Linrad with Raspberry Pi and RTL-SDR dongle.

[w6sz]

HI all,

Today Leif, SM5BSZ published one more of his fine informative Youtube video's,  this time on the Perseus SDR Receiver running
with a Raspberry Pi.

     https://www.youtube.com/watch?v=YWg6bqp-RyQ

I believe it to be the newer  Raspi 3?

This video is quite informative on the Raspberry Pi installation.

Not sure how many here are using the Raspi with Linrad or have experimented with it.

I have been using the board for some time with more or less success, though underruns were often a problem.
In particular with signal frequency selection with the mouse.

Now using this video I did a fresh installation of Linrad with a rtl-sdr dongle and it worked out of the box so to say.
All I did was changing the first FFT amplitude ( for using the 8 bit RTL-SDR dongle )

The load is 14% / 25%. I guess the real  key is the use of framebuffer. for video

Use the " ../configure's --with-help "  in the install process at the right time, read the results carefully.

Thought that some here on the reflector might want to know about this.

It is like Leif is mentioning, the use of this setup is for portable work with batteries as a back-end on 10 GHz.
radio also running off batteries.;

73 Rein W6SZ.

[Larry Dighera]

Hello Rein,

Today I built xlinrad on the RPi3 under Raspbian Jessie. Of course, I have
zero experience with linrad, so I'm embarking on yet another learning process.
Guessing at most of the xlinrad configuration options for use with the Airspy
Mini hardware, I chose the 3 MSPS option, however linrad seems to have
configured for 2.5 MSPS. So, I built the latest Airspy library, ran configure
and make, and still suffered a 1401 error (as I recall). As time permits, I'll
continue to explore linrad, and provide more specific information. And I'll
more closely scrutinize the configuration of linrad as shown in the new video
<https://youtu.be/YWg6bqp-RyQ> at the link address you provided.

Here Leif demonstrates how to begin to mitigate USB interference with the
Airspy Mini: <https://youtu.be/YFfyWBL4xpY>. That appears to be worth the
effort.

In the mean time, I'm excited to have yet another SDR program to run on the
RPi3 in addition to gqrx <https://github.com/csete/gqrx>, CubicSDR
<https://github.com/cjcliffe/CubicSDR>, OpenWebRX
<https://github.com/simonyiszk/openwebrx>, and qtcsdr:
<https://github.com/ha7ilm/qtcsdr>.

With regard, to sluggish performance, are you overclocking your RPi3?

Best regards,
Larry
WB6BBB

[Conrad PA5Y]

I am using Linrad v04-10 as both master and slave and the polarisation
control seems to be at odds with that reported by MAP65. MAP65 is
correct, I have been using it for many months and the reported
polarisations agree very well with a few others in my area that I trust.
However the max signal in the slaves seems to be 90 deg out. If I set
the master polarisation fixed in Linrad to the same polarisation that
MAP65 reports a peak I get very good correlation, however if I set the
slaves to the same the peak occurs 90 deg away. It is as though the
BBraw24 signal on the network is inverted somehow.

So 45 is actually 135 and 135 is actually 45 deg.

Can someone else please check this and confirm please.

Regards

Conrad PA5Y

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[Leif Asbrink]

Hi Rein,

You wrote:
"I believe it to be the newer Raspi 3?"

At the beginning I show the board and the box it came in.
It is surely a Raspberry Pi 2!!

You can compare with photos here:
https://www.stewright.me/2016/03/raspberry-pi-1-2-3-compared/

As far as I can understand, the Pi3 runs from a 30% faster
clock, but for Linrad, cache size is important as well as
latency somewhere in the system. On modern x86 platforms
I can run at nearly 100% cpu load without glitches, but
I can not run the Pi 2 much above 50% before USB fails on
the input. To what extent a more modern ARM cpu would allow
a higher utilization is unknown to me.

On the Pi2 it is essential to make the fft1 bandwidth
large in order to make the fft fit in the cache memory.

73

Leif

On Tue, 27 Sep 2016 13:48:47 -0700 (PDT)

[Leif Asbrink]

Hi Conrad,

Press 'X' on the processing screen. Then press 'D'
"Select pol. default" and finally press 'D' for
"Toggle sign for channel 2"

73

Leif

> --
> You received this message because you are subscribed to the Google Groups "Linrad" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to linrad+un...@googlegroups.com.
> To post to this group, send email to lin...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.

