Does Air Spy mini work with EZRA?

[Conrad Cardano]

Hi

I have the SARA "scope-in-a-box" and wonder if upgrading to the Airspy mini is worth it?

Thanks for reading.

Conrad

[Stephen Arbogast]

I am using the attached ezCol version with an AirSpy, not  the mini.  Check with Ted Cline, he might have newer version.

Stephen

[Conrad Cardano]

Hi

Which model of the AirSpy are you using?

Do you notice a difference?

Thank you for reading!

Conrad

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/sara-list/e01e49c4-f5e7-43e7-aa73-c98fcae7f019n%40googlegroups.com.

[Stephen Arbogast]

This one...    https://airspy.com/airspy-r2/

It has;

12 bit ADC @ 20 MSPS (10.4 ENOB, 70dB SNR, 95dB SFDR)

0.5 ppm high precision, low phase noise clock

which are better than say the RTL-SDR.

I'm not sure the difference with the RTL-SDR is all that noticeable.

Stephen

[Robert Meade]

If not, adding support wouldn't be too difficult to write yourself. Alternatively, record baseband and process it into fft magnitude spectra yourself like I do. You could also just run it through DSPIRA in gnu radio. 

To view this discussion visit https://groups.google.com/d/msgid/sara-list/98a99b70-6786-444e-8d85-18fb15d6e3d8n%40googlegroups.com.

[Stephen Arbogast]

In my  humble opinion  there are  many  very  good  software packages  to collect  Hydrogen  Line data.  What  I  like  about  ezRA  is that it  is  a  suite  of  software programs that does  so much more  than   collect  raw  data  and save the  spectrum in a file.  Yes, it does  a lot  and takes time  to learn  but  worth the  effort.

Here is one of  my favorite programs  to  collect spectral  data and save to file,  SDRAngel.  SDRAngel  does not save spectral data in a file format  that  ezRA -  ezCon can  process.  I  have on my  todo list to use one of  Ted Cline's  data  file  conversion programs  to modify  to  convert  SDRAngel  saved file to a  file format that ezCon  can  work with.

This screen shot does not show  any  hydrogen line  but  you get   the idea..

[Stephen Arbogast]

Another   screen shot  from  Durango, Colorado ,  USA   using    AirSpy R2...   now  about    11:40   pm  local  time  no hydrogen  line but  notice the  spike    left  in the  spectrum...   This is  is  characteristic  for the  AirSpy R2 It  can be removed using    the many  features  of  ezRA  -  ezCon........

[Stephen Arbogast]

Note   the  bump   to  right  as my   Discovery   Dish   drifts ....    

The plot is corrected  for LSR....

[Andrew Thornett]

Someone shout at me! Following does come from AI. However, it does summarise the bandwidth advantage and options of the Airspy Mini over the RTL-SDR and equivalents with their maximum 2.4 Msps bandwidth. 

Andy

The Airspy Mini has a maximum IQ sample rate of 6 MSPS, giving a maximum instantaneous bandwidth of approximately 6 MHz.

However, there are a few subtleties:

Available sample rates: 10, 6 and 3 MSPS are supported by the hardware, but only 6 and 3 MSPS are officially supported on the Mini. 

Usable spectrum: At 6 MSPS, the outer edges of the spectrum suffer from filter roll-off and aliasing, so the truly usable bandwidth is closer to 4.8-5.0 MHz. SDR# typically displays about 80% of the nominal bandwidth, i.e. about 4.8 MHz. 

Unofficial 10 MSPS mode: Some users run the Mini at 10 MSPS (≈10 MHz bandwidth) using a debug setting in SDR#, but this introduces fixed spurious signals because the Mini lacks the clock circuitry of the Airspy R2. It is not recommended for serious work. 

For radio astronomy applications:

Hydrogen line (1420 MHz): 6 MHz is more than sufficient, since the HI line is only a few hundred kHz wide.

Continuum observations: The full 6 MHz bandwidth can improve sensitivity because noise decreases approximately as the square root of bandwidth.

