URL
https://opencores.org/ocsvn/astron_r2sdf_fft/astron_r2sdf_fft/trunk
Subversion Repositories astron_r2sdf_fft
[/] [astron_r2sdf_fft/] [trunk/] [rTwoBF.vhd] - Rev 3
Compare with Previous | Blame | View Log
-------------------------------------------------------------------------------- -- -- Copyright 2020 -- ASTRON (Netherlands Institute for Radio Astronomy) <http://www.astron.nl/> -- P.O.Box 2, 7990 AA Dwingeloo, The Netherlands -- -- Licensed under the Apache License, Version 2.0 (the "License"); -- you may not use this file except in compliance with the License. -- You may obtain a copy of the License at -- -- http://www.apache.org/licenses/LICENSE-2.0 -- -- Unless required by applicable law or agreed to in writing, software -- distributed under the License is distributed on an "AS IS" BASIS, -- WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -- See the License for the specific language governing permissions and -- limitations under the License. -- -------------------------------------------------------------------------------- -- Purpose : Butterfly -- Description : -- Default the rTwoBF is combinatorial and it can not be pipelined because -- of the feedback shift register. -- However for the FFT input stages with larger feedback shift registers it -- may be beneficial for achieving timing closure to move some of the z^(-1) -- shift delay out from the d to a feedback shift register into this rTwoBF. -- The shift must only occur for valid data, so therefor then the in_val -- input is also needed. -- The g_in_a_zdly allows getting a delay shift into this rTwoBF for input -- in_a. The g_out_d_zdly allows getting a delay shift into this rTwoBF for -- output out_d. Externally the feedback shift register depth must then be -- decreased by g_in_a_zdly+g_out_d_zdly. -- Remarks: -- . For the last FFT output stages the feedback shift register depth is ..., -- 4, 2, 1 so then there is less need to use g_in_a_zdly or g_in_a_zdly -- other than 0. -- . Default use g_in_a_zdly=0 and g_out_d_zdly=0, so then clk and in_val can -- be left not connected. -- . Alternatively one can use g_in_a_zdly=0 and g_out_d_zdly=1 for all -- stages. library ieee, common_pkg_lib, common_components_lib; use IEEE.std_logic_1164.all; use common_pkg_lib.common_pkg.all; entity rTwoBF is generic ( g_in_a_zdly : natural := 0; -- default 0, 1 g_out_d_zdly : natural := 0 -- default 0, optionally use 1 ); port ( clk : in std_logic := '0'; in_a : in std_logic_vector; in_b : in std_logic_vector; in_sel : in std_logic; in_val : in std_logic := '0'; out_c : out std_logic_vector; out_d : out std_logic_vector ); end; architecture rtl of rTwoBF is signal in_a_dly : std_logic_vector(in_a'range); signal out_d_ely : std_logic_vector(out_d'range); begin -- Optionally some z-1 delay gets move here into this BF stage, default 0 u_in_dly : entity common_components_lib.common_delay generic map ( g_dat_w => in_a'length, g_depth => g_in_a_zdly ) port map ( clk => clk, in_val => in_val, in_dat => in_a, out_dat => in_a_dly ); u_out_dly : entity common_components_lib.common_delay generic map ( g_dat_w => out_d'length, g_depth => g_out_d_zdly ) port map ( clk => clk, in_val => in_val, in_dat => out_d_ely, out_dat => out_d ); -- BF function: add, subtract or pass the data on dependent on in_sel out_c <= ADD_SVEC(in_a_dly, in_b, out_c'length) when in_sel='1' else in_a_dly; out_d_ely <= SUB_SVEC(in_a_dly, in_b, out_d'length) when in_sel='1' else in_b; end rtl;