Can I use the lowpass filters in direct cascade ?

[saulbit]

I am dealing with a troublesome circuit. The first IF is 5.4GHz, and the bandwidth of the first IF filter is about 100MHz. What is more, the second IF is about 300MHz, so I have to rejecte the possible (5.4GHz-2*300MHz) outputs of the first mixer (The first LO is higher than first IF). I have to choose a bandpass filter, with center frequency of 5.4GHz. The parastic band of such filter is a bit severe in 9GHz. So I want to use a LFCN-5500+, which is a lowpass filter and could offer about 30dB rejection in 9GHz, direct cascade with the bandpass filte ? Can I just do it? I can't find the answer in many books.
By the way, since the passband of the LFCN-5500+ is beyond 5.4GHz, so the impedance of bandpass filter in 5.4GHz band is not influenced by the LFCN-5500+. I mean in direct band, not for the case of insertion of an amp.
Here is the datasheet of the LFCN-5500+.

 

[Fovakis]

Hi friend. Is it easy to upload a block diagram of the RF Downconverter you make and write the frequency ranges of L.O IF Baseband etc...

Regards,
George

 

[vfone]

Actually the definition of the Image Frequency is:
f_imag = f_RF + 2*f_IF in case of high side injection (LO higher than RF) and:
f_imag = f_RF - 2*f_IF in case of low side injection (LO lower than RF)

In case of high side injection (LO higher than RF) a Low Pass Filter can be used as an Image Filter in front of the mixer, and yes in this situation the RF Band Pass Filter can be followed by a Low Pass Filter to improve image rejection.

In case of low side injection, cannot be used a LPF as Image Filter (but only a BPF or a HPF), because Image Frequency is situated on the other side.

 

[shashy.br]

You can cascade the LFCN filters but the Radiation effects reduce the combined rejection,

Better design for 3 times the attenuation at the output that required at the stop band.

 

[pstuckey]

The important thing to remember about cascading filters is that the S-parameter data you get from the manufacturer is taken with a very good 50 ohm load on input and output. Outside the passband, your BPF will not look like 50 ohms, so you may not get as much attenuation as you expect from the LPF. Putting a small resistive pad between the two filters will improve performance if you can take the gain hit.

 

[biff44]

ps is right, u can not cascade the two. Your bpf rejection will suffer, and your lpf will not perform as specified. in fact, if you had a line length between the two filters of half wavelength at 9 ghz, you might have 0 dB rejection!


If you were to add a 3 dB pad between the BPF and LPF, then the combined rejections can be achieved...you have provided close to a 50 ohm load to both of them. Of course, you could also add an amplifier stage or a ferrite isolator between the 2 filters and achieve the same good performance.