help need in conversion problem

[senthil]

hello friends,

i do my ifft implementation with respect to complex number input and
got my output as complex number in simulation side.

Modelsim S.E 5.5 -simulator.

With respect to real type in vhld coding, that one is not synthesised.
so, i'm going to convert the real to std_logic_vector. for that i
fixed by quantization of corresponding value of real type to 8 bit.
ok.

for eg. if i have real value = 0.316 i can represent in bitvector as
"00101000"
and 0 .949= "01111001". if i do some calculations like
a <= 0.316+0.949 <= "00" & "00101000" + "00" & "01111001" =
"0010100001"

0.316+0.949 = 1.265 => "10100001"

both come correct ans.. if suppose some of the carry generate means
what can i do. and if i multiply the sum of number with the some
twiddle factor(that i put in lookup table as std_logic_vector of 8
bit), the total bit will increase at that time manuall calculation of
real converted to bit values and computed result are vary..
for that what i can do .. tell me and give some suggestions...

expecting ur reply
cheers
Senthil.R
MTech vlsi design

 

[senthil]

hi ,,
i in need how to convert the real value to bit of 8 bit vector.. ie.,
normalization techniques..
pls give some guidance.
regrds

bhuvasen@hotmail.com (senthil) wrote in message news:<3045a319.0402202124.3162f7d9@posting.google.com>...

hello friends,

i do my ifft implementation with respect to complex number input and
got my output as complex number in simulation side.

Modelsim S.E 5.5 -simulator.

With respect to real type in vhld coding, that one is not synthesised.
so, i'm going to convert the real to std_logic_vector. for that i
fixed by quantization of corresponding value of real type to 8 bit.
ok.

for eg. if i have real value = 0.316 i can represent in bitvector as
"00101000"
and 0 .949= "01111001". if i do some calculations like
a <= 0.316+0.949 <= "00" & "00101000" + "00" & "01111001" =
"0010100001"

0.316+0.949 = 1.265 => "10100001"

both come correct ans.. if suppose some of the carry generate means
what can i do. and if i multiply the sum of number with the some
twiddle factor(that i put in lookup table as std_logic_vector of 8
bit), the total bit will increase at that time manuall calculation of
real converted to bit values and computed result are vary..
for that what i can do .. tell me and give some suggestions...

expecting ur reply
cheers
Senthil.R
MTech vlsi design