MMDVM/MMDVMHost
1
0
Fork 0
mirror of https://github.com/g4klx/MMDVMHost synced 2025-12-25 19:05:39 +08:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #595 from F4FXL/FM_Ext

Browse Source 
Make FM Network work
This commit is contained in:
Jonathan Naylor
2020-05-31 11:28:15 +01:00
committed by GitHub
parent 1caffc1dad 17b49fde88
commit f18e253eac
6 changed files with 181 additions and 76 deletions
Download Patch File Download Diff File Expand all files Collapse all files

146
FMControl.cpp
View File

@@ -24,20 +24,40 @@
#include <cstdio>
#endif
const float EMPHASIS_GAIN_DB = 0.0F; //Gain needs to be the same for pre an deeemphasis
const unsigned int FM_MASK = 0x00000FFFU;
#define SWAP_BYTES_16(a) (((a >> 8) & 0x00FFU) | ((a << 8) & 0xFF00U))
const float DEEMPHASIS_GAIN_DB = 0.0F;
const float PREEMPHASIS_GAIN_DB = 13.0F;
const float FILTER_GAIN_DB = 0.0F;
const unsigned int FM_MASK = 0x00000FFFU;
CFMControl::CFMControl(CFMNetwork* network) :
m_network(network),
m_enabled(false),
m_incomingRFAudio(1600U, "Incoming RF FM Audio"),
m_preemphasis(0.3889703155F, -0.32900055326F, 0.0F, 1.0F, 0.2820291817F, 0.0F, EMPHASIS_GAIN_DB),
m_deemphasis(1.0F, 0.2820291817F, 0.0F, 0.3889703155F, -0.32900055326F, 0.0F, EMPHASIS_GAIN_DB)
m_preemphasis (NULL),
m_deemphasis (NULL),
m_filterStage1(NULL),
m_filterStage2(NULL),
m_filterStage3(NULL)
{
m_preemphasis = new CIIRDirectForm1Filter(8.315375384336983F,-7.03334621603483F,0.0F,1.0F,0.282029168302153F,0.0F, PREEMPHASIS_GAIN_DB);
m_deemphasis = new CIIRDirectForm1Filter(0.07708787090460224F,0.07708787090460224F,0.0F,1.0F,-0.8458242581907955F,0.0F, DEEMPHASIS_GAIN_DB);
//cheby type 1 0.2dB cheby type 1 3rd order 300-2700Hz fs=8000
m_filterStage1 = new CIIRDirectForm1Filter(0.29495028f, 0.0f, -0.29495028f, 1.0f, -0.61384624f, -0.057158668f, FILTER_GAIN_DB);
m_filterStage2 = new CIIRDirectForm1Filter(1.0f, 2.0f, 1.0f, 1.0f, 0.9946123f, 0.6050482f, FILTER_GAIN_DB);
m_filterStage3 = new CIIRDirectForm1Filter(1.0f, -2.0f, 1.0f, 1.0f, -1.8414584f, 0.8804949f, FILTER_GAIN_DB);
}
CFMControl::~CFMControl()
{
delete m_preemphasis ;
delete m_deemphasis ;
delete m_filterStage1;
delete m_filterStage2;
delete m_filterStage3;
}
bool CFMControl::writeModem(const unsigned char* data, unsigned int length)
@@ -63,62 +83,47 @@ bool CFMControl::writeModem(const unsigned char* data, unsigned int length)
bufferLength = 255U;
if (bufferLength >= 3U) {
#if defined(DUMP_RF_AUDIO)
FILE* audiofile = ::fopen("./audiodump.bin", "ab");
#endif
bufferLength = bufferLength - bufferLength % 3U; //round down to nearest multiple of 3
unsigned char bufferData[255U];
m_incomingRFAudio.getData(bufferData, bufferLength);
unsigned int nSamples = 0;
float samples[85U]; // 255 / 3;
// Unpack the serial data into float values.
unsigned int pack = 0U;
unsigned char* packPointer = (unsigned char*)&pack;
unsigned short out[168U]; // 84 * 2
unsigned int nOut = 0U;
short unpackedSamples[2U];
for (unsigned int i = 0U; i < bufferLength; i += 3U) {
short sample1 = 0U;
short sample2 = 0U;
unsigned int pack = 0U;
unsigned char* packPointer = (unsigned char*)&pack;
//extract unsigned 12 bit unsigned sample pairs packed into 3 bytes to 16 bit signed
packPointer[0U] = bufferData[i];
packPointer[1U] = bufferData[i + 1U];
packPointer[2U] = bufferData[i + 2U];
unpackedSamples[1U] = short(int(pack & FM_MASK) - 2048);
unpackedSamples[0U] = short(int(pack >> 12) - 2048);
//extract unsigned 12 bit samples to 16 bit signed
sample2 = short(int(pack & FM_MASK) - 2048);
sample1 = short(int(pack >> 12) - 2048);
//process unpacked sample pair
for(unsigned char j = 0U; j < 2U; j++) {
//Convert to float (-1.0 to +1.0)
float sampleFloat = float(unpackedSamples[j]) / 2048.0F;
// Convert from unsigned short (0 - +4095) to float (-1.0 - +1.0)
samples[nSamples++] = float(sample1) / 2048.0F;
