Connect with us

Motor Control Circuit Problem

Discussion in 'Electronic Design' started by Steve, Oct 28, 2004.

Scroll to continue with content
  1. Steve

    Steve Guest

    Hello,
    My group and I built the following dc motor control circut:
    http://forum.webzila.com/index.php?act=Attach&type=post&id=2441

    I replaced the SK100 transistor with NTE262 and the SL100 with TIP110

    We have a high speed 9-18V DC Motor (18,000 RPM / 1.98A max).
    When we plug a 12 V DC power supply directly to the motor it spins very
    fast but when we use the circuit to drive the motor then it seems to
    spin much slower.

    With power supply set to 12V, we just measured to voltage going to the
    motor and it says 1.5V .... what is wrong?

    We just now set the power supply to 20V and the leads that connect to
    the motor still only output 1.5V
     
  2. Terry Given

    Terry Given Guest

    Its because you have a built-in "will not work" feature. The npn
    transistors Q3,Q5 in your H-bridge have no series resistors. Consider
    only Q1,Q3,Q4 (the same is true for Q2,Q5,Q6) we see that:

    with Q1 on, its collector sits at Vcesat = 0.2V (say), and pnp Q4 is on
    (base current flows thru the 1k into Q1). Q3 is off. When Q1 turns off,
    its collector 1k resistor pulls the collector high - at least it tries
    to, but when Q1 collector rises to 0.7V (say) then Q3 base-emitter
    junction is forward-biased, clamping Q1 collector to 0.7V - thereby
    ensuring BOTH Q3 and Q4 are on. oops.

    If you stick a resistor in series with Q3 base, it will work a bit
    better - but the value of base resistor is important to ensure Q3,Q4
    both turn off. use say 22k resistors for the base of Q3 and Q4, and also
    slap 10k across the b-e junctions of Q3,Q4. This ensures that Q1
    collector will pull up to 22k/23k*9V = 8.6V, ensuring Q4 will be off.

    You will also have cross-conduction at every switching edge - at some
    point both Q3 and Q4 will be on simultaneously, attempting to short out
    the supply rail. A careful choice of base resistor (and perhaps a
    speedup cap) is required. or a better circuit.

    Cheers
    Terry
     
  3. Terry Given wrote:

    [snip]
    Cracking phrase Grommit.

    Never stopped Sinclair, Amstrad or Tandy.

    Gibbo
     
  4. Steve

    Steve Guest

    Originally we got the circuit from here:
    http://www.electronic-circuits-diagrams.com/motorimages/motorckt4.shtml

    Then someone told us that there is no current limiting between the
    bases of the transistors. So we modified the circuit like this:
    http://forum.webzila.com/index.php?act=Attach&type=post&id=2441

    Would the original design work better?

    Would the advise of modifying the various resistors that you and Terry
    provided fix this issue? At this point we are not asking for the motor
    to receive the full voltage that is supplied to the circuit but we need
    as much as possible so that it can pull a 2lbs door.

    We cant really afford to build a new circuit unless it can use
    identical parts as the current one.

    Can you again explain the changes that we should make or show a simple
    circuit diagram of the way we should modify the circuit? When you say
    to use a 22K resistor for Q3/Q5 base, do you mean put a resistor going
    from the base of Q3 to the base of Q5? Also when you talk about the
    Q4/Q6 resistor, are you refering to the 1K resistor that is between Q4
    and Q3 & Q5 and Q6?

    Also, when you say make Q1 and Q2 resistors 220 ohm...are you refering
    to the 10K resistors that are going to the base or the 1K resistors
    that are connected to the emitter?

    Thanks
     
  5. Terry Given

    Terry Given Guest

    the original circuit was even worse, for the same reasons I outlined -
    except with no base resistors at all (as opposed to half the amount you
    require) its a recipe for frying transistors whilst maintaining a
    stationary shaft on the dc machine.

    John's fix is essentially the same as that I proposed - base resistors
    about, but choose values wisely; some semblence of understanding is
    required.

    Cheers
    Terry
     
  6. nospam

    nospam Guest

    The circuit diagram is wrong, and so is what you built if you followed it.

    Swap Q3 with Q4 and Q5 with Q6. The emitters of Q3 and Q4 should be
    connected and drive one side of the motor, Q5 and Q6 similarly driving the
    other.
     
  7. Steve

    Steve Guest

    I changed the Q1 and Q2 resistors to 220 ohms
    Based on this circuit diagram can you show me what you mean by the
    following:
    http://users3.ev1.net/~srudenko/img/hbridge.gif

    I also have the EWB file (maybe if you have EWB you can show me what
    you mean by making the changes in there).

    Or just explain exactly where to plug in the resistors you are refering
    to.

