a tank bandpass filter's problem
but the capacitance value 1.7 pf , is already exist , u can get it from murata , or epcos
khouly
Agreeded, you can get capacitors nowadays down to 0.1 pF.
IF you want to use a bigger value than 1.7 pF, you can use a transformer (an actual transformer, or an L-C impedance match network) at the input and output of the filter to make the bandpass filter work at a different impedance level.
In any event, you should be sure to simulate this filter with lossy components (ie. add the Q of the inductors and capacitors), as your filter will have a lot of loss and distortion of the reject bands.
Fig 1 shows the response of your filter, fig 2, using ideal inductors.
The proposed filter topology is best suited for BW>30 %, but your filter has a BW of 7 %. This gives a low value of C2 and L1/L3. This may give problems with parasitics from the PCB layout (stray capacitances/inductances from interconnection pads, etc).
A better approach is the schematic in fig 3, which gives more convenient component values. However, 7 capacitors instead of 3 are used. The frequency response is in fig 4 with ideal inductors. Using a Q of 150 of the inductors gives the response in fig 5.
The filter screenshot you have posted has 9 MHz BW, not 18 MHz (136-118=18 MHz). An 18 MHz wide filter will have the schematics in fig 6 and frequency response in fig 7, with an inductor Q of 150.
Hi
If you use an L-C impedance network, will this only work if you are dealing with narrow band filters?
Regards
