meassuring receiver IM
Analog UHF receiver intermodulation meassurement according to ETSI 300 086 shall be performed at 25/50 KHz and 50/100 KHz from wanted channel.
When one of the interfearing generators is only 25 KHz from wanted channel, the side band noise from receiver VCO may allready decrease SINAD thereby reducing the IM rejection figure.
The result is a meassure of VCO noise rather than real intermodulation, even if sideband noise otherwise is acceptable.
The best way to measure IM is to use two signal generators and a spectrum analyzer. Then there is no effect of VCO noise on IM value. Using only power meters is not recommended.
In general, maybe, but the measuring proceedure in this case is determined by the standard. The two unwanted signals must be adjusted to a reduction of signal/noise ratio to 14dB.
You are right, there is a tough spec for phase noise on the signal generators on situations like this.
And this is about UHF. In HF (freq<50MHz) there are even smaller spacing between IM tones, down to 10kHz.
The only way to do this IM measurements (if you don't have tens of thousands of S to spend on top-notch signal gens), is to build your own low phase noise UHF signal generator using two crystals, with desired frequency spacing.
A Driscoll schematic would be a good option:
http://www.ko4bb.com/~bruce/XTALOSC.html
Sometime, for a single generated frequency, a simple 50 cent crystal oscillator can have much better phase noise than a very expensive synthesized signal generator.