Analog Band Pass Tunable Filters (ADMV8420,ADMV8416,HMC892A,HMC891A)

Dear All!

I'm going to use Analog Band Pass Tunable Filters (ADMV8420,ADMV8416,HMC892A,HMC891A) for DDS base LO.

Unfortunately the specifications do not provide any typical schematic or circuit  for manage Vcnt control Pin's.

The main idea is to control the bandpass filters by DAC with OP AMP with the higher update rate  up to 200 ns settling time, without reducing the LO phase noise performance.

In my opinion the Vcnt pins are very sensitivity to noise, and If you can provide me some high frequency low noise application circuits to do it I will be apparated.    

Thanks in advance

Aram Khachatryan

Parents
  • +1
    •  Analog Employees 
    on Apr 22, 2021 8:12 PM

    Hi Aram,

    Sorry for the late reply. You are on the correct thinking regarding speed and noise performance.

    In the past we have used the AD5760 DAC followed by the ADA4898 op-amp as a unity gain buffer. The DAC reference was the LT3042 voltage regulator set to 15V. This resulted in excellent baseband noise spectral density on the order of 10nV/rtHz, so the residual phase noise floor of the filter would not be impacted (better than -150dBc/Hz). The downside was the voltage tuning speed was on the order of 1usec, which isn't quite your desired 200nsec time.

    You might find a faster DAC or op-amp in the catalog, but keep in mind the baseband noise spectral density may impact phase noise results. As a rough order of magnitude for phase noise floor, you could use small angle phase modulation equation of 20*log(Vnoise_in_rms * Filter_Phase_Sensitivity / sqrt(2)).

    Best Regards,

    David

Reply
  • +1
    •  Analog Employees 
    on Apr 22, 2021 8:12 PM

    Hi Aram,

    Sorry for the late reply. You are on the correct thinking regarding speed and noise performance.

    In the past we have used the AD5760 DAC followed by the ADA4898 op-amp as a unity gain buffer. The DAC reference was the LT3042 voltage regulator set to 15V. This resulted in excellent baseband noise spectral density on the order of 10nV/rtHz, so the residual phase noise floor of the filter would not be impacted (better than -150dBc/Hz). The downside was the voltage tuning speed was on the order of 1usec, which isn't quite your desired 200nsec time.

    You might find a faster DAC or op-amp in the catalog, but keep in mind the baseband noise spectral density may impact phase noise results. As a rough order of magnitude for phase noise floor, you could use small angle phase modulation equation of 20*log(Vnoise_in_rms * Filter_Phase_Sensitivity / sqrt(2)).

    Best Regards,

    David

Children