BOOSTXL-DRV8323RS: DRV8323RS drive question

Hello, and thanks for your question.

What are your gate drive settings (IDRIVE) set via the SPI interface? We often see that if IDRIVE is set too high it can result in improper operation or damage to the device. [FAQ] Selecting the Best IDRIVE setting and Why this is Essential. Based on your MOSFETs we would recommend an IDRIVE setting of 30mA or 60mA source and 60mA or 120mA sink.

#1:  Low-side gate drive voltage,it vary from 11V  to 15V , as the datasheet refer that  the voltage is fixed at 11V , so why ?

The low-side gate drive voltage is generated internally with an LDO. This voltage has a typical value of 11V and can be less if the load is higher (larger MOSFETs or higher PWM frequency). The voltage can also be more if there is a "pumping event" on the MOSFET gates. If there is significant ringing on the GLx or SPx, you can inject current into the internal regulator and pump the voltage up.

#2:  High-side gate drive voltage,it vary too,check the waveform

We expect that the high-side voltage waveform can vary if several phases are switching at the same time. This results in some ripple on VCP - Can you measure to see if we can observe any ripple on VCP? You can plot VCP, GHA, GHB, and GHC. The ripple should not be an issue as long as it is <1V. Also check to be sure that VM is not varying since VCP is referenced to VM.

#3:the MOS Drive  Rising edge oscillation ，how to solve the problem   as I have set the register  IDRIVEP_LS   and  IDRIVEN_LS   to  0000b, the drive current is also smallest ;

There are two possible sources to this oscillation:

• Parasitic inductance in the low-side source path (i.e. sense resistor and GND return path)--> I recommend tying C15, C16, and C17 to the low-side MOSFET source directly rather than to GND, since this can help mitigate oscillation on SPA, SPB, and SPC
• Parasitic inductance between the high-side and low-side MOSFETs --> Snubbers will be required across each MOSFET to mitigate this.

#4: the high side -MOS close delay long,how to solve ?

MOSFET rising edge can be controlled by:

• When current is flowing OUT OF the phase: IDRIVEN_LS (low-side sink current) --> IDRIVEN_LS looks to be very quick already
• When current is flowing INTO the phase IDRIVEP_HS (high-side source current) --> Can you increase the IDRIVEP_HS?

#5: the following is our schematic ,pls check and share some suggestions

Comments from above:

1. I recommend tying C15, C16, and C17 to the low-side MOSFET source directly rather than to GND, since this can help mitigate oscillation on SPA, SPB, and SPC
2. If The above point does not work, snubbers will be required across each MOSFET to mitigate ringing.
3. Make sure that the voltage rating of C11 is the same as the rating of C14.
4. I would recommend adding placeholders for filtering capacitors on ISENSEC and ISENSEB, just in case they are need.

Thanks,

Matt

Hello again!

I have a few pictures and animations that may be helpful: GateDriverSwitchingBehavior_020221.pptx

Since you are measuring at the high-side MOSFET VDS (VDRAIN-SHx) rather than the low-side MOSFET VDS (SHx-GND) your plots are the opposite of mine.

Your first shot is similar to my page 3: the high-side gate turn OFF is controlling the switching event (turning OFF of the high-side MOSFET, the low side body diode catches and HS VDS exceeds VDRAIN). In this case the low-side MOSFET goes from VDS = VDRAIN to VDS = -0.7V when the HS gate turns OFF & the low-side body diode catches through reverse recovery (Qrr = 145nC, trr = 82ns from the CSD18540Q5B datasheet). Next the low-side MOSFET turns ON and the low-side MOSFET VDS goes from -0.7V to ~0V very quickly.

Your second shot is similar to my page 5: the low-side gate turn ON is controlling the switching event (turning OFF of the high-side MOSFET, the high side body diode catches and HS VDS goes negative). In this case the low-side MOSFET goes from VDS = VDRAIN to VDS = VDRAIN+0.7V when the HS gate turns OFF, and then to VDS = 0V when the low-side MOSFET turns ON. Body Diode reverse recovery is not a factor in this case.

Thanks,

Matt

Hi,

Your case is going in the reverse direction compared to the article. So the same process is followed but in the reverse direction.

The diode is recovering from the reverse bias condition (going to the forward boas condition), and the diode requires some current & time to overcome the Qrr.

In your case (image from your previous post below):

• Current is flowing out of SHx into the motor
• The low-side MOSFET & body diode start in the OFF state when the high-side MOSFET is ON.
• The high-side MOSFET turns OFF, at this point the VDS voltage across the low-side MOSFET  and body diode is +24V
• Since current needs to flow OUT of SHx into the motor, the low-side body diode needs to turn ON to conduct current
• However, the low-side MOSFET body diode is starting with a reverse bias at 24V with no current flow
• The body diode will then start to recover from the reverse bias condition and move towards the forward bias condition (this requires current to overcome the Qrr)
• In your image below, the low-side MOSFET gate is turned ON before the reverse recovery is completed. The MOSFET starts to conduct before the body diode drops all the way to zero. Once the MOSFET turns ON, the VDS drops to about 0V.

   - YEL : Low-Side MOS VDS    BLU:Low-Side MOS VGS      Pink: High-Side MOS VGS

Thanks,

Matt