Differential VCO, Short or High Resistance at control point?

[yorkyong]

The frequency of this vco is controlled by the Vctrl at the RED point. I'm not sure whether this node should be short to ground or high resistance for the signal. That is to say whether to add a big capacitor to short it to ground or connect to a high resistor to Vctrl.

 

[tfwee]

Hi YorkYong,

May I ask why to do you need to short this ground or add a high resistance to Control Signal? This VCNTRL is the point where you applied a DC voltage to reverse-biased the diode so that you can tune the frequency of the VCO. This point will be considered as gnd when you do a AC analaysis of the circuit.

 

[yorkyong]

Re: Differential VCO, Short or High Resistance at control po

Yes, in theory, it is short to gnd. But, in the measurement, the vctrl is given by a dc probe, which cannot be considered as an ideal "short" to ground. Then, my problem is the necessity of adding a decouple capacitor for the RED node

 

[rfsystem]

Re: Differential VCO, Short or High Resistance at control po

If the bias points of the NMOS are above the diode forward voltage you stop possible oscillation if the node is connected to ground. If connected via a reasonable high impedance to ground oscillation does not stop. If connected to supply the oscillation frequency is at maximum.

 

[tfwee]

Hi Yorkyong,

I have a suggestion for you. Since we know that VCNTRL voltage comes from the LPF of the PLL system, you may have to connect output of your LPF of your PLL system to this point and apply a control voltage to the input of the LPF for testing. This way you are able to test the VCO under the actual condition and able to plot the characteristic of the VCO.

 

[pi331133]

If the red point has a large decouple cap to ground, then in the PLL loop how the outpout of LPF to change the DC value of the vctrl, it will slow down the loop bandwidth, the control point is a DC for VCO but is time variable signal for PLL. I think in case of measure a standlone VCO you can put a decouple cap to reduce noise come from control point.

 

[mazelk]

Re: Differential VCO, Short or High Resistance at control po

In order to get a symetrical AC circuit, Vcont must be AC ground. So, you can divide VCO into halves, and carry on simple small signal analysis.