TMS320F28377D: eCAP rising edge detection

Hi Luke,

Yes, you need to follow the datasheet for GPIO24 input voltage requirements i.e. as you stated minimum VIH=2V & max VIL=0.8V.

Here are the answers to your further questions:

Does it mean when the voltage input on GPIO24 change from 0V to 2.0V, it will have an eCAP event?
- Yes it should, but since 2V is the minimum VIH, I would go beyond 2V, at spec voltage i.e. VDDIO to guarantee the event.

Is there any voltage rising time spec?
- We have output timing spec for GPIO (section GPIO-Output Timing), but in your case you are using GPIO24 as the input to eCAP, we do not have spec for internal switching characteristics, but it will definitely be faster than output switching times.

No eCAP hysteresis, is this correct?
- Yes, only hysteresis here will be on GPIO.

Thanks & Regards,
Nirav

Hi Nirav,

Thanks for your response.

According to datasheet, the maximum GPIO output rising time is 8ns. So, we should be able to take this as the maximum eCAP input specification, it means that voltage input on GPIO24 should go VDDIO within 8ns.

Does it make sense?

Best regards,

Luke

Hi Luke,

Rising time of 8ns on GPIO is the output switching characteristic, when signal goes from low to high. As far as input switching time, it should be much faster. As long as the input signal switches from 0V to VDDIO (regardless of time) internally signal should be latched fairly quickly. To answer your question, there is not requirement on rising time for the input on GPIO24, for ECAP to capture a rising edge it needs voltage on GPIO24 to rise to 2V or above (as per the spec to VDDIO voltage). Internally it triggers a rising edge as soon as GPIO24 crosses minimum VIH spec of 2V.

Regards,
Nirav

Hi Nirav,

My customer is facing a system noise problem and eCAP doesn't work well. To have noise immunity, we add a RC low pass filter at eCAP input and use GPIO input qualification function, seems that eCAP works well now.

As you know, adding a RC low pass filter will change input voltage rising time/slope, we want to make sure we are adding appropriate LPF and having correct GPIO input qualification setting, this is the reason why we are asking the GPIO/eCAP input rising time spec.

The GPIO input switching time should be much faster than 8ns, if we take 8ns as the maximum eCAP input rising time spec and then determine R, C for LPF and parameters for GPIO qualification, is this reasonable?

Regards,
Luke