First real measurement results
Using a UMC202HD (24bit 192kHz USB audio input) connected to the IF port of an ADE-11X passive dbm using an isolation transformer to avoid ground loops.
A TCXCO on 10MHz is connected to the LO port. The TCXCO needs to warm for at least 5 minutes because the heater draws enough current to cause 50Hz residu pollution?
The module to measure is set to 50.010MHz?and connected to the RF port
Noise level when nothing connected to the RF port
The frequency is on top in kHz, 100Hz per tick
Noise level is average at -110, max input signal is 0dB so 110dB(?) dynamic range
When using a black module I measure
The signal top is at -6dB. Width at -80dB is about 100Hz
Many small spurs start at -85dB and disappear into the noise at 500Hz offset. They seem to be at a 100Hz grid. This could imply the spurs are caused by some residual 100Hz ripply of one of the analogue supplies I am using.
The noise floor also shows the 100Hz grid in the noise. Need to investigate.
The green module has similar spurs but they are less visible because of somewhat more noise in between
Beyond 500Hz offset there is no visible noise difference but the black module has all kind of spurs.
I modified one green module to isolate and decouple all supply lines with inductors.
There seems to be no difference in the close phase noise and beyond 500Hz offset again everything disappears in the noise floor
Notice the frequency offset, the TCXCO is helpful to calibrate the modules
Need to search for an audio FFT with more resolution, hope that will also reduce the noise floor.
Thanks for the pointers to the very low noise amplifiers but at -110dB, is there still much to gain????
