The tinySA can be used as a signal generator and in low output mode there are some options for modulation.
Using the AM modulation to create a 10MHz signal?



with another tinySA in zero span mode it is possible to look at the modulation as one full scan of 290 points only takes 232 milliseconds



and indeed the modulation is a sawtooth with a frequency somewhere below 1kHz
As you can see the output level still needs some calibration.

I hope you remember us little people when you become rich and famous :)
Can you share when you think the commercial version will be available?

Erik,

I would love to test out this TinySA and have been with the project since December of 2019 and have a version built up like you did on protoboards. I have A LOT of test equipment at hand (mostly HP and Agilent) for checkout and am knowledgeable of the design. Please do consider me for this effort - thanks,

73 de Marty, KN0CK?

Greetings Erik,

? ? ? ?A common problem for radio amateurs is the proliferation of modern day RFI sources that cause significant interference particularly to the HF bands. IMHO a portable low cost spectrum analyzer that includes some specific features intended to support tracking down noise sources would be of significant interest for many hams.

? ? ? ? Often noise sources are related to the power line frequency. It has been well documented elsewhere that because there are often multiple A/C power related RFI sources around the home and within the local hydro grid that an ability to record the time domain signature of the most offending noise is helpful for identifying the most significant cause. Once a problem is located within one's home it can usually be corrected. If a power grid problem is pinpointed it has been experienced that usually the utility will repair it expediently. Spectrum analyzers with time domain signature waveform comparison capability are now used by many utility companies for locating RFI sources.

? ? ? I found it interesting that that the tinyVNA firmware already offers a zero span mode with time domain display capability. I wonder if the time domain display could be designed to lock on to a hydro buzz type of noise and store a few cycles? If this were implemented the offending RFI?waveform signature could be compared to noise sources found in the field to help identify the source. Possibly a split screen display with the stored waveform shown within the top half and real time waveform in the bottom half would offer a convenient means for the user to compare signatures.

? ? ? ?I am interested in evaluating a tinyVNA for the purpose of tracking down RFI sources.

Best Regards,
Tom, VA7TA

Erik,
? ?Once you are ready to release the full list of console commands for the tinySA, I would like to see if I can translate? the python nanovna.py class library to the tinySA (tinysa.py?).? No guarantees but I believe my Python skills have progressed to the point since using the NanoVNA that I should be able to do it.? I might have to ask hugen to clarify a few points if he is available.

- Herb

Hello, i can help you vs test tinySA?

? My main job is the development, repair, maintenance of aviation communications systems. Therefore, I have the opportunity to test it on various equipment (compare with other spectrum analyzers, check on various signal generators and on real radio stations).
Also can help vs software (it my hobby)

Hello, i can help you test tinySA

My main work is development, repair, support of aviation communication systems (i not programmer, it my hobby, but i can help vs software)

It is possible to check the required characteristics on various equipment.

1. Very, happy to see selectable front end attenuation.? One of the best ways to determine if a signal is real or not is to add rf attenuation.? If the signal stays at the same level you know it is real but if it drops dramatically it was not.? I usually have to hunt down an external attenuator to perform this check on a low cost spectrum analyzer.

That is true for intermodulation products. For mirrors (due to limitations in shielding and filters) there is a built in function to kill mirrors

2. Any possibility of adding a menu for selectable display units (i.e. dBm, dBuV, uV)?

Can you point me to where I can find info their purpose and how to calculate?

3. Average detector not an option but video averaging can be accomplished by setting video bandwidth lower than the resolution bandwidth.

The video bandwidth in the tinySA is automatic and results from the span and the 290 display points. There is an option for averaging over multiple scans,??

4. I didn't see a sweep-time menu option on the wiki page to allow the user to manually adjust the sweep rate.? Sometimes necessary to capture some signals.

There is an option to set the sample time per display point. This is primarily meant for balancing sweep time and loading of the narrow RBW filters.?

