The Solar Radio Spectroscope consists of a radio receiver and logging system set up at the observatory, with data accessed from the house through a WiFi bridge for reduction and presentation.
The receiver scans continuously over a broad range of the radio spectrum. Data represents radio signal strength (dB) which is plotted over time on a spectrograph. Signal peaks are either transient, associated natural sources (eg the Sun) or continuous, associated with artificial sources (CB, radio, TV, mobile phones etc). Note: Astronomers refer to the artificial sources as "interference" :-) Spectrographic output provides a view of the data which allows identification and interpretation of solar events.
1. Reception
Step 1 was determining the frequency of interest that would likely capture events on the scope. This involved a bit of research on solar physics and investigation of receiver capability, resulting in a target range of 25-1800MHz.
I selected a hybrid Discone antenna to provide a balance of frequency range, gain, size and cost. Other antenna types are typically designed (or limited) to useful gain over much narrower frequency ranges, or prohibitive for a backyard project. I'm happy for a communications engineer to show me a better way here.
 |
Full band discone scanner antenna
Mounting Hardware
|
The next stage involves RF amplification over the desired frequency range. The simplest solution for me was a TV signal amp which provides a pretty good range of 40-862MHz. There's obviously scope for improvement here, with Low Noise Amplifiers being capable of say 40dB gain at 0.2dB noise. I plan to run the scope for a while before determining if this is a weak point in this application.
.jpg) |
TV Signal Amplifier
|
Finally the receiver. This is the good bit - we can now simply and cheaply receive all sorts of RF over a huge range with a TV dongle. The receiver package is completed with a Software Defined Radio (SDR) program, which provides all of the control functionality via a software package.
I've selected a tuner with the Realtek chip below to provide the best frequency coverage. Other chips can be more targeted to narrower ranges around digital TV etc. I am using RTL-SDR Scanner as the SDR driver program, which provides a user interface for direct observation, but also a good command line interface allowing easy automation. Also note the SDR program requires a specific driver to access the TV dongle.
In the past, radio enthusiasts tackled this problem with similar mindset - take apart an old TV, pull out the tuner and various amplifiers and filters and solder up a new receiver.
 |
USB TV Dongle
|
2. Logging
RTL-SDR Scanner can be run from the command line with parameters for frequency range of each "sweep", the software defined gain to apply, the dwell time at each frequency, and the bins used for Fourier transformation of the data set. Higher dwell times can improve the signal to noise ratio, but increase the duration of the sweep. The sampled bandwidth is 1.9kHz, which appears to be fixed (broader sampling bandwidth would be useful in this application).
RTL-SDR Scanner runs every 5 minutes, executed via a DOS batch file, fired off by Windows Task Scheduler.
PARAMETERS
|
OUTPUTS
|
3. Reduction & Output
The reduction process takes the raw scan data produced at 512 bits per MHz and averages this within a user specified bin size. Currently I'm using 0.25MHz as a bin size, resulting in 4 data bits per MHz, or 252 bits over the initial scan range of 25-88MHz. This seems manageable but can be readily changed to provide the appropriate resolution.
I've chosen Excel for this component, mainly as a familiar front end for data, but also VBA is a comfortable programming environment for me.
INPUTS
|
PROCESSING
|
OUTPUTS
|
I initially set up this module to execute the scanning process directly, so you could accumulate data and see it progressively. However I changed this to the current 2-stage process to make it a bit more robust. I'd rather have a number of short processes running discretely than have Excel running 24 hours at a time. This would still require a triggering mechanism in any case.
Planned Improvements
There are a few changes I'll work on over time to streamline the process and improve the output.
Efficiency
- DONE: Bottleneck is remote access to data files over the wireless bridge, so will look at remote desktop to execute reduction process (upstream of the bottleneck). Resolved with daily reduction done by script run as scheduled task.
- DONE: Seems OK so far, ironing out once-off issues from a few bugs, auto reboots and anti-virus interference. Added status logging and email notification for outages.
- Resolution may be limited by sweep interval at 5 minutes, so may need to improve this. However, time to complete one scan is the limiting factor.
- Look at upgrading RF amplifier to a purpose built LNA with broader bandpass and lower noise if required.
- Possibly changing balance of dwell time and (if possible) sampled band width to speed up scan time, allowing greater resolution with shorter sweep times.
- DONE: Option for web services to drive process and publish output. Daily reduction script also publishes spectrograph to this site.
No comments:
Post a Comment