¿ªÔÆÌåÓý

Re: RF source evaluation


 

Yes I continue slowly.? Progress is glacial compared to Erik :)
To be clear to others, the simpleSA version runs on a different processor (ESP32) and the initial motivation was to see if I could make an analyser that could be used and controlled via the web from a standard internet browser using the WiFi features of the ESP32.? It proved to be possible though it took a few months to work out why the wifi stopped at certain points in a sweep - turned out the fifth harmonic of the LO was blocking the WiFi receiver!

I also designed a set of PCB, using KiCAD.? Of course there were a few mistakes but easily fixed.? My boards are a little more cramped than Gyula's as I wanted to keep the two boards within the 100x100 for low cost PCB from JLCPCB, and I wanted all the connectors on one edge.? ?I have some of the 30MHz Si4432 clock getting in unfortunately.? The 350MHz low pass filter also needs some optimisation.? As another learning exercise I also designed a 3D printed case with cover/tilt stand using TinkerCAD?which a club member kindly printed for me.

In this build I used different frequency SAW filters and different IF for the second pair of Si4432.? With no cable connecting the TG output to the SA input you cannot tell if the TG is on or off even with RBW @ 2.6kHz.? An additional RF switch is also included to change from the low frequency range input to the high range input.

Photo below with the tracking generator being used to measure a QRP-labs bandpass filter for 20m.

I am thinking of designing another PCB set with a common clock source for all four Si4432, and where the Si4432 would be soldered directly to the PCB rather than use the ready made modules.? Not sure if buffering is needed for the clock yet and having trouble finding a low cost 30MHz TCXO - any sources anyone?? Given the limited frequency resolution maybe a TCXO is overkill.
Next plan for software development is to enable the second set of Si4432 to be used as a second SA channel, then it could be used either a two channel SA, SA with tracking generator, SA with signal generator or two channel signal generator.? It will happen when I have the need for it!

Thinking of the TinySA itself and tracking generator, with two TinySA I think a very good result could be obtained if rather than using the LO from one to synchronise the second the PC software was used to offset the IF of the two devices, and the PC software kicked off each sweep in both devices to keep the synchronisation between the signal generator frequency and the analyser sweep.? I think it should be possible to keep the two devices in good enough synch throughout one sweep just by careful timing calculation (no of samples, delay duee to filter settling, some overhead and a margin for some timing error.? Then no additional hardware such as attenuators and amplifiers would be needed.

Worth a try anyway.? Gyula's version is getting very good results with the same SAW filters in each pair and an offset in the IF but still within the passband of the SAW filters.


73

Dave M0WID


Join [email protected] to automatically receive all group messages.