5. Can a external generator be selected for more accurate level calibration instead of the built in calibrator?? Not everyone has access to a calibrated rf generator, but it would be nice to allow those that have one to use it for extra accuracy.

Any signal under marker 1 can be used for calibration. See??under "actual power"

6. Trace math - thumbs up!

7. Zero span - does x-axis change to time domain units?

Yes, but no settable scanning time yet, just full speed

8. Are the live and stored traces available for usb readout using the "data 0" and "data 1"? NanoVNA console commands?

Yes, one stored trace, one live (calculated) trace and one trace containing the last unprocessed (raw) scan

Does mono have a decent graph object?

On Mon, Apr 27, 2020 at 11:50 PM, <erik@...> wrote:

Yes, ref level and attenuation are linked as you would expect so the noise floor comes up with increasing attenuation but signals stay on same level.

=====================================================================
Erik,

1. Very, happy to see selectable front end attenuation.? One of the best ways to determine if a signal is real or not is to add rf attenuation.? If the signal stays at the same level you know it is real but if it drops dramatically it was not.? I usually have to hunt down an external attenuator to perform this check on a low cost spectrum analyzer.

2. Any possibility of adding a menu for selectable display units (i.e. dBm, dBuV, uV)?

3. Average detector not an option but video averaging can be accomplished by setting video bandwidth lower than the resolution bandwidth.

4. I didn't see a sweep-time menu option on the wiki page to allow the user to manually adjust the sweep rate.? Sometimes necessary to capture some signals.

5. Can a external generator be selected for more accurate level calibration instead of the built in calibrator?? Not everyone has access to a calibrated rf generator, but it would be nice to allow those that have one to use it for extra accuracy.

6. Trace math - thumbs up!

7. Zero span - does x-axis change to time domain units?

8. Are the live and stored traces available for usb readout using the "data 0" and "data 1"? NanoVNA console commands?

I think you have struck a good balance between features and specifications.? Wish you and hugen much success.

- Herb

On Mon, Apr 27, 2020 at 11:39 PM, <erik@...> wrote:

The SA outputs its first LO from the high connector.
The tracking generator mixes the LO with a signal at the first IF and after an LPF you get the tracking signal

==================================================
That's pretty nifty.? Sort of like the old HP8445/HP8568 set-up.? The synchronous tracking should allow for additional depth when sweeping filters. I purchased a rf explorer and rf explorer? generator a couple of years back and the tracking feature could only be evoked by computer control, no stand-alone mode.? I can already envision a future tinysa model which includes the tracking generator inside the case rather than as a separate unit.

- Herb

- Herb

Kurt,
Can you confirm I'm doing the IIP3 measurement as it should be done?

I'm not able to create such an application but as the tinySA console commands are very similar to the nanoVNA a derivative of nanoVNA-saver becoming tinySA-saver should be not too much work.

Another measurement option is measuring phase noise. This of course is limited by the internal phase noise of the tinySA which is not world class.
In phase noise measurement mode a phase noise delta marker is set at the requested offset frequency and a delta noise level is displayed in dB/Hz to eliminate the impact of the selected RBW.
Below is the result of one tinySA measuring the phase noise of the high output at 250MHz of another tinySA. This level of phase noise defines the lower limit of effective phase noise measurements with the tinySA

开云体育

This shows the measurement option to measure OIP3 and it is used to measure the internal IIP3 at 0dB attenuation.
Two tones (6MHz and 7MHz) from fairly clean signal generators are used.
Bot the left and right IIP3 are shown.
Investigations are ongoing to further improve this and an IIP3 level of +25dBm could be possible.

See the wiki for the details


Yes, ref level and attenuation are linked as you would expect so the noise floor comes up with increasing attenuation but signals stay on same level.
Both ref level and attenuation can be set to automatic


Next to store and subtract (normalize) there is also waterfall

The SA outputs its first LO from the high connector.
The tracking generator mixes the LO with a signal at the first IF and after an LPF you get the tracking signal