samples[nSamples++] = float(sample2) / 2048.0F;
}
//De-emphasise and remove CTCSS
sampleFloat = m_deemphasis->filter(sampleFloat);
sampleFloat = m_filterStage3->filter(m_filterStage2->filter(m_filterStage1->filter(sampleFloat)));
//De-emphasise the data and any other processing needed (maybe a low-pass filter to remove the CTCSS)
for (unsigned int i = 0U; i < nSamples; i++)
samples[i] = m_deemphasis.filter(samples[i]);
#if defined(DUMP_RF_AUDIO)
if (audiofile != NULL)
::fwrite(samples, sizeof(float), nSamples, audiofile);
#endif
unsigned short out[170U]; // 85 * 2
unsigned int nOut = 0U;
// Repack the data (8-bit unsigned values containing unsigned 16-bit data)
for (unsigned int i = 0U; i < nSamples; i++) {
unsigned short sample = (unsigned short)((samples[i] + 1.0F) * 32767.0F + 0.5F);
out[nOut++] = (sample >> 8) & 0xFFU;
out[nOut++] = (sample >> 0) & 0xFFU;
// Repack the float data to 16 bit unsigned
unsigned short sampleUShort = (unsigned short)((sampleFloat + 1.0F) * 32767.0F + 0.5F);
out[nOut++] = SWAP_BYTES_16(sampleUShort);
}
}
#if defined(DUMP_RF_AUDIO)
if (audiofile != NULL) {
::fclose(audiofile);
audiofile = NULL;
FILE * audiofile = fopen("./audiodump.bin", "ab");
if(audiofile != NULL) {
fwrite(out, sizeof(unsigned short), nOut, audiofile);
fclose(audiofile);
}
#endif
return m_network->writeData((unsigned char*)out, nOut);
return m_network->writeData((unsigned char*)out, nOut * sizeof(unsigned short));
}
return true;
@@ -135,42 +140,41 @@ unsigned int CFMControl::readModem(unsigned char* data, unsigned int space)
if (space > 252U)
space = 252U;
unsigned char netData[168U];//84 * 2 modem can handle up to 84 samples (252 bytes) at a time
unsigned int length = m_network->read(netData, 168U);
unsigned short netData[84U];//modem can handle up to 84 samples (252 bytes) at a time
unsigned int length = m_network->read((unsigned char*)netData, 84U * sizeof(unsigned short));
length /= sizeof(unsigned short);
if (length == 0U)
return 0U;
float samples[84U];
unsigned int nSamples = 0U;
// Convert the unsigned 16-bit data (+65535 - 0) to float (+1.0 - -1.0)
for (unsigned int i = 0U; i < length; i += 2U) {
unsigned short sample = (netData[i + 0U] << 8) | netData[i + 1U];
samples[nSamples++] = (float(sample) / 32767.0F) - 1.0F;
}
// Pre-emphasise the data and other stuff.
for (unsigned int i = 0U; i < nSamples; i++)
samples[i] = m_preemphasis.filter(samples[i]);
// Pack the floating point data (+1.0 to -1.0) to packed 12-bit samples (+2047 - -2048)
unsigned int pack = 0U;
unsigned char* packPointer = (unsigned char*)&pack;
unsigned int j = 0U;
unsigned int i = 0U;
for (; i < nSamples && j < space; i += 2U, j += 3U) {
unsigned short sample1 = (unsigned short)((samples[i] + 1.0F) * 2048.0F + 0.5F);
unsigned short sample2 = (unsigned short)((samples[i + 1] + 1.0F) * 2048.0F + 0.5F);
unsigned int nData = 0U;
pack = 0;
pack = ((unsigned int)sample1) << 12;
pack |= sample2;
for(unsigned int i = 0; i < length; i++) {
unsigned short netSample = SWAP_BYTES_16(netData[i]);//((netData[i] << 8) & 0xFF00U)| ((netData[i] >> 8) & 0x00FFU);
// Convert the unsigned 16-bit data (+65535 - 0) to float (+1.0 - -1.0)
float sampleFloat = (float(netSample) / 32768.0F) - 1.0F;
data[j] = packPointer[0U];
data[j + 1U] = packPointer[1U];
data[j + 2U] = packPointer[2U];
//preemphasis
sampleFloat = m_preemphasis->filter(sampleFloat);
// Convert float to 12-bit samples (0 to 4095)
unsigned int sample12bit = (unsigned int)((sampleFloat + 1.0F) * 2048.0F + 0.5F);
// pack 2 samples onto 3 bytes
if((i & 1U) == 0) {
pack = 0U;
pack = sample12bit << 12;
} else {
pack |= sample12bit;
data[nData++] = packPointer[0U];
data[nData++] = packPointer[1U];
data[nData++] = packPointer[2U];
}
}
return j;//return the number of bytes written
return nData;
}
void CFMControl::clock(unsigned int ms)

