page 132,66,2,2 opt nomd,nomex,nocex,loc,mu ;******************************************* ;Motorola Austin DSP Operation June 30,1988 ;******************************************* ;DSP56000/1 ;Port to Memory FFT - 1024 point ;File name: F2-56.asm ;************************************************************************** ; Maximum sample rate: 3.332 ms at 20.5 MHZ/ 2.530 ms at 27.0 MHz ; Memory Size: Prog: 151 words ; Data: 8194 words ; Number of clock cycles: 68316 (34158 instruction cycles) ; where 66240 for FFT, 2076 for data input. 5/12/93 ; Clock Frequency: 20.5MHz/27.0MHz ; Instruction cycle time: 97.5ns / 74.1ns ;************************************************************************** fftr2fn macro idata,coef,odata fftr2fn ident 1,0 ; ; 1024 Point Complex Fast Fourier Transform Routine ; ; This routine performs a 1024 point complex FFT on external data ; using the Radix 2, Decimation in Time, Cooley-Tukey FFT algorithm. ; ; Real input and complex output data ; Real data in X memory ; Imaginary data in Y memory ; Normally ordered input data ; Normally ordered output data ; Coefficient lookup table ; -Cosine values in X memory ; -Sine values in Y memory ; ; Macro Call - fftr2fn data,coef,odata,ptr1,ptr2 ; ; data start of external data buffer ; odata start of external output data buffer ; coef start of sine/cosine table ; ptr1 location of pointer pointing to data acquisition buffer ; ptr2 location of pointer pointing to FFT buffer ; ; Radix 2, Decimation In Time Cooley-Tukey FFT algorithm ; ___________ ; | | ; ar,ai ---->| Radix 2 |----> ar',ai' ; br,bi ---->| Butterfly |----> br',bi' ; |___________| ; ^ ; | ; wr,wi ; ; ar' = ar + wr*br - wi*bi ; ai' = ai + wi*br + wr*bi ; br' = ar - wr*br + wi*bi = 2*ar - ar' ; bi' = ai - wi*br - wr*bi = 2*ai - ai' ; ; Address pointers are organized as follows: ; ; r0 = ar,ai input pointer n0 = group offset m0 = modulo (points) ; r1 = br,bi input pointer n1 = group offset m1 = modulo (points) ; r2 = ext. data base address n2 = groups per pass m2 = 256 point fft counter ; r3 = coef. offset each pass n3 = coefficient base address m3 = linear ; r4 = ar',ai' output pointer n4 = group offset m4 = modulo (points) ; r5 = br',bi' output pointer n5 = group offset m5 = modulo (points) ; r6 = wr,wi input pointer n6 = coef. offset m6 = bit reversed ; r7 = not used (*) n7 = output pointer storage m7 = not used (*) ; ; * - r7 and m7 are typically reserved for a user stack pointer. ; ; Alters Data ALU Registers ; x1 x0 y1 y0 ; a2 a1 a0 a ; b2 b1 b0 b ; ; Alters Address Registers ; r0 n0 m0 ; r1 n1 m1 ; r2 n2 m2 ; r3 n3 m3 ; r4 n4 m4 ; r5 n5 m5 ; r6 n6 m6 ; n7 ; Alters Program Control Registers ; pc sr ; ; Uses 8 locations on System Stack ; ; ; main fft routine ; _points equ 1024 ;number of FFT points _intdata equ 0 ;address of internal data workspace move #idata,r2 ;initialize input pointers move r2,r0 move #_points/4,n0 ;initialize butterflies per group move n0,n4 ;initialize pointer offsets move n0,n6 ;initialize coefficient offset move #coef,n3 ;initialize coefficient base address move #_points-1,m0 ;initialize address modifiers move m0,m1 ;for modulo(points) addressing move m0,m4 move m0,m5 move #-1,m3 ;linear addressing for coefficient base offset move #0,m2 ;initialize 256 point fft counter move m2,m6 ;initialize coefficient address modifier ;for reverse carry (bit reversed) addressing ; ; Do first and second Radix 2 FFT passes as four-point butterflies ; move x:(r0)+n0,x0 ;initialize butterfly tfr x0,a x:(r0)+n0,y1 ; do n0,_twopass tfr y1,b x:(r0)+n0,y0 add y0,a x:(r0),x1 ;ar+cr add x1,b r0,r4 ;br+dr add a,b (r0)+n0 ;ar'=(ar+cr)+(br+dr) subl b,a b,x:(r0)+n0 ;br'=(ar+cr)-(br+dr) tfr x0,a a,x0 y:(r0),b sub y0,a y:(r4)+n4,y0 ;ar-cr sub y0,b x0,x:(r0) ;bi-di add a,b y:(r0)+n0,x0 ;cr'=(ar-cr)+(bi-di) subl b,a b,x:(r0) ;dr'=(ar-cr)-(bi-di) tfr