VARACTOR DIODES --AN ALTERNATIVE TO VARIABLE CAPACITORS
BY WA7SSA

On the topic of Varactors, they are virtually all 'good' BUT they are not the same as a variable capacitor.

You cannot substitute them directly and achieve equivalent performance, but you can nearly always incorporate a varactor into a circuit and achieve good performance if you take all the effects into account.

These different effects include:
- temp coefficient
- junction conductance
- non-linearities with respect to instantaneous voltage, that is the sum of both the tuning voltage and the RF voltage at any time but especially the peak - V.tune impeadance

Others have already addressed the different temperature coefficient, so I'll leave well enough alone.

All solid state juntions suffer from one significant difference from a 'real capacitor'. IF the RF voltage is large enough the diode junction will conduct during that portion of the cycle that exceeds the diode 'on' voltage. This introduction of non-linear effects includes changing the RF impeadance of the circuit, changing bias if the circuit is designed wrong, and changing the instantaneous capacitance and hence the tuning. This is to be avoided at all costs. sometimes you will see two varactor diodes in series with opposite polarity. the tune voltage will be feed to the junction of the cathodes. this is to eliminate conduction in the circuit. (one will always be non-conducting..) Preferably, the designer accounts for the peak voltages and ensures that the diode is not going to go into conduction.

In the same vein. A varactor diode (ALL!!!) exhibit non-linear relationships between V and C. This is still true with respect to the instantaneous voltage we talked about above. The C you really get is based on the sum of both the tuning voltage and the RF voltage at any instant of time. This results in spurious modulation, oscillators which will start and run for a few cycles, stop for some small portion of the RF cycle, and then restart (this is the voice of experience, and a REAL pain to find at Ghz freq's!!), and many other nasty behavior. Good designs will ensure that the instanteous variation does not substantatially deviate from the tuning point.

Although not a function of the Varactor, another potential problem introduced by the use of a varactor is the impeadance of the tuning voltage connection. this should be very high to ensure that no extraneous effects (circuit loading, stray rf/audio coupling, etc) will be introduced to the VFO. VFO's can make great mixers, product detectors, etc under the right conditions conditions. This line (V.tune) is generally a 100k resistor to each tuned point for this reason.

Niel

Niel Skousen: Sr.Eng, SCIENTECH.SPG/CFG/NUSI
208.525.3742, 524.9229 FAX 529.4721 Idaho Falls ID
nskousen@scientech.com WA7SSA QRP-L.119

Back to Oscillators Page