Page 1 of 1

Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Thu Dec 15, 2016 11:26 pm
by someYguy
Hi all,

I've been using my RSP1 along with an inexpensive noise source (from eBay) as a Scalar Network Analyzer. Similar to a conventional spectrum analyzer/tracking generator combination.

With this, I can use the RSP1 to analyze and characterize RF filters, mixers, etc. and also to tune repeater duplexer units. I also have a 10db Directional coupler which I can connect between the noise source and the RSP1, which gives the RSP1 the ability to read return loss which allows for pretty precise antenna resonance & bandwidth measurements.

I thought that it might be good to report my findings with this here so that others may benefit from my experiences.

Results are actually quite good. For comparison, I also have an Instek model GSP-810 1ghz spectrum analyzer in the RF lab here. It has a tracking generator also.

The image below shows the RSP1/Noise Source/Directional Coupler combo sweeping a homebrew 80m double-bazooka antenna, using SDRSharp for display. Other SDR programs such as SDRuno, HDSDR, etc. can be used also with good results. As can be seen, the antenna is most resonant at 3890 khz (center freq in the display):


Next is sweeping a 500khz low-pass filter. Sharp decrease beginning at about 575khz, just as it should be:


So my conclusions are that while the RSP units in combo with noise source are not yet a true substitute for a dedicated spectrum analyzer/tracking generator, they are certainly accurate enough for test purposes and most useful for this. This is a great way to create a useful scalar network analyzer without spending lots of money. But what is really needed here is a good, dedicated spectrum analyzer program for use with the RSP. Such a program would have analyzer features such as markers, separate frequency readouts, reference (top line) level setting, and provisions for power level reading calibration with offset setting for use with external attenuators. I am a software developer and would write the program myself for us all to use but my specialties are in other areas so I do not have the necessary skill.

So get a good noise source to use with your RSP1/RSP2, and experiment! :)

Re: Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Fri Dec 16, 2016 11:14 pm
by g,3vds
Thanks for the interesting post. Please let us all have more details when you have time as I would like to put together a similar setup. Best 73 Roy

Re: Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Sat Dec 17, 2016 9:44 pm
by someYguy
Hi Roy, here is some additional detail:

The noise source which I use is this one: I run it at 7-8 volts even though it is specified at 12 volts. At 7-8 vdc, it produces a very nice, flat noise signal.

The directional coupler is only needed for single port measurements such as return loss from an antenna. Mine is pretty old but it is similar to this one: ... Swqu9U2rgN

So to measure frequency response of a filter, say for example a 146mhz banpass filter, you would first set the center frequency of the RSP1/2 to 146mhz. Then connect the output of the noise source to the input of the filter. Filter output connects to RSP1/2 antenna input.

Set the Gain Reduction of the RSP manually - i.e turn AGC off. Set the gain for a good display of the waveform, so that the top (pass) and bandwidth of the filter can clearly be seen without overloading the ADC within the RSP.

For antenna measurements, simply connect output from the noise source to the input of the directional coupler, output from the directional coupler to the RSP antenna input. The antenna to be tested then connects to the third (Coupled Port) of the directional coupler.

Again, turn off the RSP AGC and control the gain reduction manually. Set the RSP center frequency to somewhere within the band where the antenna should be resonant. Set the gain so as to produce a baseline of about -70dbm. Carefully tune around and you should see a pronounced dip at the resonance point of the antenna.

I usually set the bandwidth of the RSP to 600khz directly in the RSP Device Control Panel. Set to LOW IF.

You can convert return loss (dbm) to VSWR easily, using a chart or calculator for the purpose. It is possible to get very accurate readings if one positions the cursor carefully and then read the dbm values from the display.

At present, I've found SDRSharp to be best of the available programs for this type of work. No dedicated spectrum analyzer program yet exists, but the display of SDRSharp comes closest at present. HDSDR is good also, and even allows the VBR (Video Bandwidth Resolution) to be set.

SDRuno is very fine software and is top choice for receiving, but I found it to be currently unsuitable for test & measurement purposes because it does not allow direct access to the RSP Device Control Panel, and also zooming is limited and so it is hard to set up a detailed waveform. But I have a feeling that the awesome fellows at SDRPlay will add some spec analyzer specific features to their software at some point :).

I have not tested SDR Console for this so can't comment on that but its likely similar to SDRSharp for this purpose.

Hope this helps, any further questions please let me know.

