Question about designing cascading filters

Hi friends,

I'm designing a bandpass filter by cascoding a LPF and a HPF.
For LPF, fc=2MHz. For HPF, fc = 257MHz.
After cascading them, a bandpass filter is formed with LOWER fc=2MHz and UPPER fc=234MHz. So there is drop in the UPPER fc. Why is this so??

Any comment would be very appreciated.

Will

Re: Cascading Filters

HPF is probably 2MHz and 257MHz is LPF othervice you have bandstop filter.
It may depend on filters circuit topology or impedance mismatch when filters are connected.
Post your circuit.

Cascading Filters

there should be loading effect of the other filter , u can try to put a buffer , or simply re-adjust the HPF till u get the performance u want. actually u should try to get the transfer function of the whole filter to get the exact location of poles and zeros and the effect of each component on them.


sorry i meant u should adjust the LPF to set the fc high.

Re: Cascading Filters

Well I guess you did mix up when writing your post, the LPF has a Fc=257MHz while the HPF has a Fc=2MHz.
Either that or your design is actually a band stop filter instead of a band pass.

As for your question, I agree with what Safwat has just mentioned about the loading effect between both filters. Cascading 2 filters will surely alter the characteristics of each filter operating individually since the load resistance of the first filter is the input impedance of 2nd stage, while the source resistance of 2nd filter is turned into the output impedance of the 1st filter.

Try resimulating using appropriate load and source impedances and keep us updated, maybe my opiniong is wrong, but then only you can prove me either right or wrong.

Salam for now

Re: Cascading Filters

Hi friends,

Thanks for all of your comment. ^.^" Sorry that I've made the mistake ~ fc of HPF = 2MHz, fc of LPF = 257MHz ~ so it is a bandpass filter.

I designed a BPF by cascoding a HPF and a LPF. In between them, I inserted a voltage buffer for boosting the gain above 0dB. From the 1st stage point of view, its input is the source resistor and its output is the input of votlage buffer(Rin=inf). From the 2nd stage point of view, its input is the output of the voltage buffer(Rout=0) and its output is the load resistor.

I did the simulation for them separately under the above conditions. The results looks good. But after I cascode them + a voltage buffer in between them, the results show that there is a drop at fupper by 20MHz. Please note that there is still the drop after inserting the buffer !!!!!! In view of this, as suggested by safwatonline, I set fupper higher until meeting the specification. Now the results look good.

That's all for the story. Thanks again for your comment.

Will

Re: Cascading Filters

There are two second order effects to keep in mind.

1. Parasitic capacitances in the wiring and the buffer input and output. These will change the frequency response. As an example, your 1 pF series input capacitor will form a voltage divider with the several pF wiring capacitance and buffer input capacitance to reduce the gain.

2. The order of the filters should be chosen so that the first filter cuts out the strongest out of band signals. In your case this may be the MW AM broadcast band cut out by the HPF or the UHF television band cut out by the LPF depending on where the product is used. This will minimize the IMD generated by the buffer stage.

Re: Cascading Filters

Your buffer amplifier might distorte response. Check it first without LPF.

Re: Cascading Filters

Actually, the voltage buffer I used is a voltage-controlled voltage source in HSPICE. So it has infinite input resistance and zero output resistance, which should not distort the performance.