Various architectures of receivers has been proposed in literature, but most popular architectures among them like Heterodyne, Homodyne, Wideband-IF and Low-IF are presented here.
Super-heterodyne Architecture was invented by Armstrong in 1917 and is the most widely used architecture in wireless transceivers so far. It is a dual conversion architecture, in which, at the first state RF is down-converted to IF and then, in second stage it is from IF to baseband signal. The block diagram of super-heterodyne receiver architecture is shown in Figure.1
From the incoming RF signal preselection filter removes out of band signal energy as well as partially reject image band signals. It is then amplified by LNA to supress the contribution of noise from the succeeding stages. Image Reject filter attenuates the signals at image band frequencies coming from LNA. Mixer-I downconverts the signal coming out of the IR filter from RF frequency to IF frequency with the output of a Local Oscillator. The channel selection is normally achieved through IF filter : It is a BP filter to allow the IF band of interest and other band is rejected. This filter is critical in determining the sensitivity and selectivity of a receiver. Since channel selection is done at IF1, the LO requires an external tank for good phase noise performance. In case of phase or frequency modulation, downconversion to the baseband requires both in-phase(I) and quadrature(Q) components of the signal. Mixer-II does the second down conversion of IF signal into I and Q components for digital signal processing. The LP filter acts as a channel reject filter along with job of anti-aliasing functionality.
- IR filter and channel selection.
- Good sensitivity and selectivity.
- High Q filter
- High performance oscillator or LO
- LNA output impedance matched to 50 ohm is difficult.
- Integration of HF image reject filter is a major problem
Homodyne receivers translates the channel of interest directly from RF to baseband(ωIF=0) in a single stage. Hence these architectures are called Direct IF architectures or Zero-IF architectures. For frequency and phase modulated signals, down conversion must provide quadrature outputs so as to avoid loss of information. The block diagram of heterodyne architecture is illustrated in Figure.2.
Merits of Zero-IF architecture are
- Less hardware
- No image problem. So image filter not required.
- Because of no IF stage, LPF is sufficient for filtering.
- Amplification at BB stage. Hence power saving.
- In integrated circuits LNA need not to match to 50 ohm. Because no image reject filter between LNA and mixer.
De-merits of Zero-IF architecture are
- LO Leakage : Generally there will be a imperfect isolation between LO port and input port of mixer and LNA, due to capacitve and substrate coupling. Because of this there will be LO feedthough from LO port to the input port of the mixer and LNA. This LO leakage mixes with original LO, called self-mixing, produces DC offsets in the mixer output and causes saturation of following stages in the receiver chain.
- DC offset errors : It is the most serious problem in the baseband section of the homodyne receivers. The cause of it is self mixing of LO leakage which is due to LO feed through to mixer input port and LNA and insufficient isolation between LO port to mixer input port and LNA input.
- Since LO frequency is same as carrier frequency, it leaks from receiver to antenna which interfers with same frequency-band receivers.
- Flicker noise from an active device contaminate the BB signal.
- I/Q mis-match :
- Even order distortion
Wideband-IF receiver is a dual conversion architecture in which data is downconverted from RF to IF in the 1st stage, and in the 2nd stage it is from IF to Baseband. The block diagram of Wideband-IF receiver architecture is shown in Fig.3.
In this architecture all the RF channels are complex mixed and downconverted to fixed IF after preselection filtering and amplification. In second stage an Image Reject(IR) mixer does complex mixing and translate IF to BB using a tunable channel select frequency synthesizer. All the image frequencies are cancelled by IR mixer. If the IF is chosen high enough, additional image rejection may be obtained from the RF front-end preselection filter. Channel selection is performed at baseband by using programmable integrated channel select filter. Since LO-1 is fixed frequency synthesizer generated by crystal controlled oscillator good phase noise performance is obtainded. Channel tuning is acheived by uisng programmable frequency synthesizer at IF.
In Low-IF receiver architecture all the RF signals are translated to low-IF frequemcy which is then down-converted to BB singal in digital domain. Low-IF architecture comprises the advantages of both heterodyne and homodyne receivers. The block diagram of Low-IF receiver architecture is shown in Figure.4.
After preselection filtering and amplification, all the RF channels are quadrature mixed and downconverted to low IF containing both wanted and unwanted singnlas. The IF frequency is just one or two channels bandwith away from DC, which is just enough to overcome DC offset problems. It is then amplified and filtered before sampled by ADC. Since the ADC samples both wanted and unwanted signals, there will be higher demand on ADC dynamic range requirements. The ac-coupled signal path to ADC eliminates the need of DC offset compensation circuitry. The sampled digital data is fed to image reject mixer which is implemented in digital domain.