RSP Spectrum analyser software available shortly

[SteveAndrew]

Hi all

Reading through the forums I see that there seems to be some interest in dedicated spectrum analyser software for the RSP series of SDR radios. I have been working on such an application for the past few months and it is now starting to get a stage stage where I hope to be able to release a beta version. The Alpha version is reasonably well house-trained, but not quiet yet at a stage where it can be safely introduced to polite society

Everyone will likely have their own ideas about what such software should be able to do, and what the user interface should look like. Bearing that in mind, I've pasted some images here that should give an idea of where the project stands at this point in time. I'd welcome any ideas as to what such software should be able to do, and what sort of user interface would be best suited. I have to do a bit of a re-write so I may be able to incorporate any useful suggestions.

The controls layout is not cast in stone, it just evolved as I wrote various bits of code. I have avoided lots of drop-down menus, and tried to leave all controls instantly accessible for ease of use. Some of the control panel data is for debugging only and will be removed.

I am including a tracking generator that uses a DDS module controlled by an Arduino module. The code for this is already written for a vector network analyser project I was working on and just needs porting to this application. Calibration mode will sweep the full frequency range of the DDS and RSP in set steps, taking an amplitude measurement for each step. This data will be stored and later used to normalise data when the generator is used in tracking mode. The same principal should work for a wide band noise source. In time I would like to merge my VNA project with the spectrum analyser, resulting in a usefeful tool to have.

Currently the application includes (or will include)the following - in no particular order of importance:

* The ability to save and recall complete setups - saved in classic INI files that are easy to read and/or edit.
* Take screen snapshots
* A and B cursor functions allowing frequency and levels to be measured, along with delta functions
* Optional on-screen display of sweep parameters and cursor data
* The ability to hide the control panels, allowing a full screen spectrum to be viewed
* Spectrum, waterfall, or mixed display
* Tracking generator
* Display may be frozen if required - cursors still fully functional on a frozen display
* Saving of the input IQ signal to a file, and reloading the signal for later analysis.

If all goes well, I hope to release a Beta version available in 3-4 weeks time. All ideas and comments will be gratefully received - I look forward to getting your various thoughts on this project.

Regards

Steve A

PS - It seems there is a limit of three image attachments so I am unable to include screenshots produced by the application. I'll try and post them later this week on this thread.

Display with spectrum trace only

Capture1.PNG (60.2 KiB) Viewed 70899 times

Display with on-screen sweep data

Capture2.PNG (62.84 KiB) Viewed 70899 times

Capture4.PNG (68.55 KiB) Viewed 70899 times

[IQbert]

Nice! Would this be able to do modulation analysis as well? (i.e. QPSK, 16QAM, FM deviation, analog % modulation etc.)

[Tech_Support]

Hello Steve,
This is wonderful news. We believe that a lot of people would be very keen on this. Would you like us to help in testing of this?

Sincerely

SDRplay Tech_Support

[Paul]

Congratulations - looks fantastic!
Just an "inocent" query:
Could it be possible to reduce the sweep still further, down to say 250hz or even lower, perhaps permitting even finer resolution, down to single hz values?
Or is this a really ridiculous ask?

[vk7jj]

That's wonderful Steve, what a surprise. Rather a step up from where we were at on this thread!

All of us on the forum will be ecstatic you chose to support SDRPlay hardware. Out of curiosity I wonder how far below the displayed -130dBm before things get too noisey, and if/how you've managed to juggle bandwidth and decimation to optimise performance for a given sweep width.

The extra surprise is that you've got a tracking generator in hand too, an enviable combination of hardware and software skills.

I'm so old I learned to write on a slate and slide rules were the norm at university, it's hard not to laugh when I see your screen dumps knowing they're not pictures of the IFR or HP test set that was to be my first lottery win spend.

Arthur C Clarke got it right, any sufficiently advanced technology is indistinguishable from magic.

Cheers, Phil.

[Roger]

Steve,

This looks like a great project. Hope you will let us know if you need beta testers. I have some DSP experience and would be happy to help out with testing.

Roger

[someYguy]

Thank you Steve, this is very nice and would really unlock more of the potential uses for these receivers as test instruments. Looks very good and covers pretty much everything we here have been hoping for. My feature suggestions are simple (unless these things are already here and I missed them ). 1) Amplitude lines in 1/2/5/10db increments, selectable. 2) A simple readout of the current total frequency span. and 3) A readout of the Resolution Bandwidth in hz/khz.

Again, THANK YOU!!!

[SteveAndrew]

Hi everyone

I must say that the initial response has been very encouraging - thank you all for that.

Nice! Would this be able to do modulation analysis as well? (i.e. QPSK, 16QAM, FM deviation, analog % modulation etc.)

At this point, no provision has been made for any form of demodulation, or modulation analysis. I can see the advantages of this, and it is on the 'To do' list but will probably not be included in the initial release as that will require a fair amount of additional coding. The initial aim is to release a Beta version for testing, get any bugs sorted out, and then move on to the next stage.

