Merge branch 'feat-fst280' of bitbucket.org:k1jt/wsjtx into feat-fst280

This commit is contained in:
Steven Franke 2020-08-04 10:17:16 -05:00
commit 45a1a116c9
2 changed files with 35 additions and 35 deletions

View File

@ -44,9 +44,9 @@ contains
integer ipk(1)
integer dat4(12)
logical lapdx,ltext
complex c0(0:NFFT1-1) !Spectrum, then analytic signal
complex c00(0:NFFT1-1) !Spectrum, then analytic signal
complex c0(0:NFFT1-1) !Snalytic signal
real s(900)
real s3a(-64:127,63)
real s3(-64:127,63)
real a(5)
data nc1z/-1/,nc2z/-1/,ng2z/-1/,maxaptypez/-1/
@ -64,7 +64,7 @@ contains
call packcall(hiscall(1:6),nc2,ltext)
call packgrid(hisgrid(1:4),ng2,ltext)
nSubmode=0
b90=20
b90=20.0 !8 to 25 is OK; not very critical
nFadingModel=1
! These probably need work:
@ -111,43 +111,36 @@ contains
ipk=maxloc(s)
f0=df2*ipk(1) - 0.5*baud !Candidate sync frequency
! do i=1,iz
! write(51,3051) i*df2,s(i)
!3051 format(f12.6,e12.3)
! enddo
c0(NFFT2/2+1:NFFT2)=0. !Zero the top half
c0(0)=0.5*c0(0)
call four2a(c0,nfft2,1,1,1) !Inverse c2c FFT
! call sync66(c0,f0,jpk,sync) !c0 is analytic signal at 6000 S/s
ntol=100
call sync66a(iwave,15*12000,nsps,nfqso,ntol,xdt,f0)
jpk=(xdt+0.5)*6000
call sync66a(iwave,15*12000,nsps,nfqso,ntol,xdt,f0,snr1)
jpk=(xdt+0.5)*6000 - 384 !### Empirical ###
if(jpk.lt.0) jpk=0
! Genie sync:
! jpk=3600
! xdt=jpk/6000.0 - 0.5
! f0=1510
! f0=1490
c00=c0
a=0.
! a(1)=-(f0 + 1.5*baud) !For sync66
a(1)=-(f0 + 2.0*baud) !For sync66a
call twkfreq(c00,c0,15*6000,6000.0,a)
xdt=jpk/6000.0 - 0.5
call spec66(c0(jpk:jpk+85*NSPS/2-1),s3)
call twkfreq(c0,c0,15*6000,6000.0,a)
call spec66(c0(jpk:jpk+85*NSPS-1),s3)
s3a=s3/maxval(s3)
do j=1,63
call pctile(s3(:,j),192,40,base)
s3(:,j)=s3(:,j)/base
enddo
call pctile(s3a,192*63,40,base)
s3a=s3a/base
s3lim=10.
where(s3a(-64:127,1:63)>s3lim) s3a(-64:127,1:63)=s3lim
call qra64_dec(s3a,nc1,nc2,ng2,naptype,0,nSubmode,b90, &
! Apply AGC
s3max=20.0
do j=1,63
xx=maxval(s3(-64:127,j))
if(xx.gt.s3max) s3(-64:127,j)=s3(-64:127,j)*s3max/xx
enddo
call qra64_dec(s3,nc1,nc2,ng2,naptype,0,nSubmode,b90, &
nFadingModel,dat4,snr2,irc)
snr2=snr2 + 6.0
if(irc.gt.0) call badmsg(irc,dat4,nc1,nc2,ng2)
decoded=' '
@ -171,6 +164,9 @@ contains
!###
endif
! write(61,3061) nutc,irc,xdt,f0,snr1,snr2,trim(decoded)
!3061 format(i6.6,i4,4f10.2,2x,a)
return
end subroutine decode

View File

@ -1,4 +1,4 @@
subroutine sync66a(iwave,nmax,nsps,nfqso,ntol,xdt,f0)
subroutine sync66a(iwave,nmax,nsps,nfqso,ntol,xdt,f0,snr1)
parameter (NSTEP=4) !Quarter-symbol steps
parameter (IZ=1600,JZ=352,NSPSMAX=1920)
@ -49,7 +49,7 @@ subroutine sync66a(iwave,nmax,nsps,nfqso,ntol,xdt,f0)
s1=s1/base
s1max=20.0
! Make this simpler: just add the 22 nonzero values.
! Apply AGC
do j=1,JZ
x(j)=maxval(s1(i0-64:i0+192,j))
if(x(j).gt.s1max) s1(i0-64:i0+192,j)=s1(i0-64:i0+192,j)*s1max/x(j)
@ -63,6 +63,7 @@ subroutine sync66a(iwave,nmax,nsps,nfqso,ntol,xdt,f0)
do i=-ia,ia
x=s1(i0+2+i,:)-s1(i0+i,:)
do lag=-15,15
! Make this simpler: just add the 22 nonzero values?
do n=1,4*85
j=n+lag+11
if(j.ge.1 .and. j.le.JZ) sync_sig(i,lag)=sync_sig(i,lag) + sync(n)*x(j)
@ -80,15 +81,18 @@ subroutine sync66a(iwave,nmax,nsps,nfqso,ntol,xdt,f0)
tsym=nsps/12000.0
xdt=jdt*tsym/4.0
snr1=maxval(sync_sig)/22.0
! do i=-64,64
! write(13,3013) nfqso+i*df,sync_sig(i,jj)
!3013 format(2f12.3)
! write(62,3062) nfqso+i*df,sync_sig(i,jj)
!3062 format(2f12.3)
! enddo
! do j=-15,15
! write(14,3014) j,j*dt4,sync_sig(ii,j)
!3014 format(i5,2f12.3)
! write(63,3063) j,j*dt4,sync_sig(ii,j)
!3063 format(i5,2f12.3)
! enddo
return
end subroutine sync66a