x0,a a,x0 y:(r4),b add y0,a y:(r0)+n0,y0 ;bi+di add y0,b x0,x:(r0)+n0 ;ai+ci add b,a y:(r0)+,x0 ;ai'=(ai+ci)+(bi+di) subl a,b a,y:(r4)+n4 ;bi'=(ai+ci)-(bi+di) tfr x0,a b,y:(r4)+n4 sub y0,a x1,b ;ai-ci sub y1,b x:(r0)+n0,x0 ;dr-br add a,b x:(r0)+n0,y1 ;ci'=(ai-ci)+(dr-br) subl b,a b,y:(r4)+n4 ;di'=(ai-ci)-(dr-br) tfr x0,a a,y:(r4)+ _twopass ; ; Do remaining 8 FFT passes as four 256 point Radix 2 FFT's using internal data ; and external coefficients. ; ; Each 256 point Radix 2 FFT consists of 8 passes. The first pass uses external ; input data and internal output data to move the data on-chip. Intermediate ; passes use internal input data and internal output data to keep the data ; on-chip. The last pass uses internal input data and external output data ; to move the data off-chip. ; move #odata,r4 ;load output data address do #4,_end_fft ;do 256 point fft four times move r4,ssh ;push output data address onto stack move r2,r0 ;get external data input address for first pass move m2,r3 ;update coefficient offset move #256/2,n1 ;initialize butterflies per group move #1,n2 ;initialize groups per pass do #7,_end_pass ;do first 7 passes out of 8 move #_intdata,r4 ;initialize A output pointer move n1,r5 move n1,n0 ;initialize pointer offsets lua (r5)-,n7 move n1,n4 move n1,r6 lua (r0)+n0,r1 ;initialize B input pointer lua (r4)+n4,r5 ;initialize B output pointer lua (r6)+,n4 move n4,n5 lua (r3)+n3,r6 ;initialize W input pointer move n4,n0 ; ; initialize butterfly move x:(r1),x1 y:(r6),y0 ;lookup -sine value move y:(r0),b mac x1,y0,b x:(r6)+n6,x0 y:(r1)+,y1 macr -x0,y1,b y:(r0),a ; do n2,_end_grp do n7,_end_bfy subl b,a x:(r0),b b,y:(r4) mac -x1,x0,b x:(r0)+,a a,y:(r5) macr -y1,y0,b x:(r1),x1 subl b,a b,x:(r4)+ y:(r0),b mac x1,y0,b y:(r1)+,y1 ;Radix 2 DIT butterfly kernel macr -x0,y1,b a,x:(r5)+ y:(r0),a ;with constant twiddle factor _end_bfy move (r1)+n1 subl b,a x:(r0),b b,y:(r4) mac -x1,x0,b x:(r0)+n0,a a,y:(r5) macr -y1,y0,b x:(r1),x1 y:(r6),y0 ;lookup -sine value subl b,a b,x:(r4)+n4 y:(r0),b mac x1,y0,b x:(r6)+n6,x0 y:(r1)+,y1 macr -x0,y1,b a,x:(r5)+n5 y:(r0),a ; _end_grp move n1,b1 lsr b n2,a1 ;divide butterflies per group by two lsl a b1,n1 ;multiply groups per pass by two move r3,b1 lsr b a1,n2 ;divide coefficient offset by two move b1,r3 move #_intdata,r0 ;intermediate passes use internal input data _end_pass ; ; Do last FFT pass and move output data off-chip to external data memory. ; move n7,r1 move ssh,r4 move (r1)+ move r1,n0 move r1,n1 ;correct pointer offset for last pass move r1,n4 move r1,n5 move #_points/4,n4 ;offset for output pointer A lua (r0)+,r1 ;initialize B input pointer lua (r4)+n4,r5 ;initialize B output pointer, first step move n4,n5 ;offset for output pointer B lua (r3)+n3,r6 ;initialize W input pointer move (r5)+n5 ;initialize B output pointer, second step move #0,m4 ;bit-reversed addressing for output pointer A move m4,m5 ;bit-reversed addressing for output pointer B move x:(r1),x1 y:(r6),y0 move (r5)-n5 ;predecrement output pointer A move x:(r5),a y:(r0),b do n2,_lastpass mac x1,y0,b x:(r6)+n6,x0 y:(r1)+n1,y1 ;Radix 2 DIT butterfly kernel macr -x0,y1,b a,x:(r5)+n5 y:(r0),a ;with one butterfly per group subl b,a x:(r0),b b,y:(r4) ;and changing twiddle factor mac -x1,x0,b x:(r0)+n0,a a,y:(r5) macr -y1,y0,b x:(r1),x1 y:(r6),y0 subl b,a b,x:(r4)+n4 y:(r0),b _lastpass move a,x:(r5)+n5 ;save last real result of these 256 move m0,m4 ;reinitialize pointers for modulo addr. move m0,m5 ; move m2,r6 ;get fft counter move n6,n2 ;get fft data input offset move m3,m2 ;external data pointer uses linear arithmetic move (r6)+n6 ;increment fft counter (bit reversed) move (r2)+n2 ;point to next 256 point fft input data move r6,m2 ;save fft counter _end_fft ; when fft is finished, jump back to see if data collection for next fft is completed ; jmp strt endm