[Conrad PA5Y]

Yes this fixes the problem but why does it flip in the first place? The
BB audio in the master is correct, only when it is passed to the slaves
does the sign for ch2 change. Is this a bug? I already figured out how
to fix it I just wondered why?

Regards

Conrad

[Leif Asbrink]

Hi Conrad,

It seems you have told the master to invert the sign.
That would mean that the baseband signal already has
the sign inverted so you should not invert once again
in the slaves.

If however you send raw data to the slaves you would have
to invert the sign in the slaves if you have to do it
in the master. Raw data is not processed in any way.

This is the way it works in Linrad, it is not a bug.

[rei...@ix.netcom.com]

Hello Leif,

Thanks. Yesterday I looked at this version number and found out that
the 2 and 3 look very similar. Never realized that there was a 2 version.
Have 1 and 3 here.

I can be wrong but the use of framebuffer has really changed the situation
here and, I am using the Pi 3.
I wanted to make the point that for those having given up on using the PI, should give
it another look, using your latest video.
This, besides from being able to run from batteries.

( 7.3% 11% )

FFT BW 50 Hz ( 8192 )
FFT window 2
FFT version 3
Threads 4
Storage time 0

Output delay 200 ms
Base band storage 2s secs

with x11, I had to use long baseband storage times to prevent
underruning and even just stopping altogether, particular with
use of the mouse.
When using the mouse for frequency changes I see only very small
changes in the load and it takes very short times to recover.

Not sure this all makes sense but I am happy with it.

If you have a set of critical parameters I can try to run them and
give you results.

73 Rein W6SZ

[Larry Dighera]

I'm new to linrad having only built it a few days ago, so the user interface is
very unfamiliar to me. Is there a document that explains the functions of the
various boxes and arrows?

I'm running xlinrad 04.10 on a Raspberry Pi 3 with an Airspy Mini dongle under
X11 on Jessie, and was unable to get past the 1401 error:

INTERNAL ERROR: 1401
[1401]The file par_airspy is corrupted or not consistent with other
parameters. Use the U menu to set up your USB dongle and do not forget
to press W to save data in the main menu afterwards.

While the full-size Airspy does support a sample rate of 2.5 MSPS, the slowest
sample rate supported on the Airspy Mini is 3.0 MSPS. I found that editing
airspy.c, and changing from 2.500000 to 3.000000 overcame the 1401 issue. So,
Leif may want to add an additional specific setting for the Airspy Mini in the
next linrad release.

Now, although I can at last see signals on the linrad display, audio is
stuttering, and I'm seeing "RX UNDERRUN_ERROR"s and "RX A/D SPEED ERROR:
2960472Hz (normal 3000000 Hz)" and "DA SYNC ERRORS." I haven't a clue what to
do about them, but I have taken notes from Leif's video
<https://www.youtube.com/watch?v=YWg6bqp-RyQ> of linrad running on the RPi2,
and will attempt to apply the suggestions found there.

I attempted to build command-line linrad with libsvga video, but it complains
about the video device... and fails to run. I understand that the cause may be
an issue with the outdated Raspbian libsvga package, and I have downloaded the
version from Leif's repository, but not yet attempted to build it.

I have provided more detailed information below, but I'm wondering if someone
who has successfully used the Airspy Mini with linrad on the RPi3 might be
willing to post copies of their par_* files.

Any guidance toward resolving these issues is much appreciated.

