H-Bridge Transistor options.

Discussion in 'General Electronics Chat' started by geoffers, Jul 7, 2012.

  1. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Hi All,
    Hope someone can help on this? I've built a couple of H-bridges to drive the coil on my 125khz rfid reader. The circiut is one I've pinched from another post on here, the first one I made used 2n2222's and 2n2907 transistors for the H with 2n2222's to switch them. All well so far but I was after a bit more current, not knowing as much as I should I got some tip121's and 126's. Still using the 2n2222's to drive them. This doesn't work! A quick search found out that darlington power transistors take some time to turn off! I guess what is happening is both sides of the H are 'on' at the same time? The data sheet for the tips isnt to helpfull but I assume they arnt suitable for my project? Transistor appear to be a minefield to me, can anyone point me in the right direction for a transistor pair to use? I need to switch more current than the 2n2222's will (800mA i think) they were getting quite warm!
    Thanks Geoff
     
  2. mhastie1234

    New Member

    Feb 10, 2012
    29
    2
    Post a schematic. I have made an hbridge with tips, they do work.
     
  3. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    [​IMG]
    Here is the schematic I 'pinched'! The 2n5551's are replaced with 2n2222's and the mje2955t with tip126 and mje3055t with tip121. I think my issue is the time it takes for the tip's to turn off? a 125khz square wave is driving the bridge and I think the tips are 'on' at the same time? I've had a hunt about and orderd some bd243c's and bd244c's. Hopefully these should work?! Any thoughts?
    Thanks Geofff
     
  4. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Hi all,
    My transistors have turned up! Using the circuit in my last post with the bd243 and bd244's and I should say minus the diodes as I'm driving a LC circuit. I tested the bridge just connecting the input to + and gnd and got +/- across the centre, all well so far! However when I drive the bridge with a 125khz square wave the output from the bridge is 'peaky' also with no load the transistors start to heat? I can only assume both transistors on one side of the bridge are 'on' for a overlaping period? Hence a short and the heat? Is there a way to overcome this? (the npn transistors switching the bridge are 2n2222's) The square wave is at the moment generated by a u2270b I could use a pic to drive the bridge and build in some 'dead' time but when I already have a driver this seems over complicated:(! Maybe I need to review which transistors I use?
    Any help would be welcome.
    Thanks Geoff
     
  5. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Hi all,
    Been looking at this again, put my oscilliscope on it, the square wave driving the bridge isn't quite square, it takes 1uS from full - to full+, the ouput from the bridge takes 3uS to do the same, I assume this means the bridge transistors are both 'on' at the same time, I've tried tying the base and emmiters with a resistor but it makes little difference, any thoughts?
    Thanks Geoff
     
  6. mhastie1234

    New Member

    Feb 10, 2012
    29
    2
    Have a look at "help building a h-bridge for a stepper motor" thread he is using the same circuit with the same problems
     
  7. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Hi,
    Thanks, I must admit its that post I got my initial circuit from. I think I have slightly different issues though? As a motor driver I think my circuit would work ok? But as I'm trying to create a driver for a RFID coil at ~125 to 134kHz its that giving me problems. The square wave driving the bridge has a duration of about 7uS, the output from the bridge as its is has peaks 7uS apart with a slope that takes 3uS from 0v to 12v. (Hope I'm explaining this ok) I'm guessing this means both transistors are on at the same time for nearly half the time? Hence the heat with no load? I'm sure there must be a suitable transistor pair? I ought to say I'm 'self taught' so there is alot I dont know!:confused:
    Thanks for looking, Geoff
     
  8. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Hi again,
    Been searching a bit on the transistors I'm trying to use, I couldn't easily find turn on times for the 2n2222, but its turn off time is ~250nS, the turn on time of the bd243/bd244's is 300nS and turn off time 1uS (have I understood it right that load affects turn off time?) so does this mean I have a 700nS overlap on my bridge? If so that means the bridge is shorted 10% of the time? (7uS square wave) In which case I suspect the peaks are my supply 'recovering'. Does anyone know the best way to remidy this? Do I just have to search for faster transistors?
    Thanks Geoff
     
  9. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Your circuit does not have resistors from base to emitter on the transistors to turn them off quickly. So the transistors slowly "float" off. At your high frequency the transistors never completely turn off.
     
