HEX文件格式不赘述,写里直接放上代码。请批评改正。
1 %%convert a signal data into hex file format 2 % data format:16bit 3 % signal length: less than 2^24-1 4 % author: Yang Li yangli0534@gmail.com 5 % data:2015.01.27 6 7 clear all; 8 close all; 9 clc; 10 11 %% fixed point or binary16 float point 12 fixed=1; 13 %% generate a signal waveform 14 % 15 T=10;%time length 16 fs=1e3;%sample rate 17 N=10000; 18 Period=N/fs/10; 19 f=1/Period; 20 t=linspace(0,(N-1)/fs,N); 21 22 %data=1:2^5-1; 23 data= sin(2*pi*f*t) +(rand(1,N)-0.5)/10; 24 % data=mod(50000*t,1000); 25 % plot(t,data) 26 dataBits=16; 27 byteBits=8; 28 29 %% convert into 16 bits fixed point number 30 if fixed == 1 31 maximum=max(abs(data));%maximum of absolute signal 32 scale=(2^15-1)/maximum;%scale 33 data=round(data*scale); 34 for i=1:1:length(data) 35 if data(i) <0 36 data(i) =2^16-abs(data(i)) ; 37 end 38 end 39 end 40 %% write 41 fh=fopen('data.hex','w');%open a new file (file path, file name) 42 bytesEveryRow=16;%write 16 bytes every row 43 dataEveryRow=bytesEveryRow/(dataBits/byteBits);%number of data every row 44 lengthData=length(data);%length of signal data 45 rowLengths=ceil(lengthData/dataEveryRow);%number of rows of the file to be written 46 47 for i=1:rowLengths % write a row each time 48 if(mod(i,hex2dec('1000'))==1) 49 baseaddr=dec2hex(floor(i/hex2dec('1000')),4); 50 checkSum=6+hex2dec(baseaddr(1:2))+hex2dec(baseaddr(3:4)); 51 checkSum=(dec2hex(bitand(256-mod(checkSum,256),255),2)); 52 fprintf(fh,[':02000004',baseaddr,checkSum,13,10]);%产生HEX码(end-1:end) 53 54 end 55 checkSum=0; 56 addr=dec2hex(mod((i-1)*16,hex2dec('10000')),4);%the start address of this row 57 checkSum=hex2dec(addr(1:2))+hex2dec(addr(3:4)); 58 if i*dataEveryRow <lengthData % numbers of data to be written greater than dataEveryRow 59 dataRow=dataEveryRow; 60 dataHex=''; 61 for j=8*(i-1)+1:1:8*(i) 62 if fixed == 1% 63 dataHexCurr=dec2hex(data(j),4);%16 bits fixed point 64 else 65 dataHexCurr=num2binary16hex(data(j));%number format: 16bits floatIEEE 754 half precison float point standard,result is a hex format 66 end 67 checkSum=checkSum+hex2dec(dataHexCurr(1:2))+hex2dec(dataHexCurr(3:4)); 68 dataHex=strcat(dataHex,dataHexCurr); 69 end 70 71 else 72 dataHex=''; 73 dataRow =lengthData-(rowLengths-1)*dataEveryRow; 74 for j=(rowLengths-1)*dataEveryRow+1:1:lengthData 75 if fixed == 1% 76 dataHexCurr=dec2hex(data(j),4);%16 bits fixed point 77 else 78 dataHexCurr=num2binary16hex(data(j));%number format: 16bits floatIEEE 754 half precison float point standard,result is a hex format 79 end 80 checkSum=checkSum+hex2dec(dataHexCurr(1:2))+hex2dec(dataHexCurr(3:4)); 81 dataHex=strcat(dataHex,dataHexCurr); 82 end 83 end 84 checkSum=checkSum+dataRow*dataBits/byteBits; 85 checkSum=(dec2hex(bitand(256-mod(checkSum,256),255),2)); 86 % disp([num2str(i),'--',dataHex,'--',checkSum]) 87 fprintf(fh,[':',dec2hex(dataRow*dataBits/byteBits,2),addr,'00',dataHex,checkSum,13,10]);%产生HEX码(end-1:end) 88 end 89 90 fprintf(fh,':00000001FF'); 91 fclose(fh);
代码中有定点浮点两种,其中浮点格式为IEEE754标准中的half-precision-floating point, 16bit, 1bit 浮号位,5bit指数位,10小数位,代码如下:
1 function [ hex ] = num2binary16he( number ) 2 %The IEEE 754 standard specifies a binary16 as having the following format: 3 4 %Sign bit: 1 bit 5 %Exponent width: 5 bits 6 %Significand precision: 11 bits (10 explicitly stored) 7 %S EEEEE MMMMMMMMMM 8 %INPUT: number is a random precision number 9 10 %OUTPUT : var binary is a 2's string of the binary16 11 % var decimail is the value of the number in decimal 12 13 %Author :Yang Li . 14 %Email: yangli0534@gmail.com 15 if real(number) >= 2^14 16 error('beyond the maximum :-2^14--+2^24'); 17 end 18 19 S=number<0 ; 20 E=0; 21 M=abs(number); 22 while((E>-15 && E<15)&&( M>=2 || M<1 )) 23 if(M>=2) 24 E=E+1; 25 M=M/2; 26 else 27 E=E-1; 28 M=M*2; 29 end 30 end 31 if ((E==-15 || E==15)&& M>=2 || M<1) 32 M=0; 33 else 34 M=round((M-1)*2^10) ; 35 end 36 37 38 number1=(-1)^S*(2^ E )*(1+M/2^10); 39 % binary=strcat(num2str(S),num2str(E),num2str(M)) 40 E =E+15; 41 binary=strcat(num2str(S),dec2bin(E,5), dec2bin(M,10)) ; 42 43 44 sReal=(str2double(binary(1))); 45 eReal =-15; 46 for i=2:6 47 eReal=eReal+str2double(binary(i))*2^(6-i); 48 end 49 % eReal=eReal*(-1)^str2double(binary(2)); 50 mReal = 0; 51 for i=7:16 52 mReal=mReal+str2double(binary(i))*2^(16-i); 53 end 54 55 numberReal=(-1)^sReal*2^eReal*(1+mReal/2^10); 56 err=num2str(abs(number-numberReal)); 57 decimal=numberReal; 58 % disp(['the origin data is : ',num2str(number)]); 59 % disp(['the float format is : ',binary]); 60 % disp(['so ,there is a error :',num2str(abs(number-numberReal))]); 61 62 num=0; 63 for i=1:16 64 num=num+str2double(binary(i))*2^(16-i); 65 end 66 hex=dec2hex(num,4); 67 end