Best regards,
Larry
WB6BBB

==========================================================================
Here's what in par_userint:

# cat par_userint

vga mode [12]
shm mode [0]
Screen width (%) [75]
Screen height (%) [75]
font scale [2]
mouse speed [8]
Max DMA rate [300]
Process priority [1]
Native ALSA [2081]
Rx input mode [4]
Rx rf channels [1]
Rx ad channels [2]
Rx ad speed [3000000]
Rx ad device no [10012]
Rx ad mode [0]
Rx da mode [0]
Rx da device no [1]
Rx min da speed [44100]
Rx max da speed [48000]
Rx max da channels [2]
Rx max da bytes [2]
Rx min da channels [2]
Rx min da bytes [2]
Rx soundcard radio [0]
Converter Hz [0]
Converter MHz [0]
Converter mode [0]
network flag [0]
Tx ad speed [0]
Tx da speed [0]
Tx ad device no [-1]
Tx da device no [-1]
Tx da channels [0]
Tx ad channels [0]
Tx da bytes [0]
Tx ad bytes [0]
Tx enable [0]
Tx pilot tone dB [0]
Tx pilot microsec. [0]
Tx soundcard radio [0]
Operator skil [1]
Max blocked CPUs [0]
Timer resolution [0]
Autostart [0]
Rx ad latency [0]
Rx da latency [0]
Tx ad latency [0]
Tx da latency [0]
Sample shift [0]
Min DMA rate [30]
Use ExtIO [0]
ExtIO type [0]
Transceiver mode [0]
PTT control [0]
check [2230408]

-------------------------------
# cat par_fm_wg

ytop [0]
ybottom [266]
xleft [0]
xright [1367]
yborder [152]
xpoints per pixel [25]
pixels per xpoint [0]
first xpoint [10845]
xpoints [33000]
avg1num [6]
spek avgnum [18]
waterfall avgnum [36]
spur_inhibit [0]
check [6660024]
yzero [1.000000000000000]
yrange [32768.000000000000000]
wat. db zero [20.000000000000000]
wat. db gain [0.250000000000000]

-------------------------------
# cat par_fm_mg

ytop [267]
ybottom [575]
xleft [1255]
xright [1304]
Scale type [0]
Avgnum [8192]
Tracks [0]
Check [1290024]
ygain [30.940000534057617]
yzero [-26.000000000000000]
Cal dBm [0.000000000000000]
Cal S/N [0.000000000000000]
Cal S/N sigshift [0.000000000000000]
Cal S-units [0.000000000000000]

-------------------------------

# cat par_fm_fg

ytop [562]
ybottom [622]
xleft [0]
xright [104]
yborder [601]
direction [1]
gain [0]
gain_inc [0]
fq inc [0.050000000000000002775557561563]
freq [7.000000000000000000000000000000]

cat par_fm_cg

ytop [360]
ybottom [496]
xleft [1305]
xright [1367]
Meter graph on [1]
Oscill on [0]

cat par_fm_bg

ytop [268]
ybottom [496]
xleft [2]
xright [686]
yborder [344]
fft3 avgnum [50]
pixels/point [1]
coh factor [8]
delay points [4]
AGC flag [0]
AGC attack [2]
AGC release [4]
AGC hang [2]
Waterfall avgnum [400]
Mouse wheel step [0]
Oscill ON [0]
Arrow mode [0]
Filter FIR/FFT [1]
Filter shift [0]
FM mode [0]
FM subtract [0]
FM factor [10]
ch2 phase [0]
Squelch level [0]
Squelch time [0]
Squelch point [0]
check [8880309]
filter flat [4394.531250000000000]
filter curved [0.000000000000000]
yzero [1.000000000000000]
yrange [4096.000000000000000]
dB/pixel [0.776851594448090]
yfac pwr [0.000000031292437]
yfac log [12.872471809387207]
bandwidth [1391.601562500000000]
first freq [0.000000000000000]
BFO freq [5325.000000000000000]
Output gain [5.000000000000000]
Waterfall gain [1.000000000000000]
Waterfall zero [20.000000000000000]
Oscill gain [0.100000001490116]

-------------------------------
# cat soundboard_rx_init.log

sound system = native ALSA
SETUP FOR RX INPUT IS NOW:

SDR DEVICE = AIRSPY

Serial no = 0x04A464C83618440B
Sampling rate = 2.5 Mhz, Xtal adjust = 0 ppb
Gain mode = Linearity, Bandwidth = 1.5 MHz
Bias T (5V) = OFF
SETUP FOR RX OUTPUT IS NOW:
Testing hw:CARD=0
device found: alsa_dev_plughw_pcm_name=plughw:0,0
(alsa_get_dev_pcm_name: n=0, retcode =0)