Re: Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Sun Dec 18, 2016 3:12 am
by someYguy
Here are a couple of more images showing uses for this:

Sweeping a Collins 455khz Mechanical Filter. The filter response curve just below the center frequency shows that this is an LSB filter, and is approx. 3.5khz wide. The spike to the left is generated within the RSP1 and should be ignored.


Sweeping another Collins filter, this time USB and approx. 3.1khz wide. I'm not sure where the spikes on top of the response curve are coming from but they do not hinder measurements. [Edit: This filter has a lot of ripple in its passband, thus the peaks and valleys.]


So this equipment - SDRPlay RSP1 or RSP2, along with a good noise source and appropriate software can indeed be used to test these types of IF filters. The question was whether or not this combo of gear would provide sufficient resolution, narrow bandwidth, and clean display for this, and it clearly does :).

Re: Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Tue Dec 20, 2016 1:13 pm
by g,3vds
Thanks very much for the additional information. I'm all fired up to try it!
I've ordered a boxed noise source from Shenzen which I hope is OK. Will try it at 8 volts. Expecting delivery next month.
The coupler you mention seems to be OK on 2m and up but as I'm mainly HF I'm looking for an Elecraft CP1
I have an RSP2 so only used SDRuno so far.
I may have to buy an RSP1 for this project
I would have killed for something like this when attempting to tune a 70 CMS duplexer many years ago!!!

PS. I've now ordered the coupler you pointed out as I can use it. Its coming from Hawaii which is about as far from me as you can get without going off planet but the price is excellent.
I'll still search for a CP-1. Building small fiddly thingies is a bit hard for me these days. Thanks again for your help.

Re: Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Tue Dec 20, 2016 8:42 pm
by someYguy
Very good. Let me know how it all works when you get everything together :). My directional coupler is an old Merrimac type UR-1-10. Although only rated down to 10 mhz, I've had no trouble and no significant loss with using it down to about 1 mhz. I did forget to mention that I mounted my noise source inside a metal box also but yours is already enclosed so you're good to go. Running at 7-8vdc seems to make it run much cooler also.

Regarding the RSP, I think that our SDRPlay folks are very close to releasing an IO dll for the RSP2 for use with other software so should work soon with SDRSharp.


Yup, I agree that this would have been so cool back in the days of lots of duplexer tuning :). I have an old Motorola 1500 series UHF duplexer here from back when I ran a repeater, and I did try cavity tuning of it with the RSP1 etc. and only limitation which I found was the maximum viewable bandwidth of the RSP at 8 mhz.

I could see the pass & reject and it was certainly usable but I couldn't really get the screen centered to my liking. Should work much better for 2m duplexers though due to the much smaller frequency separation :).

- Darrell

Re: Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Sun Dec 25, 2016 11:33 am
by DF1KZ
I use this Noise Source, a RSP1 and a Mini Circuits ZEDC-15-2B coupler. The noise source is really good, it becomes even better if you remove the third amplifier and insert a few dB attenuation between source and coupler. The reason for the high temperature of the source are the resistors used with the amplifiers.
I made some measurements of this Noise Source using a professional R&S spectrum analyser and EMC receiver. I also have values from 10kHz to 3GHz but they are numerical (txt). I have to convert it before I upload it.
If you use the source with all three amplifiers without any attenuation take care of your RSP1! It is overloaded by the wide signal rather fast.

Scan with 10kHz bandwidth:
EMI-Scan_ohne-Transducer.jpg (188.92 KiB) Viewed 22191 times
Scan with 120kHz bandwidth:
EMI-Scan_ohne-Transducer_120kHz.jpg (186.16 KiB) Viewed 22191 times

Re: Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Sun Dec 25, 2016 12:04 pm
by g,3vds
There seems to be a software called "Spectrum Lab" which may be useful. Merry Christmas. Roy

Re: Using the RSP1 with a noise source aka Scalar Network Analyzer

Posted: Wed Dec 28, 2016 5:56 am
by someYguy
I thought about using Spectrum Lab for display but I couldn't get it to read out the direct frequency, and the interface seemed to be pretty crowded. Seems that of the available software, SDRSharp works best for this because it has a grid very much like a standard spectrum analyzer, and the span can be easily set. For example, when sweeping the Collins 455khz filters in the photos above, I set the span to approx. 20khz (via the zoom slider) by using the frequency scale just below the spectrum display. You can do same with HDSDR, but it does not have vertical grid lines so harder to reference. SDRuno isn't very useful for this application at all because it zooms only in predefined steps.....