"2 Lecture schedule TSEK38: Radio Frequency w4: Transceiver Design • Le1: Introduction (Ch 1) • Le2: Fundamentals of RF system modeling (Ch 2) Lecture 3: Superheterodyne • Le3: Superheterodyne TRX design (Ch 3.1) TRX design w5: • Le4: Homodyne TRX design (Ch 3.2) • Le5: Low-IF TRX design (Ch 3.3) Ted Johansson, ISY w6: [email protected] • Le6, Le7: Systematic synthesis (calculations) of RX (Ch 4) w7-8: • Le8, Le9: Systematic synthesis (calculations) of TX (Ch 5) TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "3 "4 Repetion of Lecture 2 Example Output contains the Linear same frequency components as input, • Linear, Time-Invariant systems they are just stronger Vout = G * Vin A cos ω t GxA cos ω t • Linearity: 1 1 – Harmonic distortion -ω1 ω1 Both inputs are amplified by Linear – Cross-modulation the same amount – Compression point Vout = G * Vin A cos ω1t GxA cos ω1t – IP3 B cos ω2t GxB cos ω2t Input signals are • Noise Figure stronger now Different tones are Non-linear • Phase Noise amplified differently Vout = G(Vin) * Vin New frequencies K1 cos (2ω1-ω2 ) t A’ cos ω1t appear at the output A’’ cos ω1t B’ cos ω2t B’’ cos ω2t K2 cos (2ω2-ω1 ) t TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "5 "6 Time Invariance Effects of Nonlinearity • Time-variant = response depends on the time of • Consider a nonlinear memoryless system origin. x(t) y(t)= $ $ V + $ V2 + $ V3 +... • A system is time-invariant if a time shift in its input 0 + 1 in 2 in 3 in results in the same time shift in its output. • Let us apply a single-tone (A cosωt) to the input and If y(t) = f [x(t)] calculate the output: then y(t-#) = f [x(t-#)] x(t) = Acosωt y(t) Acos t A2 cos 2 t A3 cos3 t • LTI: Linear time-invariant, y(t-#) = L * [x(t-#)]. ! = α1 ω +α 2 ω +α 3 ω α A2 ⎛ 3α A3 ⎞ α A2 α A3 A time shift in the input causes the same time 2 A 3 cos t 2 cos 2 t 3 cos3 t = + ⎜α1 + ⎟ ω + ω + ω shift in the output. 2 ⎝ 4 ⎠ 2 4 • Examples: filters, isolators, duplexers, linear Second Third amplifiers. DC Fundamental Harmonic Harmonic TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "7 "8 Gain (1dB) Compression (p. 32) Desensitization (p. 32) • Eventually at large enough signal levels, output power • At the input of the receiver, a strong interference may exist close to does not follow the input power, the desired signal. • The small signal is superimposed on the large signal (time domain). If the large signal compresses the amplifiers, it will also affect the small signal. The P-1dB point correlates well to loss of linear behavior, getting out- of-spec in standards (EVM, ACPR, etc.) so for linear applications, operation beyond this point is useless. TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "9 "10 Cross-modulation (p. 32) Intermodulation – Intercept Point • Any amplitude variation (AM) of the strong interferer A2 will also For a given input level (well below P1dB), the IIP3 can be calculated appear on the amplitude of the signal A1 at the desired frequency by halving the difference between the output fundamental and IM and distort the signal. levels and adding the result to the input level, where all values are • Interferer: results in: expressed as logarithmic quantities. extra AM modulation occurs TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "11 "12 Cascaded Nonlinear Stages Noise Figure/Factor • If each stage in a cascade has a gain greater than • Noise figure (factor) shows the noise performance of a unity, the nonlinearity of the latter stages becomes system. increasingly more critical because the IP3 of each Noise Factor stage is equivalently scaled down by the total gain preceding that stage. • It can be expressed in dB: Noise Figure • NF depends on not only the noise of the circuit under consideration but the SNR provided by the preceding stage. 