Biasing mixer and VCO

I will be characterizing a mixer and a VCO (900MHz) and have some doubts about biasing. I need the following voltages :

- 3.3V and 5V for VDD purposes
- 1.2V and 2.56V for biasing, applied in MOS gates
- variable 0-5V to adjust VCO frequency, applied in MOS gates

My plan is to use a battery and get the desired voltages through low-noise low-dropout regulators, placed in a PCB.

I've read that RF chokes are used to pass DC to a circuit and suppress RF from the circuit. My questions are :

- Are chokes needed in this work frequency, or it's not critical ?
- Are chokes needed in all voltages described above ?
- Would I need an RF choke per DC output ?
- How will be a simple way to implement a choke ? inductor, inductor + shunt capacitor, inductor + resistor + shunt capacitor ?
- What about the component values (L,C) ?

thanks !

well more specifically, if I characterize a ring VCO, I suppose an RF choke is needed between DC power supply or battery and the MOS circuit (as figure shows) because in that node it may exist RF.
What about DC control voltage, also should have this kind of circuitry or not needed ?

thanks !

No, Vcontrol does not need a choke or similar circuit to isolate RF and DC because of very low Cdg capacitance.If you use a voltage source to drive tail current, this will be approx. short circuit for AC in small signal equivalent.

you mean capacitance between gate-and-source ? or Cdg ?
Gate will be fed by a DC variable source.

I mean Cdrain-gate..

In addition to, when circuit is well balanced, there won't be any RF signal on common driving point of differantial point so that this node "pseudo ground" for AC signals..

I see BigBoss, what about VDD node, it's said that RF could flow through it due to RF exists in drain node. Simulator shows just DC voltage in this node, maybe because power supplies are set as ideal ?

Of course all ideal DC supplies are short circuited for AC signals.That's why there won't be any AC components at there, it's obvious...

But in practice, every DC source has a resistance and inductance in according to supply quality but in general there will more-or-less be.To be avoid from the influences which will be perturbated from supplies and other sources, "decopuling" connections must be well maintened. To get a good result, in practice some decopuling capacitors parallel are used.Inductances are not preferable due to their self resonance frequencies.

I don't know which frequency you work on but be careful about also self resonance frequencies of the decoupling capacitors.

And finally, never use ideal components even in your simulations.Use ALWAYS equivalent components to see the realistic results..

the RF choke I am thinking to use in the outputs of my regulators for VDD purposes is :

http://www.minicircuits.com/ADCH-80.pdf

the following circuit converts from a single input to differential output. Input_1 and its bias are applied through a bias-tee, while in input_2 there is only a DC source to polarize that transistor. I assume that this last DC source doesn't need choke ?