Interferometry: The Mini lacks an external clock input, so coherent multi-receiver work is difficult compared with the Airspy R2.

In summary:

Mode Nominal bandwidth Practical usable bandwidth

3 MSPS 3 MHz ~2.4 MHz

6 MSPS 6 MHz ~4.8-5.0 MHz

10 MSPS (unofficial) 10 MHz Not recommended due to spurs

For hydrogen-line work, the Airspy Mini's 6 MHz bandwidth is generally ample, but for coherent interferometry the Airspy R2's external clock capability gives it a significant advantage.

Adc options on Airspy Mini vs Rtl-Sdr:

The main difference is not just the advertised ADC bit depth, but the effective dynamic range you actually get.

Feature Airspy Mini RTL-SDR (R820T/R820T2 based, e.g. Blog V3/V4)

ADC resolution 12-bit 8-bit

ADC sampling rate 20 MSPS ~28.8 MSPS

Effective number of bits (ENOB) ~10.4 bits ~7–8 bits

Theoretical ADC SNR ~74 dB ~50 dB

Claimed SFDR ~95 dB Typically much lower

Dynamic range Much better More limited

Oversampling/decimation Yes, up to ~16-bit equivalent at narrow bandwidths No comparable enhancement

Behaviour with strong nearby signals Usually excellent More prone to overload and images

The Airspy Mini uses a 12-bit ADC at 20 MSPS, with an effective resolution of about 10.4 bits, corresponding to roughly 70–74 dB SNR. It also uses oversampling and software decimation to improve narrowband performance further. 

By contrast, the RTL-SDR uses the 8-bit ADC built into the RTL2832U, giving around 48–50 dB of ADC dynamic range. 

What does this mean in practice?

Weak signals next to strong ones: Airspy Mini wins decisively.

Crowded bands (FM broadcast, pager transmitters, urban environments): Airspy Mini produces fewer spurious responses and images.

Hydrogen-line radio astronomy: The Airspy's greater dynamic range allows you to run higher gain without clipping and generally gives cleaner spectra.

Simple listening on quiet bands: The RTL-SDR often performs surprisingly well and offers excellent value.

For radio astronomy, the advantage is not usually that the Airspy hears dramatically weaker signals; rather, it is that it maintains linearity and a clean baseline in the presence of interference. That can make a noticeable difference when integrating spectra over long periods.

A useful rule of thumb is:

RTL-SDR: about 50 dB ADC dynamic range.

Airspy Mini: about 70–74 dB ADC dynamic range.

That extra 20 dB or more is often the difference between "usable" and "comfortable" when observing in RF-polluted environments.

Sent from Outlook for Android

---

From: 'Stephen Arbogast' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Sent: Tuesday, June 23, 2026 9:08:34 am
To: Society of Amateur Radio Astronomers <sara...@googlegroups.com>
Subject: Re: [SARA] Re: Does Air Spy mini work with EZRA?

To view this discussion visit https://groups.google.com/d/msgid/sara-list/e794e8f7-88df-484c-9113-15e8908b3351n%40googlegroups.com.

[Don Latham]

This from an ai indicates that somewhere someone has done their homework. believable.

 

------------
Don Latham
PO Box 404,
Frenchtown, MT, 59846
406-626-4304

 

---

From: 'Andrew Thornett' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>
To: sara-list <sara...@googlegroups.com>
Date: Tuesday, 23 June 2026 9:59 AM MDT

To view this discussion visit https://groups.google.com/d/msgid/sara-list/FRWP194MB27528393F33ED6D98041E211FFEE2%40FRWP194MB2752.EURP194.PROD.OUTLOOK.COM.

[Stephen Arbogast]

From ChatGPT.....

