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.

AWR1642: Power supplies for TIDA-01570 automotive reference design

Part Number: AWR1642
Other Parts Discussed in Thread: , TIDA-01570

I'm creating my own board around the AWR1642, and basing it on the TIDA-01570 design. I noticed some differences in the power supplies between TIDA-01570 and AWR1642Boost. I would appreciate some guidance in selecting my power supply components.

1. The automotive board uses only one LDO, which is for the 1.0 V RF power supply. The AWR1642Boost uses a 1.3 V supply and runs the internal LDO on the AWR chip. I understand that bypassing the internal LDO reduces heat dissipation by the AWR chip, but is there a downside to this approach? Is noise performance better if the internal LDO is used?

2. Another difference I see is that the automotive board has ferrite beads between every power supply and the AWR chip, while the AWR1642Boost has none. According to this document, LC filters like the automotive board uses are more power efficient than LDOs, but noisier, and that noise can get into the RF signal. However, it seems that some of these ferrite beads are unnecessary. In particular, the FB after the 1.0V LDO and the FBs for 3.3V and 1.24V (digital) supplies. Can I remove those without hurting the noise performance?

3. It also seems that adding an LDO for the 1.8V supply on the automotive board would benefit the noise performance, since swra557 indicates that the 1.8V and 1.0V power rails are both sensitive to input supply ripple. Is this correct?

  • Hi,

    We have asked the designer of the board for support.

    We will get back to you as soon as possible

    thank you
    Cesar
  • Hello ,
    Please find my reply below:

    >>1. The automotive board uses only one LDO, which is for the 1.0 V RF power supply. The AWR1642Boost uses a 1.3 V supply and runs the internal LDO >>on the AWR chip. I understand that bypassing the internal LDO reduces heat dissipation by the AWR chip, but is there a downside to this approach? >>Is noise performance better if the internal LDO is used?

    You are right that using 1V mode helps in heat dissipation since the internal LDO power loss is eliminated. But if the internal LDO is bypassed then the quality of external required , interms of the ripple requirements , become tighter. You can find the supply requirements in the AWR1642 datasheet (www.ti.com/.../awr1642.pdf) , table 5-2. You can see that if you use the internal LDO you get a benefit of that LDO. But if you can provide a cleaner 1V supply then that can be also used.

    >>2. Another difference I see is that the automotive board has ferrite beads between every power supply and the AWR chip, while the AWR1642Boost >>has none. According to this document, LC filters like the automotive board uses are more power efficient than LDOs, but noisier, and that noise can >>get into the RF signal. However, it seems that some of these ferrite beads are unnecessary. In particular, the FB after the 1.0V LDO and the FBs for >>3.3V and 1.24V (digital) supplies. Can I remove those without hurting the noise performance?

    The FB bead o n the 1V is required since the LDO alone cannot provide the required PSRR with the low dropout, hence additional filtering using FB is needed. Some level of filtering is needed on 1.2V as well due to internal coupling. On the 3.3V, the FB may be optional.

    >>3. It also seems that adding an LDO for the 1.8V supply on the automotive board would benefit the noise performance, since swra557 indicates that >>the 1.8V and 1.0V power rails are both sensitive to input supply ripple. Is this correct?

    Both supplies are critical for the analog/RF performance. You could meet the ripple requirements either using LDOs or using other filters like LC filters. It depends on the noise frequencies present on your supplies. In the TIDA, the supply noise is dominated by the DCDC switching frequency, which is at fixed at 4Mhz. Hence it can be filtered easily with LC filters as well.

    Regards,
    Vivek
  • Thanks, this is helpful information!