Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
18 years ago

SDRAM Controller

Now, i am researching about SDRAM Controller Core.So o need some document its.Can you help me?

Many thanks

43 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    Please post me your files so that I can verify your timing pameters and code/testbench. Also, Which version tools are you using? Is this for the Altera Cyclone II Devkit, or the DE1/2 etc...? Is this a post synthesis timing waveform that you posted? The answers will help me to further evaluate and offer advice on you design.

    Daniel L Short,

    Senior Design Engineer

    Specialized Communications Corp.

    Noventri Corp.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    Nevermind...

    I made this post before looking onto the next page...Sorry. I will also look it over(along with FVM) and see if I can offer anything more.

    Daniel L Short
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor

    Ok. I have loaded the design into Quartus II Version 8 SE, and updated the database to work in my environment. I recompiled and verified the simulation in the Quartus simulator. I have confirmed that the simlation is a timing simulation, using an unconstrained project. The target shows as a Cyclone III, Is this correct? And the timing analyzer messages indicate a critical warning on an unconstrained file for optimization.

    I would just like to take a moment as a side note to say that I totally agree with FVM in that the sense that I get is this method of simulation using a stimulus file instead of a written testbench does strike me as a potential problem spot. Did you make it or did you get it from some Altera source? I am just wondering because I generally have not had very much success with the Quartus simulator (just too limited using vectors). And I really am not certain that the back annotated delays are going to allow the operation of the RTL post fit. What I mean is...I have done simulations with the stimulus editor where I have improperly triggered logic at times when state-machines may not even care about things...For example trying to write lines during a refresh cycle or some other parrallel RTL event.

    Long story Short.... Have you seen the Functional model simulate properly yet? Start there with a known good testbench, and a tool Like the Modelsim Simulator and then do the gate-timing.

    Hope this helps, and as more information is available throughout the day(as I have time to work on it) I will post my findings.

    Daniel Short