Error 10327: Can't determine definition of operator ""+""

I have the same problem.. Error (10327): can't determine definition of operator ""+"" -- found 0 possible definitions

I write this program:

LIBRARY ieee;

USE ieee.std_logic_1164.ALL;

use ieee.numeric_std.all;

use ieee.std_logic_signed.all;

ENTITY filter_iir IS

GENERIC ( L: INTEGER :=3;

W1: INTEGER :=16;

W2: INTEGER :=32;

W3: INTEGER :=33;

W4: INTEGER :=14

);

PORT ( --x_in: IN signed (15 downto 0);; --INPUT

--y_out: OUT BITS14; --RESULT

x_in : in BITS16;

y_out : out signed (31 downto 0);

clk : IN STD_LOGIC);

END filter_iir;

ARCHITECTURE myiir of FILTER_IIR IS

SIGNAL x : ARRAY16TO3;

signal x_num : ARRAY32TO3;

constant b0 : signed(W1-1 downto 0) :=x"0119";

constant b1 : signed(W1-1 downto 0) :=x"0232";

constant b2 : signed(W1-1 downto 0) :=x"0119";

constant a1 : signed(W1-1 downto 0) :=x"FFB5";

constant a2 : signed(W1-1 downto 0) :=x"00B0";

shared variable acc : signed (W2-1 downto 0):= (others => '0');

shared variable y : signed (W2-1 downto 0):= (others => '0'); -- 28 bit

shared variable y_den : signed (W2-1 downto 0):= (others => '0'); -- 27 bit (11+16)

BEGIN

sop: PROCESS

BEGIN

wait until (clk='1') ;

x(0)<=x_in;

x_num(0)<=b0*x(0); -- b0x(k)

debugx<=x_num(0);

x_num(1)<=b1*x(1); -- b1x(k-1)

debugx1<=x_num(1);

x_num(2)<=b2*x(2); -- b2x(k-2)

debugx2<=x_num(2);

for i in 0 to L-2 loop

x(i+1)<=x(i);

end loop;

for j in 0 to L-3 loop

y(j+1):=y(j);

end loop;

y(0):=(x_num(0)+x_num(1)+x_num(2)+y_den(0)+y_den(1));

acc:=y(0);

y_den(0):=a1*y(0)(15 downto 0); -- a1y(k-1)

y_den(1):=a2*y(1)(15 downto 0); -- a2y(k-2)

end process;

y_out <= y(0)(31 downto 0);

end myiir;

when it sum these parts Y(0):=(x_num(0)+x_num(1)+x_num(2)+y_den(0)+y_den(1)) the compiler give me an error.. why?

i use ieee.std_numeric.all library and in this library there is the definition of "+"....?!?!?:confused::confused::confused:

Boh.. i search the error but i don't understad..

so initially i think that this function want the adder at the same lengths.. but i read the source code.. and seem that it isn't a problem...

x_num is array of signed with length 16 and y_den is array of signed with length 32.. and y(0) is 32bit signed register..

so i don't know where is the problem?

can somebody help me..?