Forum Discussion

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

About My VHDL Task

Hi, as a beginner in VHDL, I like to hear comments and ideas from you.

I have a task to do, my lecturer sent me a code but I couldn't find out.

This's the code.

--

library ieee;

use ieee.std_logic_1164.all;

entity singen is

port

(

clk : in std_logic;

q : out std_logic_vector(7 downto 0)

);

end entity;

architecture rtl of singen is

type regtype IS array (0 to 15) of std_logic_vector(7 downto 0);

signal reg : regtype :=

(X"00", X"18", X"2d", X"3b", X"40", X"3b", X"2d", X"18",

X"00", X"e8", X"d3", X"c5", X"c0", X"c5", X"d3", X"e8");

signal c : integer range 0 to 15 := 0;

begin

process

begin

wait until rising_edge(clk);

if (c < 15) then

c <= c + 1;

else

c <= 0;

end if;

end process;

q <= reg(c);

end rtl;

--

He told me that "Use this file as component in VHDL and create design, where two sinusoidal signals with different frequencies (31250 Hz and 78125 Hz) are summed or subtracted (depending on outside signal)."

With this above code how to generate two signals with different frequencies. It's generating only one signal and it's being q. But I don't know how he adjusted its frequency like he desired. Do you guys mind if explaining me? Thanks in advance.

9 Replies

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    7 years ago

    Do you remember your ;ecturer telling you about adders? hlw about prescalers? Clock frequencies?

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    7 years ago

    --- Quote Start ---

    Do you remember your ;ecturer telling you about adders? hlw about prescalers? Clock frequencies?

    --- Quote End ---

    I know how to make prescalers and to define frequency of clock signal. But, something that I don't know and he didn't bring up, why 'c' is assumed till 15, why not 20. I again know there's frequency calculation with this whole number, but how to calculate it we haven't learnt. That's what I posted.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    7 years ago

    The reg table only has 16 elements. It does seem a design flaw to count only from 0 to 14, rather than 0 to 15. It cannot run until 20 because it would overflow the reg array.

  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    7 years ago

    --- Quote Start ---

    The reg table only has 16 elements. It does seem a design flaw to count only from 0 to 14, rather than 0 to 15. It cannot run until 20 because it would overflow the reg array.

    --- Quote End ---

    I didn't mean that. The prescaler should be meaning and adjusting the frequency. If it has ability to count to 20(such as), This situation will be changing the frequency. There should be a calculation why he has 16 elements.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    7 years ago

    --- Quote Start ---

    There should be a calculation why he has 16 elements.

    --- Quote End ---

    because that resolution of the sine is good enough for the purpose of the assignment I think.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    7 years ago

    huytergen,

    You can't do what he asks using the code he sent you. It looks like the reg table is a simple 16 element full sine wave in twos-compliment with an amplitude of 64. You should research DDS (Direct Digital Synthesis) and will find alot of info on the web. The output frequency of a DDS is set by setting a "tuning word" to an appropriate value:

    F = M * Fc / 2^n

    where F is the sine wave frequency, Fc is the clock frequency, M is the tuning word, and n is the width of the phase accumulator in bits. Normally, the phase accumulator is much wider than the lookup table in order to get the desired frequency resolution. For example, if your clock is 100 Mhz and you want a resolution of 1 Hz, set M to 1 to find 2^n of 10.0e8 so you need n = 27. Most people would just use 32 bit for the phase accumulator and M. Then use the upper 4-bits as an index into your lookup table.

    So you will need to change his code to have a phase increment M input into the entity. The way he is incrementing c is very clumsy. It is going from 0 to 15 so it makes much more sense to make it an unsigned 4-bit number that you increment by M in one-line of code. M will also be an unsigned 4-bit number. He has M hardwired to 1, so the output frequency is fixed. In practice, you probably cannot achieve the desires frequencies with a 4-bit phase accumulator and will have to make it much wider.

    You can figure out the rest.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    7 years ago

    --- Quote Start ---

    He has M hardwired to 1, so the output frequency is fixed.

    --- Quote End ---

    That is assuming that the clock frequency is fixed. I don't know the level of this course but for beginners changing the clock speed, and given the numbers devide too beautifully to be a coincidence I think just a prescaler will fulfill the requirement, with a verry crude sine, but good enough I think.
  • Altera_Forum's avatar
    Altera_Forum
    Icon for Honored Contributor rankHonored Contributor
    7 years ago

    --- Quote Start ---

    That is assuming that the clock frequency is fixed. I don't know the level of this course but for beginners changing the clock speed, and given the numbers devide too beautifully to be a coincidence I think just a prescaler will fulfill the requirement, with a verry crude sine, but good enough I think.

    --- Quote End ---

    You're probably right since one freq is 2.5 times the other. If he's lucky, the instructor was showing a clumsy way to do it so they would appreciate the next lesson on how to do it more elegantly .:)