DRV8704: Ringing Noise

NAOKI AOYAMA

Guru 11215 points

Part Number: DRV8704
Other Parts Discussed in Thread: CSD18540Q5B,

Hello,

Our customer use the DRV8704.

The customer evaluate it on their board.

It occur ringing noise.

They use the CSD18540Q5B as external FET.

The configuration register is default setting.

I attach the file include the scopeshots.

DRV8704_noise.pptx

What is the cause of this noise? Wrong layout?

May this noise damage the FET?

Best Regards,

Naoki Aoyama

over 4 years ago

0 Hector Hernandez Luque over 4 years ago

TI__Expert 8980 points

Hi Aoyama-san,

Are the captures taken upon driver startup or is it upon driving A1HS?

Can you provide captures of the PWM signals and the other gate signals?

Has the result been conducted on the DRV8704EVM as well to isolate the issue to layout differences?

Is a motor load utilized?

The FET should not be damaged.

It is worth noticing that, the FET Vth =1.9 V, which is reached while Vg slews to ~3.5V. The spike seen in the current during the slew on it and on the other signals appear to be close to that 1.9 V.

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hello Hector,

Thank you for your reply.

>>Are the captures taken upon driver startup or is it upon driving A1HS?

It is upon driving A1HS. They observed the motor current at the point between FET common source and GND.
They use short break, so the current is observed "zero" during short break.

>>Can you provide captures of the PWM signals and the other gate signals?

I will ask the customer to get another captures. Which points do you want to check?

>>Has the result been conducted on the DRV8704EVM as well to isolate the issue to layout differences?

No. I will ask the customer to check the EVM.

>>Is a motor load utilized?

Yes, it is.

>>It is worth noticing that, the FET Vth =1.9 V, which is reached while Vg slews to ~3.5V. The spike seen in the current during the slew on it and on the other signals appear to be close to that 1.9 V.

Sorry, but they use the 10:1 voltage probe. They use the VM as 24V.

Best Regards,
Naoki Aoyama

0 Hector Hernandez Luque over 4 years ago in reply to NAOKI AOYAMA

TI__Expert 8980 points

Hi Aoyama-san,

Replies inline to your points:

>>Can you provide captures of the PWM signals and the other gate signals?

-> I will ask the customer to get another captures. Which points do you want to check?

>>> IN1, IN2 and AINSENP.

>>Is a motor load utilized?

-> Yes, it is.

>>> Can the customer verify without a motor load to see if the driver's output signals behave correctly? Can the customer share the motor information?

>>It is worth noticing that, the FET Vth =1.9 V, which is reached while Vg slews to ~3.5V. The spike seen in the current during the slew on it and on the other signals appear to be close to that 1.9 V.

-> Sorry, but they use the 10:1 voltage probe. They use the VM as 24V.

>>> It still appears that the spike happens when the FET is turned on. Is this spike only seen the first time A1HS is turned ON? What about when PWMing is occuring? Is the spike gone?

0 Hector Hernandez Luque over 4 years ago in reply to Hector Hernandez Luque

TI__Expert 8980 points

Hi Aoyama-san,

I have brainstormed with a colleague and we have a theory:

Given the speed at which the FET is slewing from low to high, the low side and high side FETs of the A1 bridge might be turning ON. This is possible due to turning on the high-side FET so fast that the the high-side FET is energized and, due to the gate to drain capacitor of the low side FET, the gate of the low side FET is pulled up enough to turn on the low side FET while the high side FET is on.

If possible, please share a capture of A1HS, A1LS, AOUT1, and current to see what is happening on the gate of the low side FET during this test. Ignore my previous capture request.

If the A1LS voltage goes up above the FET's threshold voltage, 1.9 V, then the problem is what we established here. I solution is to increase the TDRIVEP setting to a larger value and decreasing the IDRIVEP value.

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hello Hector-san,

Thank you for your reply.
I will ask the customer to observe A1HS, A1LS, AOUT1 and motor current.

For your information, the specification of the motor is as follows,
R= 1.67 ohm , L=3.46mH

Best Regards,
Naoki Aoyama

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hi Hector-san,

I got the scopeshots.

Sorry, but the customer changed the setting from default.

Now, TDRIVEP/N : 2.1us, IDRIVEP : 100mA, IDRIVEN : 200mA.

I attach the scopeshots.

DRV8704_noise_ver2.pptx

It seems to disappear the noise of HS FET. But the noise of LS FET is still there.

Best Regards,

Naoki Aoyama

0 Hector Hernandez Luque over 4 years ago in reply to NAOKI AOYAMA

TI__Expert 8980 points

Hi Naoki-san,

Can the customer run IDRIVEP: 50 mA? Please, inform me when the customer runs tests with EVM as well.

0 Hector Hernandez Luque over 4 years ago in reply to Hector Hernandez Luque

TI__Expert 8980 points

Additionally, can the customer share register settings?

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hi Hector-san,

I will asked the customer to set the IDRIVEP:50mA.

Regards,
Naoki Aoyama

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hello,

I will ask the customer to provide us the register setting.

Regards,
Naoki Aoyama

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hi Hector-san,

I got the scopeshots and register setting.

I attach them.

DRV8704_noise_ver3.pptx

Register_setting.xlsx

Could you check it?

Best Regards,

Naoki Aoyama

0 Hector Hernandez Luque over 4 years ago in reply to NAOKI AOYAMA

TI__Expert 8980 points

Hi Naoki-san,

Is it possible for the customer to share a schematic?

If so, let me know so I can inform you on how to share it.

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hi Hector-san,

Yes, it is possible.
How to share it?

Regards,
Naoki Aoyama

0 Hector Hernandez Luque over 4 years ago in reply to NAOKI AOYAMA

TI__Expert 8980 points

Hi Naoki-san,

I sent you a friend request. Let's start a private conversation for me to receive the files.

0 Hector Hernandez Luque over 4 years ago in reply to Hector Hernandez Luque

TI__Expert 8980 points

Hi Naoki-san,

Thank you. I am reviewing the files.

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hello Hector-san,

Thank you for your a lot of support.

I will suggest the customer to correct the schematic and layout.

Thanks,

Naoki Aoyama

0 Hector Hernandez Luque over 4 years ago

TI__Expert 8980 points

Hi Naoki-san,

If the error is gone on the EVM but still relevant, then ask the customer to do the corrections we discussed and to follow the EVM layout at a baseline for their layout revision.

0 NAOKI AOYAMA over 4 years ago in reply to Hector Hernandez Luque

Guru 11215 points

Hello Hector-san,

I see.

I will suggest to the customer to refer to the layout of the EVM.

Thanks,

Naoki Aoyama

Motor drivers

Motor drivers forum

DRV8704: Ringing Noise