low-IF receiver architecture
its a common fact that zero-IF receiver can be implemented with atleast one local oscillator stage less than their low-IF counterparts..... but still low-IF architecture is more popular as compared to their zero-IF counterparts... can anyone tell me the reason for it...
First, nowadays Low-IF is NOT more popular than Zero-IF receivers, and second, both use just one local oscillator.
The big problem with Low IF receivers it the 1st Image problem. You're relying on the I/Q matching of the mixers to cancel out the Image, and it's tough to get more than 30 dB of image rejection. As the Image for Low IF is pretty close in, you'll get hammerred with Adjacent/Alternate selectivity tests. It does help you avoid the problem of LO Leakage, but the other problems are tough to deal with.
All of Navini's (bought by Cisco late 2007) WiMAX product line subscriber equipment used direct conversion receivers, and their "P4" (Phase 4 product, the last one designed but not sold circa 2010) base station/infrastructure 'carrier class' also used direct conversion receiver.
The saving in power was great over the P3 (Phase 3 product) because the SAW filters (with high insertion loss) did not require additional gain to offset the loss, and a 2nd LO was necessary for P3 as the IF was a "direct conversion receiver" in the 400 MHz range!
Today, ceramic RF filters are cheap enough and small enough to 'chop' up the bands like licensed WiMAX bands in the USA to several segments (for possible spurious signal filtering) and the "IF" filtering can be done by the TI TRF3711 series of direct conversion ICs which have a built-in BW control/selection to match the occupied channel bandwidth.
Jim