PLL Frequency Accuracy
I designed a Synthesizer with ADF4156 for 4.4 to 5 GHz. Vtunemax=10V. I′m using an active loop filter(AD820) and a TCXO (+-1.5ppm). The PLL locks very fast and well. But after analyzing the outputsignal in more detail I found out, that there is a small deviation in frequency. The frequency changes in a range of about 100Hz. Is it possible to get the PLL more stable? Is the small Vtunemax the problem? Is there any rule of thumb how accurate the outputfrequency can be?
..my next step will be to remove the ceramic C of the loop filter.
best regards
I didn't understand if you have a small frequency instablity (i.e.: thefrequency continuosly changes around the wanted value) or a fixed frequency shift. However I think simulations can hael you tofix the problem. I suggest ADISimPLL
I used ADISimPLL for the design. I ′ve a small frequency instablity. The frequency changes around the wanted value (+-50Hz)
There is not really enough information here to help you very much beyond generalisations.
You don't say how fast these frequency variations are happening.
What is your reference frequency and how stable is it over the time frame of these variations
Assuming it is a 10MHz TCXO then it would have to be stable to within 0.5Hz to ensure that the frequecny drfft at the output is less than 100Hz.
The reference would be my first point to check if it is a slow change.
If it is more rapid then look for interference coupling into the loop filter somwhere or general loop stability.
Peter
1.5 ppm means your microwave frequency will be +/- 7500 Hz accuracy. And you are complaining that it is not, what, accurate to 1 Hz?
Yes, I agree, this is why I asked if there is an instability or a frequency shift (or both of them).
If your filter is "inverting type", then you should have a resistor divider connected to the non-inverting pin of the opamp: a decoupling capacitor in parallel with the resistor to ground is often required.
If your TCXO is actually a VCTCXO the Vctrl pin if left unconnected it could pick-up noise from the environment.
I think you should detail a little bit more the situation.
No, there is no instability in your circuit. I wish to have 100Hz drift (0.02 ppm) in all my 5GHz PLL's that I am using.