• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 98 subscribers
  • Views 741 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

The problem of using LM10010+TPS56121 as CVDD of C6678s

brian feng
Prodigy 180 points
Other Parts Discussed in Thread: TPS56121, UCD9222, LM10010, UCD7242, LM10011

Hi，

We have a design using LM10010+TPS56121 as the CVDD power supply of C6678s as the reference design of  PMP7256.But the ouput voltage of TPS56121 is 1.1v,much higher than design using UCD9222+UCD7242.and we have check the four VCNTL pins of c6678,It is exactly different from one C6678 device to the next,and the most important thing is that there are two DSPs  which sent out 111111 through the VCNTL pins,and the output voltage of TPS56121 is 1.125V(not 1.103V as described in page 77 of sprabi2b.pdf ).I am confused  what  can determine the output of the VCNTL pins.Is there something wrong with the DSPs or the cricuit external cricuit of the LM10010 and TPS56121?

Best regards!

Brian.

  • 0
    TI__Mastermind 42545 points

    Hi Brian,

    I updated your post on the following thread. (http://e2e.ti.com/support/dsp/c6000_multi-core_dsps/f/639/p/237138/906888.aspx#906888) I will repeat that post here. 

    Hi Brain,

    The C6678 will command the voltage needed using the VCNTL lines. The four VCNTL pins should be connected to your LM10010and should send a sequence once after the C6678 is removed from reset to set the CVDD voltage level. The value sent could be different for every C6678 but many of the parts do request a voltage of 1.1V. Let me ask some specific questions.

    1) Do all of your boards with the LM10010 and the TPS56121 have a CVDD voltage of 1.1V or just some of the boards?

    2) Do you observe an initial voltage of 1.0V which then changes to 1.1V? The initial voltage setting for the LM10010 is 1.0V.

    3) Do you observe the command on the VCNTL pins to set the voltage level to 1.1V? You would see the VCNTL0-2 pins remain high and the VCNTL3 pin toggle once.

    The second issue you mentioned was that a command of 111111 results in a voltage of 1.125V instead of 1.103V. The LM10010requires very precise resistors to achieve the 5% voltage requirements for the C6678. This is discussed in detail in chapter 3 of the test report (http://www.ti.com/lit/ug/snvu148/snvu148.pdf).

    4) Are you using the 0.1% or 0.5% precision resistors to minimize the voltage feedback error?

    Regards, Bill

  • 0 in reply to Bill Taboada
    Prodigy 180 points

    Hi Bill,

    Thanks for your replay very much!

    Fristly,Let me answer you questions,(Let`s affirm that  the boards with the LM10010 and the TPS56121 are A boards and the boards with the UCD9222 and the UCD7242 are B boards!)

    1)Yes,almost all A boards have a CVDD voltage of 1.1V.But the B boards  have a CVDD voltage of about 0.9V.

    2)Yes,I checked A boards just now！I can exactly observe the CVDDs' change from 1.0V to 1.1V when powered on!

    3)I observed the VCNT pins through oscillorscape. A boards output  the following codes:111111,111110,111101(not the correct range mentioned in page 77 of  Hardware Design Guide for KeyStone I Devices——http://www.ti.com/lit/an/sprabi2b/sprabi2b.pdf).But B boards have the code 100001. And the VCNTL3 pin does toggle once.

    4)I didn`t try to use higher percision resistors .But I can do make external adjustments of TPS56121 to get a lower voltage!

    So,I have some questions:

    1)As i have  questioned before,why the DSPs have the abnomal VCNT codes?Is there something wrong with the LM10010 and the TPS56121?

    2)RandyP recommended me to use LM10011.I had seen the datasheets of the LM10011 and the LM10010,I thought  there is nothing different to my design(6 bit code settings with the MODE pin and 1.0v initial voltage settings with the SET pin ).

    Best regards!

    Brian.

  • 0 in reply to brian feng
    TI__Mastermind 42545 points

    Hi Brian,

    1) You refer to the VCNTL codes of 111111, 111110 and 111101 as abnormal. These are valid codes which will command a voltage within the 0.85V to 1.1V range of CVDD. The code presented on the VCNTL lines has nothing to do with which type of power supply is installed on the board. It is simply requesting the best operational voltage for that particular part. The next set of parts you order may request a different voltage. 

    2) Recently we announced that the initial voltage for parts in the KeyStone I family should be set to 1.1V. The LM10011 adds the ability to set an initial voltage using an Rset resistor. It is pin compatible with the LM10010 and the maximum resistance on the Rset line will command 1.1V.  This allows the LM10011 to replace the LM10010 on your board without requiring any changes. 

    Regards, Bill

  • 0 in reply to Bill Taboada
    TI__Mastermind 42545 points

    Hi Brian,

    Let me reemphasize that the LM10010 is no longer compatible with the C6678. The initial voltage for the C6678 must be 1.1V for proper operation. To support this you must replace the LM10010 with the LM10011 in your design. This should not require any changes to your PCB. 

    Regards, Bill

  • 0 in reply to Bill Taboada
    Prodigy 180 points

    Hi Bill

    Thank you very much!

    I'll take your advice!

    Regards,

    Brian

  • 0 in reply to Bill Taboada
    Prodigy 180 points

    Hi Bill,

    I replace the LM10010 with the LM10011 and set initial voltage at 1.1 V,But the CVDD remain 1.125V.Is there somthing wrong?I wonder why it is different between UCD9222+UCD7242 and LM10011+TPS56121!

    Regard,

    Brian