Linearity problem in designing VCO

Hi everybody,

I have a non-commercial HBT (got the ADS model from specific foundry) for designing 6.67GHz VCO. Everything works fine when I used lumped resonator (4 element resonator circuit LC in parallel and 2 C's in series, like pi shape)

The moment I transform this lumped model of resonator into distributed model, the linearity error (tuning error --> derivative of frequency voltage slope) increases.

In distributed model, I m using interdigital capacitor for C's in series, the OC stub for shunt C and non-ideal transmission line for shunt L.

Does anybody have an idea where the error is?

I tried different model of distributed resonator but all in vain!

--X--

Over what % bandwidth are you trying to tune the VCO?

The tuning BW is about 320 MHz with fo=6.67GHz, while varactor voltage is varied from 1 to 10V.

Hello somf0872,

Your HB model with lumped element(spacial case of distributed elements) work maybe problem is that you use infinite Q for lumped elements ? If it is case you can try with real Q you can expect from distribudet elements around 20-80 max. Real Q have to give quite different linearity frequency to tuning voltage (different Q giving different slope).

GL XTASA