    Thanks
     
  8. john jardine

    john jardine Guest

    The circuit is not right.
    When Q1 is switched ON then the circuit works OK with Q3 OFF, (.7V at base),
    and Q4 ON (11.3V at base).
    When Q1 is switched OFF its collector voltage will rise to switch ON Q3
    *but* the voltage at the base of Q4 cannot rise to 12.0V to allow it to
    switch off. That's because there maybe a couple of ma's being taken by Q3
    base when it's ON and this current is supplied by the 1k Q2 resistor and the
    1k Q4 base resistor. This causes a few volts to be dropped and Q4 is still
    switched ON.
    The same thinking applies to both halves of the circuit so at any given time
    one half of the circuit must be passing a big current through 2 of the
    transistors (hot!) and not the motor. The effect is to bring the motor
    voltage nearly to nothing
    A redesign is needed. Not brilliant but maybe make Q1 Q2 resistors 220ohm.
    Stick a 2k2 resistor in Q3 Q5 base. Change Q4 Q6 resistor to 2k2.
    Where did'y'get the circuit from originally?.
    regards
    jhon
     
  9. Out of interest I tried simulating the original (symmetrical) circuit
    http://www.electronic-circuits-diagrams.com/motorimages/3.gif
    with CircuitMaker. But after an hour or so I'm darned if I can get it
    working. Typical error messages include:
    Warning: Gmin step failed
    Warning: source stepping failed
    doAnalyses: Iteration limit reached'

    Anyone care to try it in some other package please, such as LT
    SWCADIII? I'm assuming there's some nifty trick for getting an
    H-Bridge to converge in Spice...
     
  10. nospam

    nospam Guest

    Geees the original circuit shorts out the power supply with the power
    transistor base emitter diodes - what did you expect to simulate?

    As you have gone to the trouble of drawing it simply drag Q4 to where Q3 is
    and Q3 to where Q4 is, the same for Q5 and Q6.

    You then have a pair of simple class B complimentary amplifiers driving the
    motor in a bridge. Why no one else can see this is how the original circuit
    was supposed to be is beyond me.
     
  11. Rich Grise

    Rich Grise Guest

    This is the circuit configuration known as:
    "Blow up four transistors simultaneously." ;-)
    (note current path from +9v-Q4E-Q4B-Q3B-Q3E-GND, Q5,6 same)
    Cheers!
    Rich
     
  12. I understand that, but that wasn't the point. The circuit should still
    simulate for each of the 4 input conditions.
     
  13. See my reply to nospam.
     
  14. Jim Meyer

    Jim Meyer Guest


    You are of course 100% right. That's the only way the original
    circuit could have worked. The schematic got screwed up. No current
    limit resistors needed at all.


    Jim
     
  15. What happens is that the current goes too high on one of the top
    transistors, causing the computation to fail. Try putting a 1 ohm
    resistor into the base of both Q4 and Q6. That will enable the
    simulation to complete without failing. Then, decrease the value of that
    resistor, and see what happens. When you get down to some very small
    value, it'll start to fail again.

    --
    Regards,
    Robert Monsen

    "Your Highness, I have no need of this hypothesis."
    - Pierre Laplace (1749-1827), to Napoleon,
    on why his works on celestial mechanics make no mention of God.
     
  16. Rich Grise

    Rich Grise Guest

    Yeah. Time warp. Yours is better, of course, since it gives a
    useful answer. :)

    Thanks,
    Rich
     
  17. Terry Given

    Terry Given Guest

    Still, their revised circuit is an improvement on the original rubbish,
    even if it doesnt yet work right. its getting closer.....

    Cheers
    Terry
     
  18. Thanks, Robert, that sorted it immediately!
     
  19. john jardine

    john jardine Guest

    Q1Q2 keep as 220ohms, then bridge resistors like this ...

    -----+---------------------+-------- +V (eg 12V)
    | |
    | |
    | Q4 Q6 |
    | NTE262 NTE262 |
    ___ |< >| ___
    ,-|___|--| |-|___|--,
    | 2k2 |\ /| 2k2 |
    | | | |
    | | | |
    | | | |
    | | Motor | |
    | | ___ ___ | |
    to Q1-----+ +-----o-|___|--UUU-o--+ +--to Q2
    | | | |
    | | 9ohm 30mH | |
    | | | |
    | | | |
    | | | |
    | ___ |/ \| ___ |
    '--|___|-| |-|___|--'
    2k2 |> <| 2k2
    | |
    | Q3 Q5 |
    | TIP110 TIP110|
    | |
    ---------+---------------------+-----0V
    created by Andy´s ASCII-Circuit v1.24.140803 Beta www.tech-chat.de

    This setup's not optimum as it's a wee bit reliant on the actual base
    current pulled by Q3 or Q5.

    A nicer and simpler design (by Antonio Raposo) can be found at ...
    http://www.cs.uiowa.edu/~jones/step/circuits.html

    Antonio's circuit design uses a total of only 4 transistors and all could
    be say TIP110 Darlingtons.
    regards
    john
     
  20. john jardine

    john jardine Guest

    I've wondered if newcomers are ever put off the subject, after spending
    hours trying to wrestle some sense out of useless designs.
    I take a mechanical engineering mag' that recently did a short series on
    'electronics in the workshop'. The author showed an H bridge circuit with
    just the same problem as Steves. Poor proof reading or whatever, the author
    should have known better. Shame on him :).
    (ps, If I'd seen the group posts 'updating' I wouldn't have bothered with my
    post. I was wasting bandwidth just repeating what you had said)
    regards
    john
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-