Observing the neutral hydrogen line (HI) at 1420.4058 MHz is an incredibly rewarding backyard project. It allows you to "see" through cosmic dust, map the spiral arms of the Milky Way, and measure galactic Doppler shifts. [1, 2, 3]

While both the RTL-SDR and AirSpy R2 are capable of detecting the hydrogen line, the AirSpy R2 will provide a significantly better experience. However, the SDR is actually the least critical piece of the puzzle—your choice of antenna and Low-Noise Amplifier (LNA) dictates whether you succeed. [1, 2, 3]

---

AirSpy R2 vs. RTL-SDR for Radio Astronomy

Because the hydrogen signal is a faint, broad cosmic hum rather than a sharp radio station transmission, device differences drastically change how you interact with the data:

• Shorter Integration Times: The hydrogen line is weak. To see it above the background static, software must average (integrate) thousands of samples over time. Thanks to its 12-bit ADC and incredibly stable clock, the AirSpy R2 pulls the signal out of the noise floor much faster. An RTL-SDR (8-bit) requires longer integration times (often 5 to 10 minutes per scan) to smooth out its higher quantization noise floor. [1, 2, 3, 4, 5]
• Seeing the Whole Doppler Curve: As the Milky Way rotates, different spiral arms move toward or away from us, shifting the 1420.4 MHz signal via the Doppler effect. This stretches the hydrogen line into a wide curve spanning 1 to 2 MHz. An RTL-SDR can only view a maximum of 2.4 MHz stably. If your frequency drifts or you want to see a wider background baseline to calibrate your signal properly, the RTL-SDR's narrow window feels very cramped. The AirSpy R2's 10 MHz view allows you to see the entire Doppler curve alongside plenty of "empty" background space for seamless baseline calibration. [1, 2, 4, 5]

---

The Essential Setup You Need

You cannot just plug an SDR into a standard antenna and see the galaxy. To successfully build a backyard radio telescope, you need this hardware chain: [1, 2]

[ Antenna ] ---> [ Filtered LNA ] ---> [ Coaxial Cable ] ---> [ SDR ] ---> [ PC / Pi ]

1. The Antenna (Choose One):
   • WiFi Parabolic Grid Dish: A 2.4 GHz WiFi grid dish (about 100 cm x 60 cm) is highly popular. Even though it is built for 2400 MHz, it works surprisingly well at 1420 MHz.
   • Horn Antenna: A DIY pyramidal horn antenna made out of cardboard lined with aluminum foil. It has excellent noise rejection but is physically bulky.
   • 1420 MHz Yagi: A commercial or home-built directional Yagi tuned specifically for 1.42 GHz. [1, 3, 4, 5]
2. The Filtered LNA (Mandatory):
   • Cosmic signals are incredibly weak, and cell phone towers (LTE/5G) will easily overload your SDR at 1420 MHz. You must place a high-gain LNA with a built-in 1420 MHz bandpass filter directly at the antenna.
   • Excellent off-the-shelf options include the Nooelec SAWbird+ H1 or the GPIO Labs Hydrogen Line LNA. These provide ~40dB of gain and filter out everything except the hydrogen line. [1, 2, 3, 4, 5]
3. The Software:
   • You need software capable of long data integration. On Windows, SDR# paired with the "IF Average" plugin is the easiest way to start.
   • On Linux/Raspberry Pi, specialized command-line tools like rtl-obs or PICTOR automate the averaging and produce clean data plots. [1, 2]

Summary Recommendation

If you already own an RTL-SDR v3 or v4, do not buy a new radio yet. Invest your money into a WiFi Grid Dish and a SAWbird+ H1 LNA first. If you find yourself limited by the narrow bandwidth or want to map the galaxy more accurately, upgrading to the AirSpy R2 will be a massive quality-of-life improvement later on. [, 2, 3]

Would you prefer to build a DIY antenna (like a horn or patch antenna) or purchase an off-the-shelf parabolic grid dish to get started? [1]

[Stephen Arbogast]

Andrew,

Respectfully, not  shouting , in  reply to your  statement... "Continuum observations: The full 6 MHz bandwidth can improve sensitivity because noise decreases approximately as the square root of bandwidth."  This not  true....  see   https://en.wikipedia.org/wiki/Noise_spectral_density

Noise power is meaningful only  in a stated bandwidth.

