why zero-IF architecture is not used in FM receivers

[yixiusky]

Could anyone teach me, why zero-IF architecture is not used in FM receivers design ? Thank you very much

 

[biff44]

Well, if you have an rf carrier, say 1000 MHz with fm deviaton +/- 20 KHz, and you downconvert it to zero center freuqency, the "IF" output will consist of negative frequencies and positive frequencies. Most FM demodulators can not tell the difference between negative and positive frequencies, so the demodulator stops working. it would see 20 KHz and then 20 KHz and not see a change.

Zero IF can be done, it is just more complicated.

One approach is to do zero IF for some modulations, and "Low IF" for FM.

 

[mtwieg]

Well, if you have an rf carrier, say 1000 MHz with fm deviaton +/- 20 KHz, and you downconvert it to zero center freuqency, the "IF" output will consist of negative frequencies and positive frequencies. Most FM demodulators can not tell the difference between negative and positive frequencies, so the demodulator stops working. it would see 20 KHz and then 20 KHz and not see a change.
Zero IF can be done, it is just more complicated.

One approach is to do zero IF for some modulations, and "Low IF" for FM.

I can see how low-IF could work by biasing the IF at half the channel bandwidth or greater. But in order to prevent adjacent channels from aliasing/imaging in, you'd need larger channel separation which is a pretty big deal.

 

[biff44]

Yes, you are right. If you are trying to make a very high adjacent channel rejection receiver...a zero IF or low IF are not the best choices. Chips like the ADF7023 use a low IF frequency who's actual center frequency depends on what IF filter bandwidth you choose. That somewhat optimizes things, but you are still only going to get maybe 23 dB of adjacent channel isolation.

If you really want good channel rejection (including image), you probably need to:
1) use a double conversion
2) have analog filtering (perhaps at 10.7 MHz) as well as digital dsp filtering.

I just did a receiver with 1st IF at 433.92, and 2nd IF at 10.7 MHz, and it achieved an honest 50 dB of adjacent channel rejection. (note, most data sheets bend the truth a little by testing adjacent channel rejection with a CW interferer tone--not a similarly modulated interferer)...beware.

Rich