The optimal configuration of Linrad depends very much on your application.
For example you will want to configure Linrad quite differently for 144MHz EME
than for 7MHz CW.
It also depends on your hardware setup; my station for 144MHz JT65B EME is shown
in the diagram below, with the transverter output on 28MHz..
The SDR-IQ is SMALL, less than 100mm in size, and draws its power entirely from the USB2 computer port.
The SDR-IQ samples its RF input at a rate of 66.667MHz, which gives it
continuous coverage up to 30MHz.
However the maximum rate of data transfer to the USB port is only around 196KHz;
the exact value depends on both hardware and operating system. This gives
a bandwidth up to about 190kHz.
You do have to understand just a little of how the SDR-IQ works, because you
have to program decimation factors from Linrad.
The downsampling from 66.667MHz to no more than 196kHz is called Decimation.
The term Decimation originally referred to a form of punishment in the Roman
army, whereby which one soldier in every ten, chosen randomly, was
executed by his comrades.
The word has come to be used to describe any large reduction in digital sample
rate.
The decimation process is analogous to frequency mixing in a conventional
analogue receiver. Just as you would not choose to mix down from 28MHz to 455kHz
in a single step because of image frequency problems, so digital decimation is
also done in several smaller steps.
The SDR-IQ uses 3 stages of decimation, each of which must be an integer. They
are called CIC2, CIC5 and RCF, and the product of these three integers gives the
total decimation, which is the factor by which the sample rate is reduced.
The final bandwidth transferred to Linrad is equal to 66.667MHz divided by the
total decimation, and the usable bandwidth a bit less than this. The centre
frequency of this bandwidth is set within Linrad.
You may perhaps wonder why, after the sample rate is reduced to 196kHz, we can achieve a bandwidth of as much as 190kHz, which seems to run counter to the Nyquist limit. This is because the IQ mixing process is a dual arrangement using quadrature local oscillator signals, which leads to two channels (I and Q), the combination of which provides twice the bandwidth that a single channel could provide at the same sample rate.
The choice of the three separate decimations is, within limits, up to you. There
is a very useful discussion by Leif SM5BSZ of the issues in choosing decimations
in the
Linrad Message Archive.
I am presently using values of 15, 10, and 4 for CIC2, CIC5 and RCF
respectively, giving a total decimation of 600, and therefore a sample rate of
111.111kHz and a bandwidth rather less than this.
I find that with a centre frequency of 144.105MHz (28.105MHz) these values give
me a comfortable span from 144.060MHz to 144.150MHz.
The A/D converter in the SDR-IQ has an output of 14 bits, which is increased to
23 bits during the decimation process.
However only 16 bits are
transferred to the USB port. You get to choose which 16 bits by selecting a
value for the RCF Shift in the Linrad setup.
The SDR-IQ has two "gain" controls set from within Linrad, one is an initial
-10dB attenuator (either in or out), and the other is a variable control set in
1dB increments. These are combined in Linrad in a single attenuation control
running from 0 to -36dB, with the -10dB attenuator in place below a selectable
level in the range -10 to -25dB. You can hear the microswitches for this
attenuator operate as you cross the threshold attenuation level.
It is not a good idea to use the Linrad control for setting the coarse gain of
the system. Rather you need to run the SDR-IQ with as much input as possible to
overcome the subsequent quantization noise, and to limit the signal
level into Linrad with the RCF Shift setting. A shift of 1 bit is a factor of 2
in voltage level (6dB). Leif SM5BSZ has indicated that for EME work a noise
baseline in Linrad of about 20 to 25dB is best (measured above the quantization
noise level). When power is removed from the preamp, the noise baseline should
fall by at least 15dB, which is why you almost certainly need the -10dB
attenuator OUT! I am currently using an RCF Shift of 5 bits.
The Linrad Setup Parameters can be somewhat daunting at first acquaintance.
You can simply accept all the default values and this will certainly get you
going, but it is unlikely to be optimal for any particular application.
I have put together some notes on each parameter to help others and especially
myself to understand them better, but you need to be aware that these are aimed
primarily at JT65B EME on 144MHz:
The VK2KU Notes on Linrad Setup.
If you just want to look at the set of parameter values I am currently using, here is a
list of them:
The VK2KU Parameter Values.
The Linrad screen can be configured pretty much any way you like by dragging
windows around and resizing them.
Here is a typical screenshot over the frequency range 144.050 to 144.160MHz:
Linrad071229.gif
There are a couple of EME signals around 144.120MHz, and a burst of SSB on
144.140MHz.
The on-screen controls can also be a bit daunting at first, so here is a version of
the above screenshot with all the controls labelled. I have inverted the colours
so that you can more easily read the labels:
LinradScreen1.pdf
Note that the above screens are only for a single channel system; also that my
system is not yet calibrated, so the Smart Blanker is not available; and finally
the Auto Spur Removal function is not activated.
There is actually much to say about the Linrad on-screen controls. If the mood takes me I may add a separate page on that subject too ...... sometime!