1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

Yet another H-Bridge

Discussion in 'Electronic Projects Design/Ideas/Reviews' started by idiot7.0, Apr 11, 2012.

  1. idiot7.0

    idiot7.0 New Member

    Joined:
    Sep 9, 2011
    Messages:
    23
    Likes:
    0
    Hi,

    I'm trying to design this H-Bridge but it's not functioning in multisim simulator. I would appreciate getting a hint.

    I would prefer using mosfets, but I can't get high side drivers, I'll try later with a charge pump. but actually I'm interested in the darlington circuit. The current is intended to be limited at about 3.5 - 3.75 amps.

    I apologize for the circuit, it's a mess.
     
    Last edited: Apr 11, 2012
  2. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,444
    Likes:
    935
    Location:
    Canada, of course!
    R2 and R5 connect to only +5V so the upper PNP darlingtons will ALWAYS be turned on and will NEVER turn off.
    Connect R2 and R5 to +12V.

    What are the diodes for? Why are there switches at the inputs? Where is the logic input signal and what is it from?
     
  3. idiot7.0

    idiot7.0 New Member

    Joined:
    Sep 9, 2011
    Messages:
    23
    Likes:
    0
    audioguru,

    R2 and R5 are pullups to keep the PNP off.

    The diodes are in the setup to allow only positive signal to get to the darlington. Negative signal is allowed thru the NPN transistor only if the other input is high. This to get a truth table of:
    inputs = outputs
    high - high = 0V - 0V
    high - low = 0V - +V
    low - high = +V - 0V
    low - low = NC - NC

    The input should be from a μC or optocoupler, but the switches are for simulation purpose only.
     
  4. dave

    Dave New Member

    Joined:
    Jan 12, 1997
    Messages:
    -
    Likes:
    0


     
  5. idiot7.0

    idiot7.0 New Member

    Joined:
    Sep 9, 2011
    Messages:
    23
    Likes:
    0

    I placed R2 and R5 as audioguru advised but it wasn't the only problem.

    It's worst, even the logic part isn't functioning as shown in the schematic.
     
  6. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,444
    Likes:
    935
    Location:
    Canada, of course!
    Why does Multisim cross as many wires as is possible? It makes a schematic look like a nightmare.

    Your input transistors should have their emitters at 0V, not floating all over the place. Then a logic high to their series base resistor turns it on and a logic low turns it off.
     
  7. wakoko79

    wakoko79 New Member

    Joined:
    Sep 28, 2010
    Messages:
    110
    Likes:
    0
    I think it would be better if you just use L298 driver. You are using BJTs so I'm guessing your load is pretty low. If your load is high, use MOSFETs instead.

    You don't need a charge pump. It's much easier if you use P-MOS since you only need an NPN inverter at the gate drive (at low frequencies ~ below 20kHz).
     
  8. idiot7.0

    idiot7.0 New Member

    Joined:
    Sep 9, 2011
    Messages:
    23
    Likes:
    0
    Thanks wakoko79,

    I used an L298 in a paralleled setup (I mean in1 connected to in3 and in2 to in4) to supply a max of 4 amps, but it overheats a lot. It's driving a position controlled servo motor at about 30V and drawing 3.26 amps. That's why I prefer using 4 transistors (mosfets) that way, I can substitute parts with higher rated ones if needed.

    I would like using N channel mosfets. I ignore the frequencies being applied (as it depends), but theoretically with the actual setup it can reach about 200,000 updates per second.
     
  9. Grossel

    Grossel Member

    Joined:
    Sep 16, 2008
    Messages:
    918
    Likes:
    22
    Location:
    Norway
    Regular BJT's can't be driven just like any fet.
     
  10. wakoko79

    wakoko79 New Member

    Joined:
    Sep 28, 2010
    Messages:
    110
    Likes:
    0
    200,000 updates per second? Does that translate to 200kHz (one pulse per update)? That is just soooo high. I'm doing a thesis right now and I'm using a P-MOS for high side and N-mos for low side. I'm limiting the input frequency to about 15kHz because the system can't keep up. Well, I have opto-isolation at input that slows the system a lot. But still, 200kHz is damn high. I think you need further research by using N-MOS for high side and low side, but its much more difficult to design.

    I think this will help: http://www.ti.com/lit/ml/slup169/slup169.pdf
     
  11. idiot7.0

    idiot7.0 New Member

    Joined:
    Sep 9, 2011
    Messages:
    23
    Likes:
    0
    What do you mean? voltage vs current driven? I think here it's current driven.

    What's the source of the short that you're seeing? I think the bases are set to switch off the BD transistors. Unless you see current flowing between the emitters and collectors.

