PLL active loop filter for relatively high minimum tuning voltage
I have designed and built a VCO with a 12...25V tuning range and I would like to design an active loop filter for an integer PLL IC to control the VCO frequency. I would need some advice, help, or just verification that my train of thought is not wrong.
The charge pump output can produce a voltage between 0 and 5V (more likely 0.5 and 4.5V), so I would need an active loop filter with positive supply voltage say 28...32V and a grounded negative supply as the charge pump minimum output voltage is 0V as well. The topology would be a second-degree passive filter (parallel C1 and R2C2), then a non-inverting amplifier (with ~6 voltage gain to produce ~30V from 5V maximum charge pump voltage), then a third RC pole. (Type B for illustration on this link.)
My question is the following: does it cause any problems that the minimum VCO tuning voltage is higher than the minimum voltage that the filter can produce? Far below the minimum tuning voltage the VCO doesn't oscillate or spuriously oscillates. Or should I use another active filter topology in which half of the charge pump supply is lead on the non-inverting input of the amplifier (so called type A active filter)? To be honest, I don't exactly understand what happens because of that (I just read that it is optimal for spurs), am I right in thinking that the "initial" tuning voltage will be half of the charge pump supply? (So my problem would be solved because upon power-up the tuning voltage would be 2.5V*gain of the opamp?)
Thanks in advance for any insights!
Hi,
It sounds more complicated than it has to be.
May I ask why you don't use a ready to buy PLL IC?
Klaus
It is a school project, the emphasis is on the development and study of the microwave VCO, the PLL is just an enhancement for it - if your question was that why don't I use a PLL with VCO integrated. Otherwise, the PLL will be an IC, only the loop filter has to be realized with external components. To simplify things, the only question is: can a non-zero minimum tuning voltage cause problems during locking?
Depends on the specific type of PFD your PLL uses...
Simplest solution would be to diode clamp the output of your opamp circuit to be within the oscillation range of the VCO, even when the charge pump output is near zero.
My question why the VCO has a lower limit for tuning voltage ? It must work along whole tuning voltage range.
If it doesn't work, the VCO can always create a problem.Otherwise the VCO should re-designed regarding to actual voltage range without off-set.
Because the varactors have a high 0...30V tuning range, however, feedback filter Q changes with tuning so I can't use the whole range. Another reason is that electric length is a function of frequency and the phase shift of the feedback path cannot be approx. constant for a wider band, so I had to use a smaller portion of the varactor voltage range, and because of the geometry of the filter (reliable manufacturability) this range proved to be the best.
By the way I think this is not really uncommon, a lot of VCOs have a minimum tuning voltage, the question is if it causes any problems for phase-locking.
@mtwieg thanks for the clamping idea, simple but great!
Why don't you use another varactor which gives you the same amount of capacitance values between 0-15V tuning voltage ? It will be trouble-free..
I do not recommend to use anything after OPAMP including clamping diode or something else because they will contribute directly into Phase Noise.
You would want to use an op amp that was "single supply, rail to rail operation".
something like this
https://www.digikey.com/product-deta...BRZ-ND/6490582
watch out for the op amp equivalent noise though, you might have to search around for a lower noise op amp based on the output phase noise requirement of your PLL