I have 200W DC motor application, with input voltage of 36V connected with Schottky Diode as input voltage to DRV8332. This same input voltage,
is connected to buck=step-down controller TPS54260 to generate 12V for DRV. BLDC or DRV8432 controller is "perfect" solution for high current DC motors, because:
OUT_A and OUT_B are controlled with "same=parallel" PWM input (using parallel branches / conections for high current inductors/coils).
The Third output, OUT_C is used/connected as "direction"/*2-nd port/2-nd pin of Motor for ability to rotate left and rotate right.....
So function (motor rotation) in one direction is simple straightforward, OUT_C is forced to GND, and microcontroller generates PWM, which
generates on OUT_A = OUT_B (+LC filter) PWM voltage, which defines speed (Speed=Linear Function od PWM dutycycle) of this motor rotation...... Easy!
BUT real challenge is to rotate motor on opposite direction. I dont mean to modify CPU program to change PWM dutycycle (1-D) to get same linear speed,
BUT Issue, which I have tryied with partial SUCCES is to force OUT_C to force HIGH for loooong (infinity) time !!!!
FIrst, you need to have input voltage + cca 12V. Solution: Take TPS54260 + charge pump (2capacitors+2..3diodes) and you have it !
Second: It is necessary to "recharge/update" Boost capacitor for OUT_C. (Because it is decharged internally with time =some leackage!)
bad partially working solution is to use bipolar transistor current source (supplied from "top"), and connect 12V zener Diode in parallel with Boost Capacitor.
This "solution" is strongly not recommended!!! and see details bellow !!! (I have killed after some time (reason nknown), and some "normal PWM" operations DRV8332)
Second - better solution is to use comparator, and if voltage on Boost capacitor is higher than Vin (this happens when OUT_C is going up (reaching Vin), than
extra transistor is connecting resistor between Boost capacitor, and between Vin+12V (see charge pump details above). I have good succes here,
but this solution need a litle bit more schematic tuning/time/optimising as:
- speed of comparator is critical (to stop Boost capacitor charging, if OUT_C is already near GND!)
- there is more option how to measure OUT_C being HIGH. Either sensing resistor divider between VIn+12V and OUT_C, or divider between Boost Capacitor and Vin+12V or directly voltage on Boost Capacitor.
- carefullness is needed for operation with floating OUT_C.
- (zener diodes are not so easily matched with 12-13V max, and are not power current strong-enough).
Please, Questions for Motor Control Team:
- what is current consumption (typical, corners min/max, temperature min/max) for Cboost, when OUT_C is HIGH???? I have measured cca 1,5mA
(this number is needded to design reliable VIn+12V supply, and reliable supply for Boosting Capacitor)
- what is connected to Cboost pin internally? I ask because behaviour and function in NOT described in Datasheet AT ALL. I have only measured,
that mostprobably during active RESET_C (as PWM input signal) Boost Capacitor is connected to 12V. But it is not clear how, because with HAlf bridge disconnected, I should be able
to rotate wth motor (as Generator) and to Apply ANY !!!! voltage in range of 0...upto Vin on OUT_C. So from this point, is TI making wrong assmption either for DRV5432 or for DRV8332 ? (mentioning back EMF? to sense rotation speed??) This is right question, helping to solve puzzle how the Boost Capacitor is connected internally.
I needed for my application to go from OUT_C unconnected=floating directly to OUT_C = HIGH operation. unfortunatelly, I was not able to do it, even if I forced correctly digital inputs!!?! last, third question for Motor Application Team: is it caused somehow by internal state-machine sequence? because I have succes with sequence unconnected -> GND -> HIGH- How the Boost Capacitor is connected internally, and which minimal mandatory STATES are present internally? I have measured with
osciloscope some 700kHz pulses - are those regular pulses switching (reconfiguring) symptoms from internal state machine clock? It sounds logical, first you charge
boost capacitor, THAN you connectthis Capacitor somewhere else, AND THAN you.... (this is description of state machine connectivity...)
Those informations (missing in Datasheet) are needdeed for this "simple" task, to drive DC motor left and right with variable velocity...............
Is it possible to get those? Or if there will be some Application Note, how to solve it officially??
Thanks, Marek Stepan