circuitry
             
            When you connect an external signal to the base of  Q1-4 the input current must be controlled or you will blow up the base-emitter junction. It is much safer to add a series base resistor to your circuit to protect the transistors than to rely on the external signal source to be current limited.
            With the transistor used as a "switch" there is not connection between the base resistor and the "signal delay circuitry".
             
            4) 2K2 Resistor (R1-4) are not recommended in series with the L293
            inputs.

            2K2 Resistor (R1-4) is recomended as its part of the signal delay circuit (CR), how do you suggest I implement CR?
             
            Ironically, R1-R4 would be quite useful as base resistors for Q1-4.  The input of a L293 is high impedance and at best, a series 2K2 resistor does not do anything including affecting the RC timeconstant. The L293 input can be safely directly connected to a logic input signal without the series input resistor .
             
            I am not sure what the purpose of your "signal delay" is but in general such a delay is used to prevent "shoot-through" current.
             
            AFAIK your interface circuit converts 2 control input to 3 control outputs for bidirectional control of dc motor using a L293 (left, right, brake, freewheel states). This would mean that any other h-bridge you want to control with this circuit would also need 3rd input (ENABLE). Just to be clear, is that what you want?
             
            wilf 
             
            ----- Original Message -----
            Sent: Thursday, December 01, 2005 9:56 PM
            Subject: Re: [beam] Re: Protecting an H-bridge

            At 06:38 p.m. 2/12/2005, you wrote:
            Hi Gary, 

            I'm not sure if I am looking at the right drawing but here are a few
            comments:

            1) The transistors (Q1-4) should each have a resistor in series with
            the base.

            This is a work in progress, and any ideas and suggestions are welcome.
            I havn't yet breadboarded it and probably won't, but I cannot see why I need a bias resistor, which also destabilzes the signal delay circuitry.

            2) I understand you would control an external h-bridge with the outputs
            of this circuit but the L293 already is a fullfledged 1A motor driver.

            The L293 is indeed a complete driver, and so it has been selected to run further testing before implementing my circuit to custom driver circuits.

            3) The L293 output levels are not compatible with most H bridge inputs

            Who, what?
            I don't understand what your saying.

            4) 2K2 Resistor (R1-4) are not recommended in series with the L293
            inputs.

            It is recomended as its part of the signal delay circuit (CR), how do you suggest I implement CR?

            5) U2 pin 11 connected to pin 12 will cause bad oscillations.
            Funny story, as I was working on it, I ran out of space, moving whole parts was the only soloution, and as result many connections appeared, I tried to find them all before posting, obviously I missed one.  Thanks.



            wilf

            --- In beam@yahoogroups.com, Gary Tolley <grogyan@w...> wrote:
            >
            > So Wilf,
            > What do you think of it?
            >
            > Gary T
            >






            YAHOO! GROUPS LINKS



          • Gary Tolley
            ... If thats the case then I should be able to use just a 1K resistor, right? ... In this case the CR circuit is to prevent any damage to an H-bridge. ...
            Message 5 of 7 , Dec 3, 2005
              At 06:17 p.m. 3/12/2005, you wrote:
              >On December 01, 2005 9:56 PM, you wrote:
              >
              > >The L293 is indeed a complete driver, and so it has been selected to run
              > > further testing before implementing my circuit to custom driver circuits.
              >OK, so the external h-bridges would be used in place of the L293! That's
              >the bit that I was not clear on. I thought that the L293 was part of the
              >logic or some kind of pre-driver for a much larger h-bridge.
              >
              > >I cannot see why I need a bias resistor (in series with the base of
              > Q1-4) , which also destabilzes the signal delay circuitry
              >
              >When you connect an external signal to the base of Q1-4 the input current
              >must be controlled or you will blow up the base-emitter junction. It is
              >much safer to add a series base resistor to your circuit to protect the
              >transistors than to rely on the external signal source to be current limited.
              >With the transistor used as a "switch" there is not connection between the
              >base resistor and the "signal delay circuitry".
              >
              >>4) 2K2 Resistor (R1-4) are not recommended in series with the L293
              >>inputs.
              >
              >2K2 Resistor (R1-4) is recomended as its part of the signal delay circuit
              >(CR), how do you suggest I implement CR?
              >
              >Ironically, R1-R4 would be quite useful as base resistors for Q1-4. The
              >input of a L293 is high impedance and at best, a series 2K2 resistor does
              >not do anything including affecting the RC timeconstant. The L293 input
              >can be safely directly connected to a logic input signal without the
              >series input resistor .

              If thats the case then I should be able to use just a 1K resistor, right?

              >
              >I am not sure what the purpose of your "signal delay" is but in general
              >such a delay is used to prevent "shoot-through" current.

              In this case the CR circuit is to prevent any damage to an H-bridge.

              >
              >AFAIK your interface circuit converts 2 control input to 3 control outputs
              >for bidirectional control of dc motor using a L293 (left, right, brake,
              >freewheel states). This would mean that any other h-bridge you want to
              >control with this circuit would also need 3rd input (ENABLE). Just to be
              >clear, is that what you want?

              Exactly, as I can just use trannies to turn the H-bridges on and off
              automatically on the control pwm signals.

              >
              >wilf

              I tried entering the robo olympics at Massey University, unfortunatly the
              night before the competition I blew up both h-bridges whilst debugging the
              rest of the bot, with one completely unusable, and the other capable of
              forward movement. Look, lots of smoke, t o x i c.
              This is why I am trying to prevent any such re-occurance.
            Your message has been successfully submitted and would be delivered to recipients shortly.