SOUNDCARD = bcm2835 ALSA
device = bcm2835 ALSA, bcm2835 ALSA (plughw:0,0)
device number = 0
hostapi = Native ALSA
min da sample rate = 8000
max da sample rate = 48000
min da bytes = 1
max da bytes = 2
min da channels = 1
max da channels = 2
DMA rate min=30 max=300

USER STARTED THE SELECTION OF A NEW INPUT DEVICE

goto begin_rx_set_io


sound system = native ALSA
SETUP FOR RX INPUT IS NOW:

SDR DEVICE = AIRSPY

Serial no = 0x04A464C83618440B
Sampling rate = 2.5 Mhz, Xtal adjust = 0 ppb
Gain mode = Linearity, Bandwidth = 2.0 MHz
Bias T (5V) = OFF
SETUP FOR RX OUTPUT IS NOW:

NOT YET SELECTED: (Select Menu Option 'B')
DMA rate min=30 max=300

goto begin_rx_set_io


sound system = native ALSA
SETUP FOR RX INPUT IS NOW:

SDR DEVICE = AIRSPY

Serial no = 0x04A464C83618440B
Sampling rate = 2.5 Mhz, Xtal adjust = 0 ppb
Gain mode = Linearity, Bandwidth = 2.0 MHz
Bias T (5V) = OFF
SETUP FOR RX OUTPUT IS NOW:

NOT YET SELECTED: (Select Menu Option 'B')
DMA rate min=30 max=300

USER STARTED THE SELECTION OF A NEW OUTPUT DEVICE
SELECT SOUNDCARD DEVICE FOR RX OUTPUT FROM LIST:
device soundcard-device-name device
sample-rate channels bytes
seq-no (soundcard-name) pcm-name min /
max min/max min/max
------ -------------------------------------------------- ----------
----------- ------- -------
Testing hw:CARD=0
device found: alsa_dev_plughw_pcm_name=plughw:0,0
(alsa_get_dev_pcm_name: n=0, retcode =0)

Testing hw:CARD=0
device found: alsa_dev_plughw_pcm_name=plughw:0,0
(alsa_get_dev_pcm_name: n=0, retcode =0)

0 bcm2835 ALSA, bcm2835 ALSA ( bcm2835 ALSA) plughw:0,0 8000/48000
1/2 1/2
Testing hw:CARD=0
device found: alsa_dev_plughw_pcm_name=plughw:0,1
(alsa_get_dev_pcm_name: n=1, retcode =0)

Testing hw:CARD=0
device found: alsa_dev_plughw_pcm_name=plughw:0,1
(alsa_get_dev_pcm_name: n=1, retcode =0)

1 bcm2835 ALSA, bcm2835 ALSA ( bcm2835 ALSA) plughw:0,1
44100/48000 2/2 2/2
Testing hw:CARD=0
alsa_get_dev_pcm_name:end of available card/devices reached
(alsa_get_dev_pcm_name: n=2, retcode =-1)

Enter selected device-seq-nmbr >
User selected 1
Testing hw:CARD=0
device found: alsa_dev_plughw_pcm_name=plughw:0,1
(alsa_get_dev_pcm_name: n=1, retcode =0)

Testing hw:CARD=0
device found: alsa_dev_plughw_pcm_name=plughw:0,1
(alsa_get_dev_pcm_name: n=1, retcode =0)


goto begin_rx_set_io


sound system = native ALSA
SETUP FOR RX INPUT IS NOW:

SDR DEVICE = AIRSPY

Serial no = 0x04A464C83618440B
Sampling rate = 2.5 Mhz, Xtal adjust = 0 ppb
Gain mode = Linearity, Bandwidth = 2.0 MHz
Bias T (5V) = OFF
SETUP FOR RX OUTPUT IS NOW:
Testing hw:CARD=0
device found: alsa_dev_plughw_pcm_name=plughw:0,1
(alsa_get_dev_pcm_name: n=1, retcode =0)


SOUNDCARD = bcm2835 ALSA
device = bcm2835 ALSA, bcm2835 ALSA (plughw:0,1)
device number = 1
hostapi = Native ALSA
min da sample rate = 44100
max da sample rate = 48000
min da bytes = 2
max da bytes = 2
min da channels = 2
max da channels = 2
DMA rate min=30 max=300

Normal end
---------------------------------------------

==========================================================================

[Leif Asbrink]

Hello Rein,

> I can be wrong but the use of framebuffer has really changed the situation
> here and, I am using the Pi 3.
> I wanted to make the point that for those having given up on using the PI,
> should give it another look, using your latest video.
> This, besides from being able to run from batteries.
>
> ( 7.3% 11% )

Low cpu load - but you did not tell what sampling rate you have set.

