Help me remove spurs from PLL output

Hello friends

I designed a PLL+VCO circuit at 160MHZ with LMX2316 and minicircuit VCOs and 25MHZ crystal osc as ref.

Output is stable (Fcom=20KHZ,loop filter BW = 1KHZ),but as shown below have spurs at output near 160MHz carrier.Please guide me to remove the spurs.

hi 7rots51,

for good answer i need the schematics. if you can please show only the high side of the signal with span of 2 MHz?

What is the delta frequency between the 160MHz and the spur?
Is your charge pump comparison frequency 20 kHz?
Are you using a switching power supply? Power supply noise will appear as a spur on the PLL.
Do you have the extra RC network on the output of the loop filter (R3/C3) or are using just the two caps/1resistor network (R1, C2/R2)?

I am going to bet you are using a .714286 MHz phase detector frequency, and this is just that spur (R=35, N=224). Try:

1) Using lower divisor numbers (higher phase detector frequency), such as R=15 and N=96. This will give you a higher phase detector frequency of 1 2/3 MHz

2) Add an RC lowpass filter at the VCO tune port to filter out phase detector spikes at 1 2/3 MHz. I would put in a lowpass with a 3 dB corner at around 20 KHz, and see if you like it better. (Whenever you add an additional lowpass filter you compromise the phase margin of the control loop, so you want to keep the lpf corner frequency say 20X the loop bandwidth, which I assume you correctly state as 1 KHz).

I agree, it is probably your detector frequency which is too low, or your loop filter does not remove the spurs enough. You can consider selecting a higher order loop or simulate the filtering performance at the phase detection frequency.

By adding a 10uf tan cap to VCO supply (there was only 1u ,100n ) and changing channel spacing to 25khz and loop filter to 250Hz ,I got this spectrum,Is it ok?

the phase noise is -90dBc ,is it enough?

the sch of circuit is in pdf file(simulation with analog device PLL software)

bye

hi , it is a good results the spurs are disappeared , and it is great

about the phase noise u should u know is it good or band from the specs u need to have
and also u forget to mention the offset u measure the phase noise for


khouly

Hello 7rots51;
It's an old post, but I have an opinion. I think that your new narrow LPF bandwidth increases the settling time of PLL. Is it true?