11
FMControl.h
View File

@@ -25,8 +25,8 @@
// Uncomment this to dump audio to a raw audio file
// The file will be written in same folder as executable
// Toplay the file : aplay -f FLOAT_LE -c1 -r8000 -t raw audiodump.bin
//#define DUMP_RF_AUDIO
// Toplay the file : ffplay -autoexit -f u16be -ar 8000 audiodump.bin
// #define DUMP_RF_AUDIO
class CFMControl {
public:
@@ -45,8 +45,11 @@ private:
CFMNetwork* m_network;
bool m_enabled;
CRingBuffer<unsigned char> m_incomingRFAudio;
CIIRDirectForm1Filter m_preemphasis;
CIIRDirectForm1Filter m_deemphasis;
CIIRDirectForm1Filter * m_preemphasis;
CIIRDirectForm1Filter * m_deemphasis;
CIIRDirectForm1Filter * m_filterStage1;
CIIRDirectForm1Filter * m_filterStage2;
CIIRDirectForm1Filter * m_filterStage3;
};
#endif

2
Modem.cpp
View File

@@ -160,7 +160,7 @@ m_rxNXDNData(1000U, "Modem RX NXDN"),
m_txNXDNData(1000U, "Modem TX NXDN"),
m_txPOCSAGData(1000U, "Modem TX POCSAG"),
m_rxFMData(1000U, "Modem RX FM"),
m_txFMData(1000U, "Modem TX FM"),
m_txFMData(5000U, "Modem TX FM"),
m_rxTransparentData(1000U, "Modem RX Transparent"),
m_txTransparentData(1000U, "Modem TX Transparent"),
m_sendTransparentDataFrameType(0U),

4
SerialController.cpp
View File

@@ -304,6 +304,10 @@ bool CSerialController::open()
::cfsetospeed(&termios, B230400);
::cfsetispeed(&termios, B230400);
break;
case 460800U:
::cfsetospeed(&termios, B460800);
::cfsetispeed(&termios, B460800);
break;
default:
LogError("Unsupported serial port speed - %u", m_speed);
::close(m_fd);

