Spectrometer Source Block Settings

The default source block settings in the spectrometer_w_cal.grc GNU Radio program are for the Airspy R2 SDR. If a different SDR is used, changes in the source block, the samp_rate Variable block, and sometimes the freq Variable block may be needed. The settings for some common SDR’s used with the horn telescopes are described below.

Why freq sometimes has to change too. The program records a band samp_rate wide, centred on freq. The default is freq 1419 MHz with samp_rate 10e6, which records 1414 – 1424 MHz — the hydrogen line at 1420.4058 MHz sits comfortably inside. Narrow the sample rate without moving the centre and the band shrinks around 1419 MHz, and at some point the line falls off the end of it. Halve the sample rate to 2.4e6 and the band becomes 1417.8 – 1420.2 MHz: the line is 0.2 MHz outside it and the telescope records no hydrogen at all. That is why the RTL-SDR and Pluto settings below change freq as well, and the Airspy Mini and Lime do not.

A quick check: your band runs from freq - samp_rate/2 to freq + samp_rate/2, and 1420.4058 MHz needs to be inside it with room to spare at both ends.

Options:

The Airspy Mini
- Source block: the osmocom Source block is used; the “Device Arguments” are the same as for the Airspy R2. So no changes are needed in this block.
- samp_rate Variable block: This block is in the upper left corner of the canvas in the spectrometer_w_cal.grc program next to the Options block. Open this block by double-clicking it. Change the “Value” to “6e6” (which is 6 MHz).
- freq Variable block: leave it at 1419e6. The band becomes 1416 – 1422 MHz, which still holds the line.
RTL-SDR

Two changes are needed, not one. The samp_rate change alone moves the hydrogen line out of the recorded band — see the box above. Make both.
- Source block: Change the “Device Argument” to: rtl=0, bias=1, pack=0, as shown:
- samp_rate Variable block: This block is in the upper left corner of the canvas in the spectrometer_w_cal.grc program next to the Options block. Open this block by double-clicking it. Change the “Value” to “2.4e6” (which is 2.4 MHz).
- freq Variable block: next to samp_rate. Change the “Value” to “1420.5e6”. The band then runs 1419.3 – 1421.7 MHz, with the line at 1420.4058 MHz inside it.
- Two things to know about an RTL-SDR at this sample rate, neither of them a fault you can fix in the flowgraph:
  - The band is only ±230 km/s wide, so the faint wings of the line are clipped and there is no stretch of it guaranteed free of Galactic hydrogen to fit a baseline on. Maps are good for seeing where the Milky Way is; treat the intensity numbers as indicative.
  - Every SDR of this type puts a spurious tone at the exact centre of its band, which at this tuning is 0.1 MHz from the line. map_h1_hdf5_drift.py blanks a 120 kHz strip there, and because the strip is next to the line that costs roughly a quarter of the measured intensity. The comment on DC_MASK_HALFWIDTH_HZ in that script has the measured figures and the two ways to improve it.
Lime
- Source block: the Lime uses the LimeSDR Source (RX) block. Click on the osmocom block and hit Delete. Then in the search window on the tool bar at the top, type “LimeSDR”. Grab the LimeSDR Source (RX) and drag it onto the canvas where the osmocom block was. Then one-by-one connect the blue output of the LimeSDR Source (RX) block to the Stream to Vector block, the three Delay blocks, and the Complex to Real. The final connections should look like the following:
- Open the LimeSDR Source (RX) block (by double-clicking) and set the following:
  - On the “General” tab, set “RF frequency” to “freq” [without the quotes], and check that the “Sample rate” is “samp_rate” [without the quotes]. “Channel” should be on “A” [without the quotes].
  - The settings on the Channel A tab should be as shown:
- samp_rate Variable block: The Lime SDR can use a 10 MHz samp_rate; so no change is needed in this block. Leave freq at 1419e6 as well — the band stays 1414 – 1424 MHz, the same as the Airspy R2.
- Install gr-limesdr - software needed for the Lime SDR block to run in GNU Radio.
  - Open a terminal window.
  - Type and enter: sudo apt install gr-limesdr
- POWER TO THE LNA: The Lime SDR does not power the LNA. Therefore, it is necessary to provide +5 V dc external power to the LNA through a bias-T, which is connected to the LNA and Lime as shown:
  
  An SMA female to female connector/adapter will be needed for the connection from the bias-T to the LNA cable, as indicated in the diagram above.
ADALM-PLUTO
- The Adalm-Pluto SDR uses the PlutoSDRSource block that will need to be installed. Complete the steps outlined here to install this block on your computer.
- Source block: the Adalm-Pluto uses the PlutoSDRSource block. Click on the osmocom block and hit Delete. Then in the search window on the tool bar at the top, type “PlutoSDR”. Grab the PlutoSDRSource and drag it onto the canvas where the osmocom block was. Then one-by-one connect the blue output of the PlutoSDRSource block to the Stream to Vector block, the three Delay blocks, and the Complex to Real. The final connections should look like the following:
- Open the PlutoSDRSource block (by double-clicking) and set the following:
  - On the “General” tab, set the values as shown:
- The samp_rate and freq Variable blocks should be set to the values shown — samp_rate 3.5e6 and freq 1421e6. Both are needed: 3.5 MHz around the default 1419 MHz would stop at 1420.75 MHz and clip the line. At 1421 MHz the band is 1419.25 – 1422.75 MHz.
- POWER TO THE LNA: The Pluto SDR does not power the LNA. Therefore, it is necessary to provide +5 V dc external power to the LNA through a bias-T, which is connected to the LNA and Pluto SDR as shown:
  
  An SMA female to female connector/adapter will be needed for the connection from the bias-T to the LNA cable, as indicated in the diagram above.

Cable Hardware: A coaxial cable is needed to connect the LNA to the SDR. Typically a 10 ft length is adequate, but any length up to 25 ft should work fine.