Hello Steve,
This is wonderful news. We believe that a lot of people would be very keen on this. Would you like us to help in testing of this?
Tech_Support

That would be great. I only have an RSPII available so am not able to test with any other hardware. I may also have a few queries for you, I'll get in touch nearer the time.

Congratulations - looks fantastic!
Just an "inocent" query:
Could it be possible to reduce the sweep still further, down to say 250hz or even lower, perhaps permitting even finer resolution, down to single hz values?
Or is this a really ridiculous ask?

This should be possible with the existing software although I will probably have to use a low-IF option rather than using zero-IF. I'm getting some problems at sweep widths below 10KHz but I suspect it is a coding problem. One of more than a few problems I still have to address.

Out of curiosity I wonder how far below the displayed -130dBm before things get too noisy, and if/how you've managed to juggle bandwidth and decimation to optimise performance for a given sweep width...

I'm so old I learned to write on a slate and slide rules were the norm at university, it's hard not to laugh when I see your screen dumps knowing they're not pictures of the IFR or HP test set that was to be my first lottery win spend.
Cheers, Phil.

I suspect we are about the same vintage. I still have my original twelve-inch 'Guessing stick' and still use it on occasions. I sometimes think a slide rule is similar to quantum computing - an infinite number of solutions available at a single glance, a bit like the old-fashioned nomographs. Like us, antiquated, but still of some use

Noise-floor tests have not yet been done, but the performance should be the same as an RSP being used with SDRUno or SDRF#. Juggling bandwidth, decimation and the number of Fourier transform bins for optimal performance is something I am still playing with. I'm currently using the RSP in zero-IF mode for all sweep settings but I'll have to change to a low IF mode for narrow bandwidth sweeps due to problems with the DC spike. All good fun, and designed to drive strong men crazy.

With regard to the tracking generator: It is not yet tested with the analyser. As mentioned, I have used a similar idea for a VNA project I have been working on. A wideband sweep can take several seconds. In addition the RSP must change frequency at various points in the scan, perhaps dozens of times depending on scan settings. Changing the RF frequency on an RSP can take up to 15ms, and up to 100ms if a band boundary is crossed. This is something I still have to try. I'll post any updates and results on this thread.

I've included a few more images taken with the applications 'Screenshot' function. Looks like I need to include a function to allow an optional text note to be included in the snapshot image. Also looks as if the cursors need to be a bit easier to see -more for the ToDo list. Sigh...

Thanks again to everyone for an encouraging response and kind words

Best regards

Steve Andrew

144MHZ - FM, 1KHz modulation @ 5KHz deviation

Screenshot-6.JPG (47.1 KiB) Viewed 70797 times

144MHZ - FM, 1KHz modulation @ 5KHz deviation

Screenshot-5.JPG (46.82 KiB) Viewed 70797 times

35MHZ, 1KHz modulation @ 90%

Screenshot-1.JPG (33.39 KiB) Viewed 70797 times

[SteveAndrew]

My feature suggestions are simple (unless these things are already here and I missed them ). 1) Amplitude lines in 1/2/5/10db increments, selectable. 2) A simple readout of the current total frequency span. and 3) A readout of the Resolution Bandwidth in hz/khz.

Amplitude lines resolution is on the current ToDo list. Total frequency span is (should be) reported, I'll check that. Readout resolution is reported in Mhz/KHz per screen division on-screen. I do have the actual resolution available in terms of Hz per FFT bin, but it's display is currently disabled, but easy to re-instate if required.

Steve,
This looks like a great project. Hope you will let us know if you need beta testers. I have some DSP experience and would be happy to help out with testing.
Roger

Thanks for that Roger. I'll let everyone know when the Beta version is available.

[K9AQ]

THANK YOU. Christmas is coming early this year. I have been wishing for an SDR Play based spectrum analyzer software for a long time. I experiment with high power HF MOSFET amplifiers and I have tried several different "solutions" for a low cost spectrum analyzer. I started out with the "poor mans" spectrum analyzer which uses a swept varactor TV tuner and a dual channel scope for display. After that I picked up a very old HP SA at Dayton. It gave me the bandwidth that I was looking for to measure harmonic content, but the resolution wasn't very good for measuring IMD. I later switched to using the SDR play with HDSDR software. While it didn't give me the swept frequency range to measure the harmonic output, it did give me excellent narrow bandwidth for IMD measurements.

When I needed to check for harmonics, I would simply tune the SDR Play to a higher frequency, but it sure would have been nice to see them on a single, wide frequency display.

I tried some using some of the software products for the RTL dongle, but they didn't work anywhere as well as the SDR play.

I have also been using a noise generator with the SDR Play to observe the frequency response of my low pass filters.

I will be watching this forum for updates and want to be one of your early beta users. The addition of the tracking generator will make this a very useful piece of RF test equipment.

Thank you for your contribution to the SDR Play community.

Don Solberg
K9AQ