1 1 G G G = + A + A B • If ideally a system adds no noise, F=1. IIP3total IIP3A IIP3B IIP3C • If the input signal contains no noise, NF=%. TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "13 "14 Noise Figure of Cascaded Stages Phase noise • The phase of the oscillator varies as Acos(&c(t)+'n(t)). • The term 'n(t) is called the “phase noise.” • Can also be viewed as a random frequency variation, leading to a broadening of the spectrum called phase noise. Gain is power gain, which depends on the impedance of each stage. TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "15 "16 Phase noise From TSEK03 RFIC, Razavi's book Ch. 8 Example 8.23 From TSEK03 RFIC, Razavi's book Ch. 8 • Since the phase noise falls at frequencies farther from &c, it • At high carrier frequencies, it is difficult to measure the noise must be specified at a certain “frequency offset”, i.e., at a power in a 1-Hz bandwidth. Suppose a spectrum analyzer certain difference with respect to &c. measures a noise power of -70 dBm in a 1-kHz bandwidth at • We consider a 1-Hz bandwidth of the spectrum at an offset of 1-MHz offset. How much is the phase noise at this offset if the average oscillator output power is -2 dBm? (f, measure the power in this bandwidth, and normalize the result to the “carrier power”, called “dB with respect to the carrier”, dBc. TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "17 "18 Example 8.23 From TSEK03 RFIC, Razavi's book Ch. 8 Outline of lecture 3 • At high carrier frequencies, it is difficult to measure the noise • Introduction RF TRX architectures (3) power in a 1-Hz bandwidth. Suppose a spectrum analyzer measures a noise power of -70 dBm in a 1-kHz bandwidth at • Superheterodyne architecture (3.1, 3.1.1) 1-MHz offset. How much is the phase noise at this offset if the • Frequency planning (3.1.2)! average oscillator output power is -2 dBm? - IF selection (3.1.2.1)! - Spurious analysis (3.1.2.2) • Design Considerations (3.1.3) • Summary TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "19 "20 Generic RF Transceiver RF transceivers main building blocks • frequency filters • amplifiers • frequency converters • modulator/demodulators • oscillators • synthesizers • ADC/DAC • signal coupler/divider/combiner/attenuators • switches • power/voltage detectors • … TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "21 "22 Transceiver architectures • Superheterodyne (Ch 3.1) • Armstrong invented – Most popular (and still is), invented in 1918. the superheterodyne – Somewhat complex and limited flexibility by the fixed in 1918. filters • Homodyne (direct conversion, zero-IF)(Ch 3.2) – Integratabtle – Flexible • Low IF (Ch 3.3) to overcome some drawbacks with the homodyne S. Maas, IEEE Microwave Magazine, Sep/ Oct 2013, p. 34 • IF bandpass sampling (3.4), Software-defined radio, … "Additional readings" folder TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "23 "24 Quality Factor (Q) The SAW filter • Quality factor of a filter is a quantitative measure of how • SAW (surface acoustic wave) filters are electromechanical devices much loss the filter exhibits commonly used in radio frequency applications. – Lower quality factor indicates more losses • Electrical signals are converted to a mechanical wave in a device constructed of a piezoelectric crystal or ceramic; this wave is – Practical filters (especially on-chip filters) have losses delayed as it propagates across the device, before being converted and therefore low Q back to an electrical signal by further electrodes. The delayed outputs are recombined to produce a direct analog implementation • It can be shown that the quality factor is inversely of a finite impulse response filter. proportional to the fractional bandwidth of the filter: • This hybrid filtering technique is ! )f = BW / fc! also found in analog sampled ! ! filters. fc is the center frequency, BW is the -3 dB limit. • SAW filters are limited to ! frequencies up to 3 GHz. To have a small BW at high fc, a filter with very high Q is needed [Wikipedia] TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "25 "26 The SAW filter Cavity filters for basestations • SAW filters combine low insertion • Wikipedia: "Cavity filters loss with good rejection, can achieve broad bandwidths and are a tiny are the basic circuitry fraction of the size of traditional behind a duplexer and cavity and even ceramic filters. are sharply tuned • Because SAW filters are fabricated resonant circuit that allow on wafers, they can be created in large volumes at low cost. only certain frequencies • SAW technology also allows filters to pass. and duplexers for different bands to • The Q of larger be integrated on a single chip with little or no additional fabrication microwave cavities may steps. exceed 10 000. • Huge! Something like ! 40 x 70 x 15 cm. [Wikipedia] TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson TSEK38 Radio Frequency Transceiver Design 2019/Ted Johansson "27 "28 Channel Selection The problems of channel selection: • Most communication systems divide the frequency • Problem: We need to limit the bandwidth for band into several narrower channels.