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.

TPS54318: Query Regarding Load-Induced Transient on 1.8V Rail with RF Module

Part Number: TPS54318
Other Parts Discussed in Thread: IWRL6432,

Hi TI Support Team,

We are using a TI DC-DC buck regulator (TPS54318RTER) to generate a 1.8 V rail (rated 3 A) for powering a radar module (which  uses IWRL6432 chipset from Texas Instruments).

Radar module spec for Min operating voltage and Max Current is belowimage.png

During system-level testing, when the radar module is actively transmitting, we observe periodic voltage dips and overshoot on the 1.8 V rail. These events are correlated with the radar TX activity and appear to be load-induced transients rather than steady-state switching ripple.

image.png

Measurement summary:

  • Nominal VOUT: 1.8 V

  • Observed transient deviation:

    • Undershoot: ~62 mV

    • Overshoot: ~32 mV

    • Total transient excursion: ~94 mV p-p

We have also validated the regulator independently using an electronic load with a 0.2 A to 1.3 A load step at a slew rate of 0.025 A/µs, where we observed approximately 38 mV(Current steps 1.3A to 0.2A) and -50mV(Current step 0.2A to 1.3A transient deviation, indicating stable loop behavior.

The radar module datasheet specifies an operating supply range of 1.71 V to 1.89 V, so how we can measure minimum voltage out of below two case.  since in case-2 we get minimum voltage is out of sepc from moduel minimum requirement.

Minimum Voltage =1.8V-Undershoot=1.8V-62mV=1.738V  Case -1

Minimum Voltage =1.8V- Ripple (Undershoot+Overshoot)=1.8V-(62+32)=1.8V-94mV=1.706V  Case -2

 

We would like your guidance on the following:

  1. How should such load-induced transient deviations be evaluated versus traditional ripple specifications?

  2.  Should We measure ripple Transient + to Tarnsient -, in our case can we call our ripple is 94mV ( Addition of Undershoot: ~62 mV and Overshoot: ~32 mV).

  3. Or Ripple should be seperatly measure when load is get stable, in that case ripple we get 10-15mV with radar module (ignore Transiet part). 
  4. From a regulator perspective, is this level of transient deviation considered acceptable for fast, pulsed loads?

  5. Are there any recommended PDN optimization or output capacitance guidelines (type, value, placement) to further reduce transient droop for high di/dt loads?

  6. When the module datasheet specifies a minimum operating voltage (for example, 1.71 V in this case), does this limit assume a steady-state DC voltage, or does it also account for short-duration transient voltage deviations?

Partial SCH belowimage.png

Your inputs will help us determine whether the observed behavior is acceptable as-is or if further optimization is recommended.

Thank you for your support.

Best regards,
Kunal Barot

  • Hello Kunal,

    thanks for reaching out in E2E.

    Here is my feedback to your questions:

    1. - 4. Load transient and VOUT ripple are usually evaluated separately under reproducible conditions for the datasheet charts and for device validation. However, in an application the overall voltage requirements need to get achieved under application load conditions. Therefore I would suggest for you to focus on the application conditions and the application requirements.

    5. In general it is important to avoid long traces for the output filter, which add parasitic inductance to the control loop. This includes also the GND return path to the TPS54318. Also the VSNS line is susceptible to noise. Further layout guidelines can be found in section 10 of the datasheet.

    6. I'm not an expert for the radar module requirements. I would suggest to submit a separate E2E post, referencing only the radar module. Then this request will be automatically assigned to the right expert team.

    Do you observe the transient spikes at VCC_1V8 or at RDR_VCC_1V8? Please let me know.

    Best regards,

    Andreas.