Hello to all,
I describe the problem we notice trying to use a Texas DRV8353RH-EVM card to drive a 48V BLDC motor.
This engine in our application can take up to 25 amps, but on the lab bench where we are using it now we normally do not exceed 3 amps with a 100% PWM.
We purchased this card to test the Texas controller as a possible alternative to the one we use today: a Rohm BD62012FS.
We have created connections between the EVM and a working system equipped with the BD62012FS, of which we are therefore sure that the motor with its Hall sensors and the 48V supply are correct.
Since the card is configured to work with an additional card equipped with a CPU and not autonomously with the Hall sensors, we have made the appropriate modifications to the EVM itself:
• mounted 0 Ohm 0603 in R48, R50 and R52 (as indicated by Texas for reading Hall sensors)
• removed R22 and R23 to put the MODE pin in floating, in order to set the 1xPWM mode
We have therefore set the command signals: §
• a 20 kHz PWM at 3.3V on pin 2 of J2 (the signal is the one we use on our current board and in this configuration makes a slow ramp from 0 Hz to 25 Hz)
• DIR and nBRAKE at 3.3V (pins 10 and 12 of J2)
• ENABLE (pin 10 of J1) switchable to 0V / 3.3V with one of our switches and the power supply:
• 48V dc (the system works on batteries)
During our attempts, with this configuration, we were able to run the engine for a fraction of a second because the board goes continuously in the FAULT state (there is a red LED that lights up).
We noticed, among other things, that even in static conditions (immediately after a power on and without giving any PWM) the FAULT even seems to depend on the state of the Hall sensors:
• with Hall ABC = 100, 010, 011, 101 it is not activated and, if by chance it was active, with an ENABLE pulse at 0V it is deactivated
• with Hall ABC = 110 or 001 it is activated and cannot be deactivated with an ENABLE pulse at 0V (when we reset the signal to 3.3V the FAULT is reactivated immediately)
Sometimes in the FAULT state we noticed the chip warming up more than usual and the voltage on one of the low side gate was about 5 volts.
In other cases, when the chip did not heat up more than normal, we saw the three gate voltages at values quite different from each other: 12.0 Volt on one and 10.6 on the other two.
As a last attempt, fearing to have some cables a bit long (but they are those of the configuration now functioning), we mounted the resistors and the filter capacitors in correspondence of the 6 MOSFETs (for example C2 and R4), but without significantly varying the behavior.
I therefore hope to be able to receive useful information on how to correct a possible mistaken configuration or maneuver.
Thanks in advance for your help.
Best regards, Andrea
I describe the problem we notice trying to use a Texas DRV8353RH-EVM card to drive a 48V BLDC motor.
This engine in our application can take up to 25 amps, but on the lab bench where we are using it now we normally do not exceed 3 amps with a 100% PWM.
We purchased this card to test the Texas controller as a possible alternative to the one we use today: a Rohm BD62012FS.
We have created connections between the EVM and a working system equipped with the BD62012FS, of which we are therefore sure that the motor with its Hall sensors and the 48V supply are correct.
Since the card is configured to work with an additional card equipped with a CPU and not autonomously with the Hall sensors, we have made the appropriate modifications to the EVM itself:
• mounted 0 Ohm 0603 in R48, R50 and R52 (as indicated by Texas for reading Hall sensors)
• removed R22 and R23 to put the MODE pin in floating, in order to set the 1xPWM mode
We have therefore set the command signals: §
• a 20 kHz PWM at 3.3V on pin 2 of J2 (the signal is the one we use on our current board and in this configuration makes a slow ramp from 0 Hz to 25 Hz)
• DIR and nBRAKE at 3.3V (pins 10 and 12 of J2)
• ENABLE (pin 10 of J1) switchable to 0V / 3.3V with one of our switches and the power supply:
• 48V dc (the system works on batteries)
During our attempts, with this configuration, we were able to run the engine for a fraction of a second because the board goes continuously in the FAULT state (there is a red LED that lights up).
We noticed, among other things, that even in static conditions (immediately after a power on and without giving any PWM) the FAULT even seems to depend on the state of the Hall sensors:
• with Hall ABC = 100, 010, 011, 101 it is not activated and, if by chance it was active, with an ENABLE pulse at 0V it is deactivated
• with Hall ABC = 110 or 001 it is activated and cannot be deactivated with an ENABLE pulse at 0V (when we reset the signal to 3.3V the FAULT is reactivated immediately)
Sometimes in the FAULT state we noticed the chip warming up more than usual and the voltage on one of the low side gate was about 5 volts.
In other cases, when the chip did not heat up more than normal, we saw the three gate voltages at values quite different from each other: 12.0 Volt on one and 10.6 on the other two.
As a last attempt, fearing to have some cables a bit long (but they are those of the configuration now functioning), we mounted the resistors and the filter capacitors in correspondence of the 6 MOSFETs (for example C2 and R4), but without significantly varying the behavior.
I therefore hope to be able to receive useful information on how to correct a possible mistaken configuration or maneuver.
Thanks in advance for your help.
Best regards, Andrea