I use a  sampling  rate of 2.5 Meg  which  according to Nyquist  Sampling gives   me a bandwidth of  a 1/2  the  sampling  rate  which I  believe is  sufficient  for  Hydrogen  Line observations.

Experts, correct me if I am wrong.

Stephen

[Stephen Arbogast]

See https://en.wikipedia.org/wiki/Noise_power

[Andrew Thornett]

No problem Stephen, please correct the rubbish I spout, always! You and everyone else!!

Andy

Sent from Outlook for Android

---

From: 'Stephen Arbogast' via Society of Amateur Radio Astronomers <sara...@googlegroups.com>

Sent: Thursday, 25 June 2026 04:09:41

To view this discussion visit https://groups.google.com/d/msgid/sara-list/e3aea159-5d88-44ae-9119-378ef7e201c8n%40googlegroups.com.

[fasleitung3]

Andrew, Stephen,

The benefit of a larger bandwidth is described by the radiometer equation. The basic principle is that both the power received from the source and the noise power increase both linearly as the bandwidth is increased. However, the fluctuation of the noise power averages out more and more as you increase the bandwidth. This results in the signal to noise ratio being increased with the square root of the bandwidth.

There are several explations avaialbe in the internet, just look for "radiometer equation".

Then on the bandwith vs. sampling rate topic: What all SDRs do is complex sampling rather than real sampling. While the Nyquist criterion for real sampling is that the bandwidth is 1/2 of the sampling rate, for complex sampling it is a factor of 1. Just consider that for real sampling you get one number per sample whereas for complex sampling you get 2 numbers per sample.

Best regards,

Wolfgang

[Stephen Arbogast]

Ah.... my bad.  Thanks for clearing these up.

Stephen

[Stephen Arbogast]

All  my apologies,

I  did  some   followup  from  Wolfgang's  suggestions    and here  is  what I found....

https://casper.berkeley.edu/astrobaki/index.php/Radiometer_Equation

https://www.google.com/search?client=firefox-b-1-d&q=complex+++sampling

On Monday, June 29, 2026 at 10:03:01 AM UTC-6 fasleitung3 wrote:

[Robert Meade]

HI line observation should be a very low dynamic range activity. The only time that you should be seeing huge swings in input power is if manmade interference perturbs it that much. This coupled with low (< 0.1 dB) quantization noise is how I get away with 4-bit IQ for HI work without sensitivity loss (so long as the DAQ is filled to 4-bits based on gain setting).

I believe earlier, Andrew was referring to the time-bandwidth product for integration scaling with sqrt(BW) which means more BW can enable the same BT integration in less time proportional to sqrt(N). That's not to say that total noise input power is not still kTB.

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.

To view this discussion visit https://groups.google.com/d/msgid/sara-list/9f5d7364-c2b0-4797-8386-706d9b4076afn%40googlegroups.com.

[Stephen Arbogast]

Robert ,

I am not  sure   what   you   are  saying?   

[Robert Meade]

That was in response to Andrew's email starting with "Someone shout at me!". This thread is getting too long. I was saying that SDR dynamic range shouldn't matter for HI line RX unless interferening signals would saturate the receiver. And BW above 2 MHz will not often matter for HI either b/c the target signal is narrow band. It only really makes sense for continuum or for galaxy targets with large doppler spread not just shift (i.e. M33 or M31).

To view this discussion visit https://groups.google.com/d/msgid/sara-list/aca25b05-aedc-4318-bcb2-c9bcd91e562an%40googlegroups.com.

[Stephen Arbogast]

In my humble opinion  no  thread is  too  long.  What about   The   Central  Limit  Theorem  that  Wolfgang mentioned?  See      https://casper.berkeley.edu/astrobaki/index.php/Radiometer_Equation

[Stephen Arbogast]

We  still  don't understand   our  Universe.