    Thanks for the link I'll read it. 200khz is the update frequency of the encoder reading routine. The motor is PWM driven. Then theoretically there is a case where the motor keeps ringing between 2 encoder positions that keeps the h-bridge command updating back and forth. But as mentioned, it depends, on PID parameters, external force applied and limited by inertia ... etc. Then it's a theoretical max that would never happen, but I can't calculate the applied frequency too.

    I know the mistake in my schematic is so stupid so basic, but I'm not an electronic engineer, I'm a programmer. What I fear is having a mistake in the simulation parameters (I added the darlington transistors to the library) rather than in the schematic. Maybe I'll make some chips smoke, it's better than struggling in multisim.
     
  12. ronv

    ronv Well-Known Member Most Helpful Member

    Joined:
    May 30, 2010
    Messages:
    4,725
    Likes:
    488
    Location:
    Tucson, AZ. USA
    Here is one you see quite a bit.
     
  13. idiot7.0

    idiot7.0 New Member

    Joined:
    Sep 9, 2011
    Messages:
    23
    Likes:
    0
    Thank you. I see real shorts unless I'm wrong. Take the case of switching on Q6 thru R9 which itself switches Q4 that itself switches on Q2 thru R11.

    Do you see otherwise ?
     
  14. wakoko79

    wakoko79 New Member

    Joined:
    Sep 28, 2010
    Messages:
    110
    Likes:
    0
    I'm supposing R1 is the simulated load here, it wouldn't make sense otherwise.

    Nope, I see no shorts. The circuit is just fine, it should work.
    You can opt to remove R4 and R5. Just adjust the value R10 and R11.

    If you want, I can send my circuit and you judge it yourself. It's configured to accept TTL level signals. It's not the best but it works well with loads below 10A.
     
  15. idiot7.0

    idiot7.0 New Member

    Joined:
    Sep 9, 2011
    Messages:
    23
    Likes:
    0
    Yes, it would be a pleasure to take a look at your circuit.

    But about ronv's circuit I still see Q1, Q2, Q4 switched on at the same time. What am I missing?
     
  16. ericgibbs

    ericgibbs Well-Known Member Most Helpful Member

    Joined:
    Jan 4, 2007
    Messages:
    21,185
    Likes:
    644
    Location:
    Ex Yorks' Hants UK
    hi,
    If Q1 is hard ON, Vce ~0v, so no bias current thru R11 to turn on Q2
     
  17. wakoko79

    wakoko79 New Member

    Joined:
    Sep 28, 2010
    Messages:
    110
    Likes:
    0
    Here it is. I just hope you'll give credit if ever you use this in a project. =)
     
  18. audioguru

    audioguru Well-Known Member Most Helpful Member

    Joined:
    Mar 16, 2004
    Messages:
    32,444
    Likes:
    935
    Location:
    Canada, of course!
    Please post the schematic as a normal PNG file type instead of whatever the heck it was.
     
  19. Grossel

    Grossel Member

    Joined:
    Sep 16, 2008
    Messages:
    918
    Likes:
    22
    Location:
    Norway
    Well, my eyes can be very dry. Or - you try to convince me that Vbe on a BD679 and 678 can go as high as 12V.

    Or, you can beleive as much as you want, that four diodes in series in forward direction can withstand 12V.
     
  20. wakoko79

    wakoko79 New Member

    Joined:
    Sep 28, 2010
    Messages:
    110
    Likes:
    0
    Sorry, here it is. BTW your comment made me laugh. :LMAO: View attachment 63317

    The inputs are CW and CCW. If you want to drive it, here's an example:
    CW rotation -> CCW = LOW
    -> CW = PWM signal


    BTW, if it was driven that way^, it will run with slow decay (dynamic breaking). If you want to have it run in CW with fast decay mode (coasting) use this:
    CCW = PWM signal
    CW = HIGH

    It has current limit. If you don't want current limit, just remove the cursense terminal at the upper left circuit and change it to ground. I implemented it using a switch.


    Edit: I thought some error occurred and the picture didn't get uploaded. It's on the right side of :LMAO:. The white background of the picture blends well with the thread background haha.
     
    Last edited: Apr 13, 2012
  21. idiot7.0

    idiot7.0 New Member

    Joined:
    Sep 9, 2011
    Messages:
    23
    Likes:
    0
    Thank you, I get it now. I don't know what I was thinking about I think I was assuming Q1's collector as being high.

    Before I adopt this circuit, back to my original design, I transferred it to falstadt circuit simulator, the simulation works better. I found that the hbridge's NPN and PNP transistors were all the time shorted whatever their base state was. Just as Grossel noted. Why is it happening? How could such a thing be remedied ?

    I see what you mean, but even with less than 2.8 or 2.4V between the 4 diodes I'm still getting a short.

    Thanks wakoko79.
     

Share This Page