DRV8307: Issues driving a BLDC motor

user4300350

Other Parts Discussed in Thread: DRV8307, DRV8307EVM, CSD88537ND

Hi there,

I'm using the DRV8307 to drive a 24V rated BLDC motor. I was able to drive it using the DRV8307EVM but i'm having some issues when trying to drive the same motor on my board.

Below is the schematic:

You can't see it on the schematic but LOCKn, FAULTn and HALLOUT are all pulled up to 3.3V (same rail as my MCU)

The red line represents a fly wire to connect VREG to the hall sensors.

When i PWM the driver i see a the FAULTn pin gets asserted and deasserted as shown below. I'm not sure what fault condition is actually being flagged. I probed the charge pump VCP pin and it shows 34.5V. There is no temperature rise, no rotor lockup (i can turn the motor shaft pretty easily)

Any thoughts ?????

Screen capture of the FAULTn pin

over 8 years ago

0 user4300350 over 8 years ago

Intellectual 310 points

By the way, the power MOSFETs are the CSD88537ND.

0 Phil Beard over 8 years ago in reply to user4300350

TI__Expert 4650 points

Hello,

Could you clarify a couple things for me quickly?

1.) What is the VM voltage you're operating at?
2.) What is time scale of the attached oscilloscope plot in your first post?
3.) Does the motor start spinning while the FAULT pin is asserted?

0 user4300350 over 8 years ago in reply to Phil Beard

Intellectual 310 points

Hi Phil,

1.) What is the VM voltage you're operating at?

24V

2.) What is time scale of the attached oscilloscope plot in your first post?

2.5ms/div

3.) Does the motor start spinning while the FAULT pin is asserted?

the motor does not even start spinning

Thanks

0 user4300350 over 8 years ago in reply to user4300350

Intellectual 310 points

I just turned on the motor again and, i noticed that it actually makes a sound (very high frequency sound) then a click as if it wanted to turn on but then stop (it was hard to hear because the room was noisy.)
The PWM frequency is 25KHz.

Thanks

0 user4300350 over 8 years ago in reply to user4300350

Intellectual 310 points

R3 on the schematic (high side fet gate output WHSG) is the only one that transitioning.

0 Phil Beard over 8 years ago in reply to user4300350

TI__Expert 4650 points

Hi,

Thanks for the information. The oscilloscope plot shows FAULT going low for 5 ms, which is the fixed delay for an OCP event.

An OCP event can be triggered either by the voltage being above the VSENSEOCP threshold on the ISEN pin, or the drain to source voltage on the external MOSFETs is greater than the VFETOCP threshold for longer than the tFETOCP interval.

Based on your description above it sounds like the motor is trying to pull to much current during startup.

Can you measure the Drain to source voltage across each of the MOSFET's in your circuit during startup?

Can you measure the voltage across ISEN during startup?

0 user4300350 over 8 years ago in reply to Phil Beard

Intellectual 310 points

Can you measure the Drain to source voltage across each of the MOSFET's in your circuit during startup?

The high side FETs are about 23.3V

The low sides (Q1: 2.8V, Q2: 0.1V, Q3: 1.4V)

Can you measure the voltage across ISEN during startup?

The voltage across R5 is 0.049V, which means i'm drawing less than 1A of current through it.

0 Phil Beard over 8 years ago in reply to user4300350

TI__Expert 4650 points

Can you measure the gate voltage for V_LS? Also, are you measure 23.3V between the V_MOT net and each U / V / W net?

0 user4300350 over 8 years ago in reply to Phil Beard

Intellectual 310 points

Can you measure the gate voltage for V_LS?

ULSG: 8.28V

VLSG: 9.8V

WLSG: 0V

Also, are you measure 23.3V between the V_MOT net and each U / V / W net?

Yes they go from 23.3V to ~22, 8V

0 user4300350 over 8 years ago in reply to user4300350

Intellectual 310 points

These are the screen capture on

WHSG

0 Phil Beard over 8 years ago in reply to user4300350

TI__Expert 4650 points

Hello,

Thank you for the additional information. Can you run the following tests on your board?

1.) Can you probe the HW+ / HV+ / HU+ pins with the halls from the motor connected and VREG up? You should not see all the halls logic-high or logic-low at the same time.

2.) Without the three motor phases disconnected from your board and the halls powered up from VREG, can you turn the motor with your hand and probe the HW+ / HV+ / HU+ pins? You should see them toggling as you manually turn the motor. You should not see all the halls logic-high or logic-low at the same time.

3.) What is the measured logic-high / logic-low voltage for each of the HW+ / HV+ / HU+ pins?

4.) What is the voltage measured across R7? This should be approximately 2.5V based on your schematic while VREG is up.

0 user4300350 over 8 years ago in reply to Phil Beard

Intellectual 310 points

1.) Can you probe the HW+ / HV+ / HU+ pins with the halls from the motor connected and VREG up? You should not see all the halls logic-high or logic-low at the same time.

With the motor winding connected, below is the screen capture of HW+ / HV+ / HU+.

ch1: HU+

ch2: HV+

ch3: HW+

2.) Without the three motor phases disconnected from your board and the halls powered up from VREG, can you turn the motor with your hand and probe the HW+ / HV+ / HU+ pins? You should see them toggling as you manually turn the motor. You should not see all the halls logic-high or logic-low at the same time.

With the motor winding disconnected, HW+ / HV+ / HU+ stay low when i manually turn the motor

3.) What is the measured logic-high / logic-low voltage for each of the HW+ / HV+ / HU+ pins?

V_HU+ = 0

V_HV+ = 5V

V_HW+ = 5V

4.) What is the voltage measured across R7? This should be approximately 2.5V based on your schematic while VREG is up

2.5V

Note: On the DRV8307EVM i was able to replicate this issue by switching the hall sensor pins (it seems as if there is a sequence that works better.) So, i thought i'd try different hall sensors connections on my board to see if one of them work but none worked.

0 user4300350 over 8 years ago in reply to user4300350

Intellectual 310 points

I still haven't figured out what's going on. Any other ideas from your end ?

0 Phil Beard over 8 years ago in reply to user4300350

TI__Expert 4650 points

Hi,

The hall sensor pins should be connected based on the motor datasheet.

Are you able to see the three H+ pins toggling between different states if VREG is up and you manually spin the motor? Do this without supplying a PWM signal to the driver.

Is your power supply current limiting during motor startup with your board and not during the EVM?

0 user4300350 over 8 years ago in reply to Phil Beard

Intellectual 310 points

It turns out after hours troubleshooting, the motor hall sensors required pull up resistors. After adding the modification, the motor runs fine now.

Thanks once again for your help, and assistance with this.

0 Phil Beard over 8 years ago in reply to user4300350

TI__Expert 4650 points

Not a problem, glad you figured it out.

Motor drivers

Motor drivers forum

DRV8307: Issues driving a BLDC motor