> FFT BW 50 Hz ( 8192 )

This can be critical.You are using version=3 so two transforms
plus the sine/cosine and bit reversal table have to fit in the cache.
Something like 2*2*8192floats(fft)+2*4096(sin/cos)+8192short(table)
In all 121 kilobytes. The cache would also have to store some code.
Since BW=50 Hz gives a 8192 transform, the total nominal bandwidth
is 210 kHz which means your dongle is probably running at 230 kHz.
This is a low speed, you could run it at 2.4 MHz and get 10 dB
better dynamic range, but then - at least on my Pi2 it would be
necessary to set the fft bw to 2 kHz.

> FFT window 2
> FFT version 3
> Threads 4
> Storage time 0

These look good.

> Output delay 200 ms
> Base band storage 2s secs
>
> with x11, I had to use long baseband storage times to prevent
> underruning and even just stopping altogether, particular with
> use of the mouse.

There are two options with x11. With or without MIT-SHM.

> When using the mouse for frequency changes I see only very small
> changes in the load and it takes very short times to recover.
>
> Not sure this all makes sense but I am happy with it.

It does make sense. For microwaves I guess you have adequate
dynamic range so there would be no reason to use a higher sampling
rate.

The first mixer bandwidth reduction may be critical. If you make it
bigger, the baseband sampling rate gets lower and the work to compute
the new output when you click on a new frequency would be smaller.
It could also be an adwantage to reduce the output delay margin.

73

Leif

[Leif Asbrink]

Hi Larry,

> I'm new to linrad having only built it a few days ago, so the user interface is
> very unfamiliar to me. Is there a document that explains the functions of the
> various boxes and arrows?

There is the F1 or ! help.

Try it on the main menu first.

On a processing screen, place the mouse on black screen outside windows
and press F1. This would highlight all "mouse objects."

You can place the mouse on any "mouse object" and get info about what
it is used for by pressing F1 or !

> I'm running xlinrad 04.10 on a Raspberry Pi 3 with an Airspy Mini dongle under
> X11 on Jessie, and was unable to get past the 1401 error:
>
> INTERNAL ERROR: 1401
> [1401]The file par_airspy is corrupted or not consistent with other
> parameters. Use the U menu to set up your USB dongle and do not forget
> to press W to save data in the main menu afterwards.

This is corrected in Linrad 04.11 which is not yet released.
The current development version is available here:
svn checkout https://svn.code.sf.net/p/linrad/code/trunk linrad

> Now, although I can at last see signals on the linrad display, audio is
> stuttering, and I'm seeing "RX UNDERRUN_ERROR"s and "RX A/D SPEED ERROR:
> 2960472Hz (normal 3000000 Hz)" and "DA SYNC ERRORS." I haven't a clue what to
> do about them, but I have taken notes from Leif's video
> <https://www.youtube.com/watch?v=YWg6bqp-RyQ> of linrad running on the RPi2,
> and will attempt to apply the suggestions found there.

Set the fft1 bandwidth to 2 kHz, no of B threads to 4 with fft version 3
to start with.

> I attempted to build command-line linrad with libsvga video, but it complains
> about the video device... and fails to run. I understand that the cause may be
> an issue with the outdated Raspbian libsvga package, and I have downloaded the
> version from Leif's repository, but not yet attempted to build it.

Build it!!

You can now run xlinrad in a terminal window with the front buffer device.
In 04.10 you had to compile flinrad for that. Hopefully the new strategy
with just one executable will be self-explanatory.

> I have provided more detailed information below, but I'm wondering if someone
> who has successfully used the Airspy Mini with linrad on the RPi3 might be
> willing to post copies of their par_* files.

