Hello again Eric and other possible readers. I received my nanoVNA a couple of weeks ago and I have been doing many tests with it.
First I must say my unit came without shields and I added them, but there was no considerable improvement in the base noise level.
The second thing I am going to say is that I have been trying to check the operation of the CH1 input as a spectrum analyzer and I have been able see under what circumstances it is more or less usable.
First, the level measurements appear to be around 15 dB higher than the signal applied below 300 MHz and around 20 dB higher to over this frequency. I assume that it is because the nanoVNA accepts as 0dB the value applied in its CH1 in the calibration, which is that provided by CH0 and has those values ??(in the measurement frequency). Known that offset, is not a major problem.
Secondly, below 1MHz of span two peaks appear instead of one, as the image frequency also appears separated twice the value of the IF (i.e. 10KHz). In a "real" spectrum analyzer this image frequency is separated at several GHz and is rejected by the input LPF. Of the two peaks, the "real" is the higher frequency (the one on the right).
The third problem I have appreciated is that to be
recognized, the signal must fall exactly on any of the sampled frequencies because otherwise it is strongly or completely attenuated. Properly choosing the span this inconvenience is soslayable. This makes signals with a significant frequency modulation look very bad on the screen. This contributes to the low refresh rate, which also prevents an AM signal from being displayed correctly and as an extreme case a telegraphic signal.
As a summary of the above, to say that the signal of an RF generator without modulation is looked and measured "good" but "practical" signals (the carriers alone do not serve much in real life) look bad.
Greetings and forgiveness for the extension and possible-probable inaccuracies and bad english.
Javier Muriedas EA1AWY
