S12 / MagniV MicrocontrollersのトピックRe: S12ZVML chip motor current acquisition amplification output waveform problem

S12ZVML chip motor current acquisition amplification output waveform problem

yq1 — Wed, 28 Jun 2023 01:01:45 GMT

The motor current is converted into voltage by the sampling resistor, and then amplified by the MCU internal op-amp.The amplified voltage waveform determines the ADC sampling time base through the PTU trigger module, and burrs are generated near the sampling time base.The ADC conversion time is 2us, which cannot be adjusted due to design reasons，Is there a good solution to optimize glitches？

The related code is shown in the attachment.

Re: S12ZVML chip motor current acquisition amplification output waveform problem

lama — Wed, 28 Jun 2023 12:37:32 GMT

Hi,

This is response of application team.

The implementation of external HW filter is up to you considering your measurement of glitches. However, usually the moment of the sampling is selected to the point where all oscillations are dumped.

I suggest to read....
https://www.nxp.com/docs/en/application-note/AN4704.pdf see chapter 3.3
and
https://www.nxp.com/webapp/Download?colCode=AN5207

Finally, a lot of things and final approach to design depends on work in lab and adjustment of either SW or HW to the real situation.

Best regards,

Ladislav

Re: S12ZVML chip motor current acquisition amplification output waveform problem

yq1 — Wed, 28 Jun 2023 14:15:29 GMT

The ADC acquisition point triggered by the PTU is the middle of the waveform, when the ADC is not acquired in the middle of the waveform basically no burr, once the PTU triggers the ADC acquisition, the burr is generated near the acquisition moment, this is very much like the burr generated by the ADC acquisition.

I don't know if this is caused by the ADC conversion time of 2us that I set.But this conversion time cannot be increased for design reasons.

The upper side is the voltage waveform when the ADC is not acquired, and the lower side is the voltage waveform when the ADC is acquired.

Does this burr affect the acquisition accuracy and is there a good way to optimize it?

yq1_1-1687961530102.jpeg

yq1_2-1687961545708.jpeg