I suggest you try SSB, CW or AM mode first. Linrad does not have an efficient
detector for wideband FM. It uses atan2 for each (complex) sample in the
baseband (300 kHz for wfm) and that might be impossible for the Raspberry
Pi.

The important file: par_fm is missing.

73

Leif

[Mike Waters]

Larry,

There is also http://physics.princeton.edu/pulsar/K1JT/Linrad_On-Screen_Controls.pdf.

Between that, what Leif said below, and his YouTube videos, you and I will eventually become proficient with all of that. :-)

73, Mike

www.w0btu.com

[Conrad PA5Y]

No I have not told the master to invert the sign which is why I brought
it up in the first place. Angle for Ch1 is 0 degrees and Ch2 is + on the
master instance. The start polarisation on the master is 90. Does this
effect the output?

Good news is that I finally have the slaves and MAP65 working perfectly
over the network with stable and reliable timings. I basically ended up
swapping the PC that was running the master. I am running the master and
45 and 135 deg slaves on the laptop and MAP65 and the 90 deg pol on my
main PC.

I tried several PCs and basically Dell PCs always work fine providing
the BBRaw24 signals on the network, no drift and only very small but
consistent delays. HP PCs using the same NIC chipset do not work. I even
have 2 network cards in one of the HP PCs and run one network interface
connected to my Afedri and the other connected to the network proper.
Still large variable delays.

Dell PCs (3 of them) all fine. So goodness knows what is going on, same
chipset, same drivers all on defaults?? Very odd! Anyway that is what I
found with 3 HP PCs and 3 Dell PCs, all are various flavours of i5
processor and have 8GB of memory. All but the laptop are Windows 7
professional.

Hardly scientific but that is what I observed. Now I will enjoy using
what I have.

Regards

Conrad PA5Y

[Roger Rehr W3SZ]

Interesting, Conrad!

I have never had the network timing problem that you described, whether
running Linux or Windows XP or 7 or 8.1 or 10, with Core2Duos, I5s and
I7s. Long ago my machines included some Dells, but for the past 5 years
or so all machines used here with Linrad have been homebrew. I never
had an HP machine in use with Linrad. So maybe the message that your
data is giving is, "stay away from HP computers for this purpose".
Oops...that will probably prompt 15 folks to write in about how their
HPs work fine running Linrad with 42 slaves! Even if that happens, my
advice will remain the same :)

Thanks for posting and have a great week!

73,

Roger
W3SZ

[Larry Dighera]

Hello Mike,

Thank you very much for the link to the linrad 'cheat sheet.' That's what I
need to get started.

Best regards,
Larry

[Leif Asbrink]

Hi Conrad,

Today I had a look at the polarization in master vs slave.

The master will send the baseband signal that is green in
the spectrum as H to the slave and the red one as V.

If you use adaptive polarization in the master, successful
adaption will always give a horizontal signal in the slaves.

In your case, select a fixed H polarization in the master,
then you can set polarizations as desired in the slaves.

73

Leif

[Conrad PA5Y]

Hello Leif OK I will do so, in the past I had the master set to fixed
Vertical and I used VAC as an audio output for WSJT , however I can
easily use a slave for the Vertical so that is fine.

It is really great fun having the 45 deg increments in the slaves, it
has added another dimension to my EME operating.

Regards

Conrad PA5Y

[Larry Dighera]

Hello Leif,

Thank you for your assistance and your contributions to the radio amateur
community.

The linrad user interface is a bit cryptic, but the techniques you have
provided for getting help with the various controls are useful and very much
appreciated.

AM mode is less CPU intensive on the RPi3 than FM; thanks for that insight. It
will enable me to continue familiarizing myself with linrad operation and
configuring.

I built your (development) Linrad 04.11; thank you for the link. Indeed, it
does now work with the Airspy Mini on the RPi3. Very nice, but screen
processing is CPU loading is very high.

