TMS320F28379D-Q1: Stop generating PWM signals when high dv/dt occurs

Hello, 

We are working on finding the right expert for this question. Please expect a reply in the next business day. 

Thanks, 
Anu

Hi Anu,

Thank you for your response. We will wait for the updates early next week.

Regards,

Danilo

Hi.

Can you tell me what was the dv/dt?

I have tested up to 100 V/ns in TI reference PFC design with F280049C controlCard and did not see the issue on the setup/

Best,

John  

Hi John,

According to our customer,

The dv/dt in my case is 50V/ns, I connect it totally with three 20 cm jump wire(PWML,PWMH,AGND). I also tested it on GaN evaluation board (gansystems.com/.../gs-evb-hb-66508b-on1). Still could not run it.

Regards,

Danilo

Hi Danilo,

It is always best to keep the shortest signal path from MCU to gate driver or device. jump wire it self has some inductance and therefore, it is susceptible to the noise. Does GaN evb has the same pwm jump wire?

Hi John,

Here is the feedback from our customer,

I could not configure why long PWM inductance could lead to DSP failure. My DSP is totally floating from any earth (powered by battery). If you interested, you could see Fig.15 in (www.onsemi.com/.../EVBUM2641-D.PDF). In this figure, the evaluation board is connected to controller with jump wire (the same as I used). My difference is I use DSP instead of the LLC controller IC, so I think the failure might relevant to MCU difference.

PS: my jump wire is thicker and slightly looser than that in PDF, but will it matter that much?

Regards,

Danilo

Hi Danilo,

The link provided says the pdf could not be found. Could you please try to attach the pdf directly?

Is this seen on multiple F28379D controlCARDs or just one? Is the pwm trip zone submodule configured and possibly causing the pwm outputs to go low under the experienced conditions? Can you please provide the pwm configuration so I can take a look and ensure proper initialization?

Best Regards,

Marlyn

Marlyn Rosales Castaneda20 said:

The link provided says the pdf could not be found.

There was an extra ')' at the end of link. Try EVBUM2641 - NCP51820 HB GaN Driver Evaluation Board User's Manual (onsemi.com)

Hi Marlyn,

I apologize for the mistake in the link. Here is the correct one. 

https://www.onsemi.com/pub/Collateral/EVBUM2641-D.PDF

I have requested a photo of the customer's set up and he sent this.

Regards,

Danilo

Hi Danilo,

The setup looks okay, can you please help provide the PWM configuration? 

Best Regards,

Marlyn

Hi Marlyn,

We just received this feedback from our customer. 

I make a measurement. It failed even if 20Vdc applied. In this case, the dv/dt is not very high. It's strange to explain the reason behind

testing result.pptx

Regards,

Danilo

Hi Danilo,

Thank you for sending over the PowerPoint. Can the customer please check the software to see if the duty cycle is being set accordingly or not? The control loop may be saturating because of the high current.

Best Regards,

Marlyn

Hi Marlyn,

Here is the response of our customer,

Hi, the code is operating open loop. There is a Timer0 clock to set PWM generation time 200us. After I turn on the external switch, the PWM will be triggered on through Trip Zone register. And after 200us, PWMs are turned off also through Trip Zone register. So it's totally open loop one.

Regards,

Danilo

Hi Danilo,

Thank you for relaying the response. Another critical thing we should check is the status of the MCU at the time the PWM generation stops. Can VDD/VDDIO or XRSn be checked to make sure the MCU is fully operational during this time? What if that power brick is susceptible to HV noise and is briefly turning off..

Best Regards,

Marlyn