This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AWR1843: Feedback on the measured RF supply rail ripple

Part Number: AWR1843
Other Parts Discussed in Thread: TPS65400,

Hello,

My customer have designed their system around AWR1843 and the TPS65400 PMIC. They have measured the ripple on the 1.0V RF rail (LDO bypass) while running TI's MRR demo and got the below:

The orange area corresponds to the maximum amplitude measured across the whole measurement period (10's of thousands of measurements). The dotted line is the average over time of this amplitude. The continuous almost horizontal line is the limit set in the device datasheet (Table—Ripple specifications).

Except a significant excursion at around 508 kHz which corresponds to the PMIC switching frequency, the average measurement shows that the ripple is within our specification but for some minor excursions. However the maximum amplitude is above our specification in many points, but only for a small amount of time in average.

My customer ask for our feedback, except for the 508 kHz spike which they will manage in some way. Is it acceptable to consider that the temporary spikes will just translate to a higher noise floor that their signal processing chain has to deal with?

Expert advice will be gratefully accepted!


Best regards,
François.

  • Hey François,

    The spur and ripple levels have a dB-to-dB relationship, for example, a 1-dB increase in supply ripple leads to a ~1 dB increase in spur level. The ripples seem to be quite high in value. Can the customer refer to reference solutions mentioned in the Power Management Optimizations application note here:  https://www.ti.com/lit/an/swra577a/swra577a.pdf to see if they can reduce these ripples?

    Regards,

    Swati 

  • Hey Swati,

    Thank you. My customer's point is, since the the ripple levels are not permanently above the specified the limit (as on average they are below), only some chirps will be affected by them. So, maybe they could process them in some way so they do appear as spurs in the range FFT. Would that work? This would avoid the higher bill of materials with the extra LDO or passives.


    Best regards,
    François.

  • Hey François,

    These ripples would be detected as ghost objects in the FFT. It would be extremely difficult to filter them out in the processing and the customer might miss the real data because of such filtering. The customer is also using 500kHz switching frequency which makes the ripple values worse. 

    Hence, LDOs/passives should be added in the design. To lower BOM cost, the customer can use LC filters in their design instead of LDOs.

    Regards,

    Swati