I have built libsvga from the svgalib-1.4.3.tar.gz source on your web site, but
it fails to make svgalib_helper with this error:

Mon Oct 3 14:32:15 PDT 2016
# make install
grep: /lib/modules/4.4.21-v7+/build/include/linux/device.h: No such file or
directory
rm -f /dev/svga /dev/svga?
mknod -m 666 /dev/svga c 209 0
mknod -m 666 /dev/svga1 c 209 1
mknod -m 666 /dev/svga2 c 209 2
mknod -m 666 /dev/svga3 c 209 3
mknod -m 666 /dev/svga4 c 209 4
mkdir -p /lib/modules/4.4.21-v7+/kernel/misc
install -m 0644 -c svgalib_helper.ko /lib/modules/4.4.21-v7+/kernel/misc ||
install -m 0644 -c svgalib_helper.o /lib/modules/4.4.21-v7+/kernel/misc
install: cannot stat 'svgalib_helper.ko': No such file or directory
install: cannot stat 'svgalib_helper.o': No such file or directory
Makefile:63: recipe for target 'modules_install' failed
make: *** [modules_install] Error 1

I'm guessing, that the lack of device.h will prevent svgalib_helper from ever
being built on the RPi under Raspbian Jessie.

Any insight you may be able to offer how to overcome this issue will be very
much appreciated. When I attempt to run linrad in terminal mode, it indicates
it is using the FBDEV, but fails with this error:

Using fbdev, 7239KB at 3d4da000
svgalib: Cannot open /dev/svga
Is svgalib_helper module loaded?

I also loaded the Raspbian svgalib modules (libsvga-bin, svgalib1, and
svgalib-dev as I recall), but same issue.

I have attached a zip file of my par_* files and a screen dump, so that you
might more fully appreciate what I'm experiencing. (I've also e-mailed this
message directly to you, so that you'll get the attachment.)

I'm sorry to burden you with my support issues, as I'm sure you have much more
important things to attend to. But, with over 20 million Raspberry Pies having
been sold, I believe it is worthwhile for linrad to support that platform to
the extent possible.

Alex has just released a version of gqrx that finally runs on the RPi2 and
RPi3: http://gqrx.dk/news/gqrx-2-6-released

Thank you again.

Best regards,
Larry
WB6BBB

[Dominique Dehays]

Hi all ,

I agree , I am an almost daily watcher of Leif's videos. I'm reluctant
to create a youtube account, so I'm not doing the "LIKE" , but I do like
them.
Sometimes it is not easy to see on the screen where the mouse pointer is
located .
In this video from G7CNF , he is using a big yellow mouse pointer which
is very useful.

https://www.youtube.com/watch?v=ON5G4IIx6j8

Maybe it could be an improvement if U could use the same Leif?

Tnx very much for the outstanding and very useful work

Dom/F6DRO

[Leif Asbrink]

Hi Dom,

> Sometimes it is not easy to see on the screen where the mouse pointer is
> located .

I am sorry for that.

> In this video from G7CNF , he is using a big yellow mouse pointer which
> is very useful.
>
> https://www.youtube.com/watch?v=ON5G4IIx6j8
>
> Maybe it could be an improvement if U could use the same Leif?

It is too complicated. I tried long ago to make recordings by use
of screen capturing software, but that was not practical. Such
software (under windows) could do the trick of the G7CNF video.

It would be possible to change the mouse pointer in linrad, but
that would take a lot of time for me. Maybe someone else could
supply modified versions of xmain.c and wmain.c where
cross_cursor = XCreateFontCursor(xdis, XC_diamond_cross);
respectively LoadCursor (NULL, IDC_CROSS);
would be replaced by something else that would provide more
visible cursors to use when making videos from the screens with
a camera.

73

Leif

[f6...@orange.fr]

Hello all,
I am in process of setting a Linrad Master to feed 4 x Linrad slaves and
also MAP65 around an Afredi dual SDR radio.
The SDR , Linrad master , Slaves and MAP65 are on the same computer (I7)
windows10.

On the Linrad Master I used the send adress 192.168.1.255 to broadcast at
the same time through the Linrad slaves (send baseband 24 bit port 50007)
and MAP65 ( send timf2 port 50004).

I do not know if it works ( to broadcast to the Linrad Slaves and MAP65) but
it is bloking my connection to internet . I know very little on network but
I just saw on internet that using such a broadcast adress will affect all
adress between 192.168.1.1 to 192.168.1.254 . My http livebox to internet
adress is 192.168.1.1 .

