• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 30 subscribers
  • Views 472 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.

Original question:

TMS570LC4357: EMIF as parallel bus

TMS570LC4357: TMS570LC4357: design advices using EMIF

Hector Gastaminza20
Prodigy 20 points
Part Number: TMS570LC4357

Hello.

I'm deciding the design of the memory map of my system, and I would need some information of what approach is better and why.

A. Having one big memory using one chip select

or

B. Having two or three small memories using two or three CS.

Thanks

  • 0
    TI__Guru**** 196416 points
    Hello,

    The maximum memory size for each chip select (CS2, CS3, CS4) is 16MB. If you need >32MB memory, you have to use 3 separate memories attached to CS2/CS3/CS4.

    If you only need 16MB memory, you can use either 1 memory attached to 1 CS or 2 memories attached to 2 chip selects. But I prefer to use 1 memory instead of 2. Using 1 memory can make the PCB layout easier, and save CB space.
  • 0 in reply to QJ Wang
    Prodigy 20 points

    Hello QJ Wang,

    The maximum memory size for each chip select (CS2, CS3, CS4) is 16MB. If you need >32MB memory, you have to use 3 separate memories attached to CS2/CS3/CS4.

    If you only need 16MB memory, you can use either 1 memory attached to 1 CS or 2 memories attached to 2 chip selects. But I prefer to use 1 memory instead of 2. Using 1 memory can make the PCB layout easier, and save CB space.

    Thank you for share your approach. I agree; besides using one memory the system may be more reliable, every chip add more fail probabilities I guess.

    Our system it is going to be designed for future updates, so all the CS are going to be routed in the board, so from the hardware side, have one or more chips should not be a problem.

    I was thinking on the SW side, which approach might be the best. Using just one maybe could be better if we priorize reliability (are you agree?).
    Or maybe we could have a better performance spliting the access to memory in different chips, or maybe splitting could worst.
    Do you know where I could find information to support this kind of decision?
    Hector
  • 0 in reply to Hector Gastaminza20
    TI__Guru**** 196416 points
    Hello Hector,

    We are not able to access two memories with different chip selects simultaneously. From performance point of view, I prefer to use one 16MB memory rather than two 8MB memories.