Osillator test problem(shaking in the spectrum analyze)

Dear all
An osillator is tested in a Spectrum analyzer,as shown in the picture,but the center frequency is shaking like that,the shaking bandwidth is about 2MHz ,why? is it the test problem or the circuit problem.
Thanks in advance!

I cannot see SA bandwidth and resolution settings. The spectrum shows a FM signal, so if you want a good oscillator, take care of its DC power supply. If your oscillator is a VCO, take care of the varactor DC, too.

There is a FM modulation coming from somewhere.Looks like it..
It may be supply,tuning voltage,load or something else..

Dear jiripolivka and BigBoss:
it is an VCO, how to take care of the DC supply, the DC supply is as shown below: is it a good supply? How to calculate the shaking frequency using this supply specification
? and the Kvco is about 1GHz/1V


Line regulation ≦ 0.01% + 3mV
Load regulation ≦ 0.01% + 5mV (<10A)
Load regulation ≦ 0.02% + 5mV (≧10A)


Ripple & Noise

≦ 1mVrms (5Hz ~ 1MHz)


Recovery Time

≦ 100μS ( 50% Load change, Minimum load 0.5A )


Output range

0 to rating voltage continuously adjustable



hello jiripolivka,please answer my question again.



hello BigBoss,please answer my question again.

Please take a good oscilloscope and look on DC power input to the main DC input and VCO tuning input of your oscillator. Any ripple over one mV can cause the output spectrum look as you show. Use batteries if your supply is poor. Add 100 uF plus 0.1 uF capacitors across DC terminals.

what are u using for a VCO Tune line power supply? If you are taking a free running measurement, I prefer a battery and a potentiometer with a >10 uF cap as the tune voltmeter.

also, u might want to use an inside/outside DC block for the connection to the spectrum analyzer. AC ground loops are troublesome

Now that's a term I have never heard before ! had to google it

I assume the 2 grounds shown in the schematic ( lower right) refer to the D.U.T. GND and the test equip GND and never the twaine shall meet

Dave

the two grounds shown on that web page refer to the conductor being the actual outside of the coaxial line, which grounds itself out as you attach sma connectors to it.