create new tag
view all tags

IF Filter Design Information

  • This bandpass filter design is based on a LP321010 prototype which has maximum ringing of 1% outside the defined time and frequency passbands. With a 250 MSPS ADC, we are limited to a 125 MHz passband, so I choose fcenter= 75 MHz as target. The filter bandwidth is BW=75 MHz at -40 dB (1%), with two zeros at fp=129.75 MHz and fm=43.35 MHz. You find more details in the attached documents:

  • BPF321010ideal
  • The is the BPF with ideal components. Please note the little capacitor Ci=1.061 pF between the two resonators, which I included into the design to account for stray capacitance's to ground. The resonators in the center define the frequencies with maximum attenuation (zeros in the transfer function).

  • BPF321010real
  • I used this setup to simulate the circuit with real Coilcraft aircore inductors. The first resonator is particular critical, as we have a LC resonator of a large capacitor in parallel with a small inductor. The capacitors in the simulation are still ideal, we have to use the best ultra low loss ceramic capacitors available, Mouse has a good selection. For the inductors I used:
    • Coilcraft GA3094-AL (12 nH)
    • Coilcraft Midi Spring 1812SMS-R12 (120 nH)
    • Coilcraft Square Air Core 2929SQ-361 (360 nH)
    • Coilcraft Midi Spring 1812SMS-68N (68 nH)
  • The 68 nH inductor fall a bit short, so you may add some extra trace in the layout to add 3..4 nH inductance. For the first resonator I used tow capacitors in parallel, perhaps in practice it is better to use capacitors of similar values, summing to ~345 pF.
  • It is very important that the inductors don't couple to each other, so orient them 90 degree on the PCB
  • Again, the little Ci=1 pF is used only to account for the stray capacitance's from the two central inductors to ground, it should not be a physical component!

  • BPF321010_magn
  • Frequency domain magnitude response for the simulation with ideal (thin red) and real (thick blue) components.

  • BPF321010_imp
  • Impulse response for the simulation with ideal (thin red) and real (thick blue) components. The impulse was a pulse with 1 ns duration, excited with a delay of 5 ns. Obviously we don't hit the target for 50 ns response time, it is more 60...65 ns, still acceptable I would say. More bandwidth will not help, as the ADC / digitizer cannot make any use out of it.

Physics WebpagesRHUL WebpagesCampus Connect • Royal Holloway, University of London, Egham, Surrey TW20 0EX; Tel/Fax +44 (0)1784 434455/437520

Topic revision: r1 - 28 Oct 2014 - JackTowler

This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2024 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding RHUL Physics Department TWiki? Send feedback