43
Tools/DeEmphasis.py Normal file
View File

@@ -0,0 +1,43 @@
#based on https://github.com/gnuradio/gnuradio/blob/master/gr-analog/python/analog/fm_emph.py
import math
import cmath
import numpy as np
import scipy.signal as signal
import pylab as pl
tau = 750e-6
fs = 8000
fh = 2700
# Digital corner frequency
w_c = 1.0 / tau
# Prewarped analog corner frequency
w_ca = 2.0 * fs * math.tan(w_c / (2.0 * fs))
# Resulting digital pole, zero, and gain term from the bilinear
# transformation of H(s) = w_ca / (s + w_ca) to
# H(z) = b0 (1 - z1 z^-1)/(1 - p1 z^-1)
k = -w_ca / (2.0 * fs)
z1 = -1.0
p1 = (1.0 + k) / (1.0 - k)
b0 = -k / (1.0 - k)
btaps = [ b0 * 1.0, b0 * -z1, 0 ]
ataps = [ 1.0, -p1, 0 ]
# Since H(s = 0) = 1.0, then H(z = 1) = 1.0 and has 0 dB gain at DC
taps = np.concatenate((btaps, ataps), axis=0)
print("Taps")
print(*taps, "", sep=",", end="\n")
f,h = signal.freqz(btaps,ataps, fs=fs)
pl.plot(f, 20*np.log10(np.abs(h)))
pl.xlabel('frequency/Hz')
pl.ylabel('gain/dB')
pl.ylim(top=0,bottom=-30)
pl.xlim(left=0, right=fh*2.5)
pl.show()

51
Tools/PreEmphasis.py Normal file
View File

@@ -0,0 +1,51 @@
#based on https://github.com/gnuradio/gnuradio/blob/master/gr-analog/python/analog/fm_emph.py
import math
import cmath
import numpy as np
import scipy.signal as signal
import pylab as pl
tau = 750e-6
fs = 8000
fh = 2700
# Digital corner frequencies
w_cl = 1.0 / tau
w_ch = 2.0 * math.pi * fh
# Prewarped analog corner frequencies
w_cla = 2.0 * fs * math.tan(w_cl / (2.0 * fs))
w_cha = 2.0 * fs * math.tan(w_ch / (2.0 * fs))
# Resulting digital pole, zero, and gain term from the bilinear
# transformation of H(s) = (s + w_cla) / (s + w_cha) to
# H(z) = b0 (1 - z1 z^-1)/(1 - p1 z^-1)
kl = -w_cla / (2.0 * fs)
kh = -w_cha / (2.0 * fs)
z1 = (1.0 + kl) / (1.0 - kl)
p1 = (1.0 + kh) / (1.0 - kh)
b0 = (1.0 - kl) / (1.0 - kh)
# Since H(s = infinity) = 1.0, then H(z = -1) = 1.0 and
# this filter has 0 dB gain at fs/2.0.
# That isn't what users are going to expect, so adjust with a
# gain, g, so that H(z = 1) = 1.0 for 0 dB gain at DC.
w_0dB = 2.0 * math.pi * 0.0
g = abs(1.0 - p1 * cmath.rect(1.0, -w_0dB)) \
/ (b0 * abs(1.0 - z1 * cmath.rect(1.0, -w_0dB)))
btaps = [ g * b0 * 1.0, g * b0 * -z1, 0]
ataps = [ 1.0, -p1, 0]
taps = np.concatenate((btaps, ataps), axis=0)
print("Taps")
print(*taps, "", sep=",", end="\n")
f,h = signal.freqz(btaps,ataps, fs=fs)
pl.plot(f, 20*np.log10(np.abs(h)))
pl.xlabel('frequency/Hz')
pl.ylabel('gain/dB')
pl.ylim(top=30,bottom=0)
pl.xlim(left=0, right=fh*2.5)
pl.show()