Peter Hall, VK6HP
Looking at narrowband, high resolution spectra from SDRuno, HDSDR etc. is a good start and is useful for identifying gross effects. However, I find it's very useful to go the next step and actually run WSJT-X with the SDR since it gives you a sensitive display of artifacts, and lists any secondary decodes and their relative strengths. Apart from drive level issues (including audio harmonics), things such as hum show up readily on the display and, in cases where the modulation bandwidth is preserved, as spurious decodes. Hum, for example, is often offset from the wanted carrier by power-grid related frequencies.
As I mentioned in an earlier post, for quality checks I use the SDRPlay RSPdx and SDRuno running on an old Windows laptop, with the virtual CAT and audio connections made by VSPE and Virtual Audio Cable, respectively. Both connection programs are free for basic applications. This system runs simultaneously with whatever transmit system is under test. Most often the transmitter chain is connected to a dummy load but sometimes short, discreet on-air antenna tests produce interesting results - such as the 630 m passive intermodulation distortion I referred to earlier.
In all monitoring systems you do have to work hard at keeping the level to the monitoring receiver adequate, but not excessive. Terminating the receiver input is a good place to start experimenting.