Homework #6


  1. Crusty the engineer is a well intentioned employee in your organization who, unfortunately, is caught in some kind of 70's time warp. Crusty is convinced that a FSM built using discrete logic gates will outperform your shiny new 100Mhz FPGA design. Arguing with Crusty has come to nothing and its time to put an end to this once and for all. You are going to do the calculations to see how fast a discrete logic gate realization can run.

    You are to build the FSM using the the architecture given in Lecture 9. Use the positive edge triggered flip flops inside the SN74LS273N chip, and the AND, OR and NOT gates inside the SN74LS08N, SN74LS32N, and SN74LS04N chips respectively. You are to assume that the logic inside the MIE and OE boxes are 2-level combinational logic (solved using a kmap). You will need to pull the technical documents for the chips by searching Digi-Key and clicking on the Datasheet in the part view. When given a choice, always use the worst-case time estimates from the data sheets.

    To make your point you will need to show Crusty the details of your calculations, making sure to document where the numbers came from.
    1. Calculate the minimum clock period and the maximum clocking frequency.
    2. How many times faster is the discrete logic realization compared to a 100Mhz FPGA realization?
    3. Check out the Logic Families entry on wikipedia. Excluding the G series, what TTL family type should have the highest clocking rate?
    4. Provide a rough estimate of the clocking frequency of a FSM built from this type of logic. Making sure to show your calculation.
    5. List at least two other performance related dimension could you argue that an FPGA solution would be superior. Briefly describe how you would make your case.

  2. Complete the timing diagram below assuming that the register is positive edge trigger and has an asynchronous active low reset.