Any suggestions please ?
Thank you
73
Jacques F6BKI


---
L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast.
https://www.avast.com/antivirus

[Leif Asbrink]

Hello Jacques,

Since everything is on the same computer you do not need
multicasting. Send to 127.0.0.1 or in case you use fixed
IP addresses, you might also use the address of the
particular computer.

73

Leif

> --
> There is an excellent Linrad User Guide by Gaetan, ON4KHG, at:
> http://bidelec.perso.sfr.fr/Linrad/images/Linrad%20Installation%20&%20Configuration%20User%20Guide%20-%20V1-0.pdf
> ---

> You received this message because you are subscribed to the Google Groups "Linrad" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to linrad+un...@googlegroups.com.
> To post to this group, send email to lin...@googlegroups.com.

> Visit this group at https://groups.google.com/group/linrad.

[f6...@orange.fr]

Hello Leif ,
If I set send at 127.0.0.1 on Linrad master and set 127.0.0.1 on the other
four Linrad slaves it works only on the first slave not the other.
Same if I use the particular PC adress as send , looks like Linrad master
can only feed one Linrad slave ??

Tks
73
Jacques F6BKI

-----Message d'origine-----
From: Leif Asbrink
Sent: Saturday, November 26, 2016 6:08 PM
To: lin...@googlegroups.com
Subject: Re: [Linrad] Network issues

[Roger Rehr W3SZ]

Hi Jacques,

Your experience is the same as mine. From my webpage of several years
ago: "After selecting "N" from the main menu of the Master instance of
Linrad, set the Master Linrad version up to send to a group address that
will be acceptable to both MAP65 and the slave instances of Linrad. My
network is 192.168.1.0, so I use 192.168.1.255. I found that I couldn't
use the obvious choice of 127.0.0.1 when running multiple Linrad slaves;
only one connection was permitted".

http://www.nitehawk.com/w3sz/LinradMAP65.htm

73,

Roger W3SZ

[f6...@orange.fr]

Hello all, getting close ! but

Roger 192.168.1.255 is blocking my internet box 192.168.1.1 ?
Leif 127.0.0.1 OK for my internet and MAP65 but only one Linrad slave
192.168.1.10 ( PC adress ) OK for my Internet and MAP65 but only one
slave
239.255.0.1 OK for my internet and all the Linrad slaves but not
for MAP65
192.168.0.255 OK for my internet but not for MAP65 nor Linrad slaves

None of the above settings works so far at the same time for my internet and
the Linrad slaves and MAP65 at the same time .
Any other options ??
Thank you all

73
Jacques F6BKI

-----Message d'origine-----

From: Roger Rehr W3SZ
Sent: Saturday, November 26, 2016 11:09 PM

[Alex Artieda]

Hi All

I use the Microsoft Virtual Loopback adapter to create a separate network, and then the UDP overflow will happens on that adapter and not in your entire network, if slaves runs in other computer install a second network card and make separate networks.

Regards
Alex, ZS6EME

[Roger Rehr W3SZ]

Yes,

In the past I had blocking problems when splitting the MAP65/Linrad
entities between 2 computers and I just run the MAP65/Linrad stuff on a
separate network that was only used for them. That eliminated the
blocking problems.

Currently I run everything on an I5 Windows 10 machine with Gigabit
Ethernet and I have noted no blocking problems, so I take no special
measures at all.

If you have blocking problems when using a single computer, then I would
follow Alex's excellent advice. If you don't know how to install the
Loopback Adapter from the Device Manager, this will show you:
https://www.youtube.com/watch?v=SnOZdOYkLns

Note that the Microsoft Loopback Adapter was renamed in Windows 8 and
thereafter. The new name is "Microsoft KM-TEST Loopback Adapter". So
when using the Add Hardware Wizard to manually add a network adapter,
choose Manufacturer "Microsoft" and choose network adapter "Microsoft
KM-TEST Loopback Adapter".

Good luck, and
73,
Roger W3SZ

[f6...@orange.fr]

Hello all, thank you for your input .
I am going to put MAP65 on the side for now , my main interest at the moment
is 6m EME , there are not so many stations actives .

Thank you
73
Jacques F6BKI

-----Message d'origine-----
From: Roger Rehr W3SZ

Sent: Sunday, November 27, 2016 2:41 PM