  10. PaulEE

    Member

    Dec 23, 2011
    423
    32
    On top of that brilliantly simple but effective and fundamental point, you might also consider implementing logic that only allows the diagonal transistors on at one time. If two are on, even slightly, at once, they'll heat up and fizzle, as you've seen, in no time.
     
  11. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Thank you for having a look at this, I had already tied the base and emmiters of the power transistors together with 100kohm resistors, should I do the same with the 2n2222's? I have a uC in the circuit I'm using but its already quite 'busy' interprating the rfid signal. I could fairly easily programme up another one to drive the H-Bridge, I suppose I'd have to put in 1.5-2uS of off time on both sides of the bridge? Transistors are a bit of a mystery to me at the moment, there are so many to chose from! Are the Bd243c/bd244c pair a good choice or should I try somthing else? Idealy I'd like to be able to switch 2A or so at 20ishV, am I being to optimistic or is there a suitable pair or different configuration or circuit that will do what I want, ie take a 5v square wave signal and amplify to 20v at 125-134khz
    Thanks Geoff
     
  12. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    A 100k resistor between the base and emitter will create a current of only 0.7V/100k= 0.000007A which is not enough to turn off a transistor quickly. Use 470 or 220 ohms instead.
     
  13. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Thank you for the advise,
    I put 470ohm resistors between base and emitter's and they wouldn't turn on then, went to 4k7 and the wave looks the same as before, should I play about between the two values? I have 1.5k resistors in series with the 2n2222's. I've been wondering if theres any way I could put a cap in that had to charge before the transistors turned on? As a sort of delay, no idea if this is even possible but thought I'd ask the question!
    Cheers Geoff
     
  14. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Sorry, I looked at your WRONG schematic not at the correction in your text.
    You are using very slow Darlington transistors that already have base-emitter resistors inside, not ordinary transistors that are shown on the schematic.

    Why didn't you correct the schematic you posted??
     
  15. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Sorry, I should have posted a redrawn schematic, I thought the schematic would give enough of a idea of my problem but can see its confusing. If the darlington pair I'm using aren't suitable could you point me in the right direction for a pair I could use? I know this is basic stuff but if you could help it would me much appritiated.
    Thanks Geoff
     
  16. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Darlington power transistors switch slowly, I don't know if there are faster ones.

    Why not use Mosfets? They switch extremely fast when their driver can quickly charge and dischage their high gate capacitance.
     
  17. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    Thanks I've just had a quick search on the rs website and that does look like the solution. I've had a read on the mosfet section of this website. Think I could do with reading up some more. Could I use n and p channel mosfets in the same configuration as H-bridge in the schematic? Or would they not work like that? I did notice there are many more n channel avalible than P? Is there a reason for this? Would you have a suggestion for which ones to use or shall I search and guess! I would like to switch 2-3A at around 20V. Looking at the specs on rs that shouldn't be a problem.
    Thanks again for your input and suggestions, Geoff
     
  18. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    This would appear to be quite usefull? Not quite as much current as I hoped but more than the u2270b's drivers!
     
  19. geoffers

    Thread Starter Active Member

    Oct 25, 2010
    239
    6
    I've been looking at the data sheet for the mosfet h-bridge chip, as I'm driving a coil and want it to resonant will the diodes built into the bridge affect it? I want to maximise the magnetic field produced, I guess those diodes are there to protect the chip from flyback on a electric motor?
    Cheers Geoff.
    Mosfet h bridge block.jpg
     
  20. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The diodes are part of the manufacturing steps that make as Mosfet. They cannot be removed. If your resonant signal level is a little higher than the power supply voltage or a little lower than ground then the diodes will conduct and they turn off very slowly.
     
Loading...