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.

TPS51206: LPDDR4 Multiple power line solution

Part Number: TPS51206
Other Parts Discussed in Thread: TPS568215, TPS65296, TPS51200

Hello team,

We have received inquiries from customers regarding the power supply of LPDDR4. It seems that they are in a hurry, so I would appreciate your feedback.

They are using LPDDR4 in the device under development, but they are making these 3 voltage lines as a power supply.

  • 1.8V VDD with TPS563219,
  • 1.1V VDD2 with another TPS563219
  • And 0.6V VDDQ by dividing it with another 1.2V TPS568215 to TPS51206

Since they use 4 LPDDR4, VDDQ consumes 0.74A (184mA x 4), and they are worried whether TPS51206 can withstand sudden changes.

Should they use a different DCDC (such as TPS568215) for 0.6V VDDQ?

Looking at the catalog, it seems that there is a dedicated power supply for LPDDR4 (TPS65296), but they do not know whether LPDDR4 should be one power supply like this, so I would appreciate it if you could give they some advice about it.

By the way, the input voltage is 12V.

Best Regards,

Ryotaro Fukui

  • Hi Fukui-san,

    I don't think you need TPS568215+ TPS51206 to generate 0.6V VDDQ. I think you can use TPS51200 to generate 0.6V VDDQ with 1.1V VDD2 as input. Since TPS51200 has external REFIN divider, you can use 1.1V in the LDO input.

    With 0.74A input, even single TPS568215 seems to be an overkill.

    Regards,

    Weidong

  • Hi Weidong-san,

    Thank you for your suggestion of the power line.

    I introduced the proposal to the customer, but he wants to know if there is any problem with the TPS51206 even if the IC is not changed, since they are just before mass production and don't want to make change as possible.

    Also, TPS568215 seems to be used for other purposes, and already decided in the circuit. I would like you to answer additional questions based on that too.
    I have attached a brief diagram in the PPT file in the hope that it will be easier to understand.

    Regards,

    Ryotaro

  • Hi Fukui-san,

    Thanks for the block diagram. In this case, I think TPS51206 can be used since customer already ahs 1.2V output.

    Please note TPS51206 will follow the 1.2V output of TPS568215, since this 1.2V is fed into TPS51206 as reference input. To reduce the voltage fluctuation, enough capacitance on the 1.2V output needs to be design to reduce the change. 

    TPS51206 should be able to withstand sudden change of VDDQ, with a minimum of 10uF output capacitance.

    Regards,

    Weidong

  • Hi Weidong-san,

    We have received two additional questions from our customers.

    Q1.

    They think that TPS51206 makes and supplies VTT: 0.6V for DDR4 IO drive VDDQ: 1.2V.
    When using for LPDDR4 VDDQ, the IO drive VDDQ is 0.6V, so the usage is different, but is there any problem using 51206?

    Q2.

    Regarding TPS52100, which was introduced as TPS65296 dedicated to LPDDR4, the circuit construction of VDDQ seems to be the same regulator control circuit as TPS51206. Can LPDDR4 also support VDDQ with this?
    (Since it is the IO drive voltage, I am worried whether this circuit configuration can handle voltage fluctuations due to data fluttering, etc.)

    Although there are a lot of questions, but please also check "if the TPS51206 is affected by the voltage fluctuation of the FPGA" that I sent you earlier.

    If you have any necessary materials, please contact me again.

    Regards,

    Ryotaro

  • Hi Fukui-san,

    For Q1, I don't see any problem using TPS51206 to supply 0.6V VDDQ. TPS51206 can both sink/source. If LPDDR4 needs source current only, it's also ok to use TPS51206.

    For Q2, I'm not clear about the voltage fluctuations you are mentions. What is the range of voltage fluctuation you are referring to? I think this fluctuation has to be within certain range, is it correct? TPS51206 will follow the reference input. To make sure the reference is clean, I think you can filter the 1.2V before sending to TPS51206, and then it it should be ok.

    Regards,

    Weidong