Many previous literatures considered that the operating frequency of an impulse governed oscillator （IGO） was difficult to be extended to the microwave band. One of the reasons put forward was that the width of the synchronizing pulse must be narrower than 1/5 to 1/3 of the period of the VCO

which was relatively difficult to achieve in the microwave band. Some recent literatures raised an objection to this

but still considered the narrower the synchronizing pulse for a microwave IGO in the case of non-ideal sample-and-hold

the better. This paper proves that neither of these conclusions is exactly true. In addition

previous literatures also doubted the loop stability of IGO

and held that its phase noise might be greater than that of a common phase lock loop. This paper discusses the problems concerning phase noise and loop stability. Theoretical analysis and an IGO experiment at 3GHz show that a correctly designed microwave IGO has low phase noise

better stability and a simpler

compact strip-line structure. Therefore further advances of the IGO may be expected.