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.

K2H bypass cap in the hardware design guide

Champs

I was going over the Hardware design guide for keystone 2 Table 3 and the number of bypass cap's are not relaistic. They can't fit that many under the device. So wondering why we have so many in the hardware design guide.

 

On the EVM it seems that they are very miniaml on the bypass cap's. So i would iamgine there got to be  a middle line. Or we shoudl tell teh custoemrs to their own simulation on the power plan and let them decide on the bypass cap.

 

Thanks

Regards

Mohsen

  • Hello Mohsen,

    The recommendations provided in table 3 are absolute minimum requirement. All capacitance identified assumes the use of the recommended power supplies (especially for the variable core). The recommended values are required for proper operation of the recommended power supplies.

    If the customer follows the same power architecture as given in Hardware design guide (Figure 1), they have to strictly adhere to the recommendations for proper operation. If not, i would suggest them to do their own power integrity simulation to find out the best suitable numbers.

    Regards,

    Senthil

  • Senthil
    thnaks for the reply, but it is not possible to place that many bypass cap on the back of the part. We need to really understand that what we recommend has to be relaistic and it has to fit the board. They can't but the bypass cap next to the device becasue it is useless becasue of the trace length.

    Have we tried to layout the part with the bypass that we recommend? if yes i liek to see the layout please.

    Thanks

    Regards
    Mohsen
  • Hello Mohsen,

    The necessary decoupling depends on a number of factors including the PCB stack-up, capacitor position, capacitor value and CVDD plane size and position. The table in the KeyStone II hardware design guide was based on the K2H characterization board.

    If the customer feels that they cannot accommodate the recommended capacitors on their board, they can reduce the number of capacitors but they need to do a full PDN simulation of their PCB to ensure that their network will provide the CVDD voltage to the limits defined in the specification. 

    Please refer the below thread where similar query has been addressed.

    https://e2e.ti.com/support/dsp/c6000_multi-core_dsps/int-c66x_multi-core_dsps/f/513/t/344606

    Regards,

    Senthil

  • Senthil
    I heard Bill Taboada has a spread sheet to calculate the bypass stuff. Can I get that and give it to the customer?

    Thanks

    Regards
    Mohsen
  • Hello Mohsen,

    I will check with Bill Taboada and try to get you the spreadsheet.

    Regards,

    Senthil

  • Hi Mohsen,

    We do not have a spreadsheet calculator specifically for the K2H. We do have a version of a PDN calculation spreadsheet that is available on the following wiki.

    The spreadsheet can be used as a starting point for generating a list of capacitors needed for a design but it is not a substitution for proper simulation of the PDN design for the system. Since the board layout, stack-up and materials as well as the specification for the bypass capacitors and their placement can all have an effect on the performance of the power distribution network, we can't publish a list of required capacitors for a customer design. The list in the Hardware Design Guide were the capacitors needed for the characterization board which was built for the K2H components. 

    Regards, Bill

  • Bill

    thanks for the answer, what are the chances to get the layout and the schematics for the characterization board i like to take  a look if at all possible. It will be for my eyes only and not to give to the customer.

    Also it is interesting that we have so many different cap's values for the bypass. I assume it has to be that way but some explanation would help to make the customer understand why we are doing this. Or maybe point out how we did the simulation and we came up with these values.

    Thanks

    Regards

    Mohsen

  • Mohsen,

    I'll see if I can get a copy and forward it to you in email.

    Bill