[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Linrad] Re: Networking issues

Hi Russ,

> Hi Leif. Well I am making some progress, but still some issues.  First I
> tried making the frequency range the same on both versions as you described.
> I had tried that before but missed the idea that I had to hit 'apply' after
> making the changes.  So when I tried it, I first set the lower value to 14.2
> and the upper to 14.8.  But when I hit apply, the range changed.  The bottom
> stayed close, but the top reduced down to around 14.5.  Tried that a few
> times to make sure.

Parameters in Linrad are of three kinds. 

1) The first kind relates to hardware: Soundcard, (or SDR-IQ etc), 
screen, fonts, network, parallel ports etc. These parameters can 
only be changed from the main menu in the setup functions.
S for screen, U for input, N for network etc. These parameters
are stored in par_userint, par_sdr14, par_network and similar.

2) Fixed parameters for a particular receive mode. Things like first
FFT bandwidth, use or not use the second FFT, decimation rate, use
16 or 32 bit arithmetics etc. These parameters are stored in par_cw,
par_ssb and similar. One for each rx mode.

3) Parameters that can be changed on the fly while operating.
These are changed with mouse or keyboard while receiving is 
running and they are stored in par_ssb_wg, par_ssb_bg, par_cw_wg
and so on. There is one file for each window in each receive mode.

When you first set the lower value by typing in the digits and then
enter Linrad will check whether the frequency is within range. You
the lowest possible frequency is your center frequency-48 kHz. 
(Assuming your SDR-IQ is still configured for MAP65-IQ) The
sampling rate (near 96kHz) belongs to type 1 above and can not be 
changed on the fly. The fact that you could set 14.2 sucessfully 
means that your center frequency was 14.248 or below. You can then
type 14.8 into the upper frequency box, but when you hit enter the
test would change the number to 14.096 (or below.) when you hit
You report you got the limits 14.2 and 14.5 which means a 300 kHz span
so I conclude you have set a much higher sampling rate than 96 kHz 
for the SDR-IQ. 300 kHz is very high for this hardware that uses 
USB 1.0. It would be a good idea to use a less extreme speed when you
are new to everything in Linrad. The manufacturer specification for SDR-IQ
says the highest sampling rate is 196 kHz. You can go higher in Linrad
but at some point there will be problems. I suggest you start with a sampling
speed of 150 kHz or less.

> Then I opened both par_ssb_wg files and made the slave one be equal to the
> master.  After doing that, I was able to get the range to approximate that
> of the master, but could not get it to be exactly the same.  However I was
> able to see the same signals in the range displayed.  But I still have the
> problem where the level on the master has to be at least 10 db higher than
> normal in order to get a reasonable signal on the slave.
The wide graph and its place on the screen are parameters in par_xxx_wg.
To get identical results, everything has to fit inside the screen. If your
master screen is bigger Linrad will contract the graph to make it fit and
that changes the endpoint frequencies.

> And I still see that when I change center frequency on the master, the slave
> does not know it.
OK. This is a bug. Every UDP package contains the frequency and a change should
be noted by the slave. I have to look at that but it will take some time because
to become 3.09 is in a messy state right now.

> It continues to show the old frequency range - even if I
> change to a new band.  If I x-out twice on the slave and then hit D again,
> the range will then reflect the new frequency. 
OK. That means that the bug is in the slave. You may type 'X', then 'B'
to get the new frequency.

> However after changing from
> 20 to 40 meters and then to 80 meters, I noticed that the slave range was
> somewhat expanded again.  The master showed 3.82 to 3.86 while the slave
> shows 3.825 to 3.85.  I looked at the slave's par_ssb_wg file again and see
> that some of the values had changed back to what they were before.
Linrad has few restrictions on what the user can do. You seem to do extreme

The total span you have is 40 kHz only in the main spectrum. Why? 
The purpose of the main spectrum is to display the entire frequency 
range of your hardware. (That is needed for you to monitor red/white
points in the main spectrum so you can set blanker thresholds correctly.)

You may have many reasons for the small frequency span. Maybe you have
set the sampling speed of the SDR-IQ to about 40 kHz in order to place
strong signals outside the frequency range that is transmitted via USB.
The USB channel has a limited dynamic range of 16 bits and SDR-IQ uses
24 bits internally.
> I attempted to make the files identical again, and now the slave always
> errors out when I click on the waterfall to start audio out, with a message:
>  [1058]Output can not start because the parameter "output delay margin ms"
> is too small.
>  Maybe the max DMA rate is set too low.
This error may be caused by many kinds of timing problems.

> The max DMA rate was set to 100, but I tried changing it to 500 with no
> improvement.  Something else must have changed but I am sure I did not make
> any changes except in the par_ssb_wg file.  Replacing that file with the
> original still does not cure the problem, so I cannot make any more tests
> tonight.  Hopefully you can tell me how to fix this new problem.
I have no idea what you are trying to do. Maybe you have set small value
for the first mixer bandwidth reduction in combination with a baseband 
filter in many points and also to run the baseband filter and resampler 
in the time domain. That would cause CPU overload and maybe this error 

I suggest you play a bit more with a single instance of Linrad with your
SDR-14 and get an understanding of the differend processing blocks that 
you use. For use with MAP65 you do not need to click any signal in Linrad
and then you would not need to know anyting about the baseband. But you
obviously clicked the waterfall and then baseband processing would start.

Use the mouse and the F1 help to read what all the different buttons 
in the baseband window do. (F1 with the mouse outside all windows, on black
screen, will show where the buttons are.)

Once you are happy with the processing you can do in the master and have
become familiar with the arrows in the upper corners and so on you can
copy all the par_ssb*.* files to the slave directory. Provided that you
have specified at least the same screen size (in pixels) for the slave
it should then become identical to the master.


Leif / SM5BSZ

You received this message because you are subscribed to the Google Groups "Linrad" group.
To post to this group, send email to linrad@xxxxxxxxxxxxxxxx
To unsubscribe from this group, send email to linrad+unsubscribe@xxxxxxxxxxxxxxxx
For more options, visit this group at http://groups.google.com/group/linrad?hl=en