I can't remember off hand if the L298 has an internal shoot-through prevention in place? If not, that would be one of the potential pitfalls in your scheme. You have to have some dead time between one leg of the bridge turning off and the other turning on (FET turn off and turn on times are not the same ...you might recall the theory about minority carriers lifetime in FETs from semiconductor physics). If you do not have shoot through prevention in place, the other leg of the bridge is going to turn on before the first leg has fully turned off leading to a direct short between the power rails leading to the heating you observed in addition to other nasty EMI stuff. The inverter you have between the two control signals to the H-Bridge does not do this for you (a HC chip has a propagation delay of a few tens of nano seconds and you really need a dead time in the microsecond range-check the datasheet for the difference between min turn on and max turn off times and make the deadtime larger than this figure). If the L298 does not have built in shoot-through prevention you would need to do this in software. Make sure to turn off and wait before turning on. This means the time one leg is on is slightly smaller than the time it is off even at 50% duty cycle. You can't simply put a delay line along with an inverter since you will then be delaying turn off in on leg half the time. Regards Madhu >-----Original Message----- >Hi all - > >I'm having trouble getting "locked anti-phase" drive to work with an >L298 H-bridge - seeing way too much (I think) current when the motor is >stopped. Has anyone accomplished this, or else know why it won't work? > >I'm trying to drive a small gear motor using an L298 h-bridge. I want to >use "locked anti-phase" drive (eg. the method used in all the PIC servo >application notes) where the PWM signal is applied to one of the inputs, >and through an inverter to the other; the enable is held high. This way >you just need one PWM pin - a 50% duty cycle should result in no current >in the motor, due to its inductance, as long as the PWM frequency is >above the motor's time constant. > >But this is not what I'm seeing - at 50% duty cycle, the motor does >stand still, but I'm reading a *lot* of current (measured at the Vs >motor supply terminal of the chip) and the chip gets hot. With a 1kHz >PWM frequency, the stand-still current is about equivalent to the motor >running under load. At higher frequencies, the stand-still current goes >up - way up! I expected the current to go down (close to zero) at higher >frequencies. > >What's going on? Is this normal? Shouldn't I be seeing near-zero current >at stand-still with locked anti-phase (I've only tried it with the >L298)? Is the L298 not suited to this drive mode? Any advice? > > >Here's some of my test data, comparing enable chopping (run/coast), >input chopping (run/brake) and anti-phase (run/reverse) drive modes. >Current listed is for the motor running about half-speed, 100rpm, with a >mainly intertial load, which is usually the maximum current reading. The >current measurements also seem to correspond to chip temperature >readings I took. > >125hz >enable - 100mA >input - 230mA >antiphase (didn't test) > >1khz >enable - 100mA >input - 110mA >antiphase - 113mA (125mA at zero speed) > >4khz >enable - 105mA >input - 110mA >antiphase - 135mA (135mA at zero speed) > >16khz >enable - 70mA >input - 170mA >antiphase - 410mA (500mA at zero speed) > >