LAUNCHXL-F28379D: Simulation is not matching with model setting

ArTzy

Hi Experts,

Good day.

Client is testing DAC of TI launch Xl-F28379D with MATLAB embedded coder support package. Testing is done with sine wave and visualizing in DSO.

As per their observation, a DSO 125Hz is found for 1e-6 sampling time, and 1177 samples.

They're asking why real time is not matching with the model settings.

Please provide assistance in matching the real time with the model settings.

Thank you for your support.

Regards,
Archie A.

over 2 years ago

0 Santosh Jha over 2 years ago

TI__Guru 50430 points

Hi Archie,

We have reached to MathWorks team to take a look why it is not matching. We should expect response by tomorrow.

0 ArTzy over 2 years ago in reply to Santosh Jha

TI__Genius 13959 points

Hello Santosh,

Thanks for letting us know here.

Thanks for your support.

Regards,
Archie A.

0 Ram Swaroop Alla over 2 years ago

Intellectual 710 points

ArTzy said:
with

Hi Archie,

We need to look into the model to understand the issue better. Could you please share the model and the snapshots of the waveform in DSO?

Best Regards,
Ram Alla | MathWorks

0 ArTzy over 2 years ago in reply to Ram Swaroop Alla

TI__Genius 13959 points

Hello Ram,

Thanks for supporting.

The client just shared the snapshots attached below with the following notes:

"In MATLAB 1177 samples and sample time 1e-6 given i.e 1.177e-3 sec but getting 8ms in real time DSO. as shown. Please help any settings need to be changed in model configurations to equate real and simulation time."

For your information.

Thank you.

Regards,
Archie A.

Attached:

0 ArTzy over 2 years ago in reply to ArTzy

TI__Genius 13959 points

Hi Experts,

Good day.

Client is asking an update. Any further support on this?

Thank you.

Regards,
Archie A.

0 Santosh Jha over 2 years ago in reply to ArTzy

TI__Guru 50430 points

Archie,

We reached out to MathWorks. Ram is actively looking on it and he will provide the feedback as soon as he finds the issue. Hopefully we should get some update by Friday.

0 ArTzy over 2 years ago in reply to Santosh Jha

TI__Genius 13959 points

Hello Santosh,

Thanks for this update.

Regards,
Archie A.

0 ArTzy over 2 years ago in reply to Santosh Jha

TI__Genius 13959 points

Hi Team, any update on this?

Thank you.

Regards,
Archie A.

0 MatthewPate over 2 years ago in reply to ArTzy

TI__Guru 71785 points

ArTzy,

I'll check with the MW team to see if they have an update.

Best,
Matthew

+1 Ram Swaroop Alla over 2 years ago

Intellectual 710 points

Hi Archie,

When we tried to run the model with the sample time of 1e-6, an overrun is detected. If we reduce the sample time to 1e-3, it works fine. You may need to consider to reduce the sample time. Also it is not recommended to use sine block approach as it is not codegen friendly.

If you need to run at higher sample rates, you may need to pre compute the DAC values and use DMA instead of using CPU to compute it.
Refer to the below example to know how the sine wave samples are pre generated by the callback script and use DMA to copy to ePWM registers.

www.mathworks.com/.../modify-epwm-using-dma.html

If you have any further queries, please contact MathWorks TechSupport.

Best Regards,
Ram Alla.

C2000™︎ microcontrollers

C2000 microcontrollers forum

LAUNCHXL-F28379D: Simulation is not matching with model setting