conquering noise for accurate rf and microwave signal
TRANSCRIPT
![Page 1: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/1.jpg)
Conquering Noise for
Accurate RF and
Microwave Signal
© Agilent Technologies, Inc. 2009
Microwave Signal
Measurements
Presented by: Ernie Jackson
![Page 2: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/2.jpg)
The Noise Presentation
• Review of Basics, Some Advanced & Newer Approaches
• Noise in Signal Measurements-Summary
• Basic Noise Reduction Approaches
© Agilent Technologies, Inc. 2009
• Basic Noise Reduction Approaches
• New Technologies
• Examples
• Resources, More Information
![Page 3: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/3.jpg)
Noise is the Enemy(Almost Always)
• Creates Uncertainty in Time, Frequency, Power, Modulation
• Converts Deterministic Values to Statistical
Communications, RADAR or other Imaging, Metrology, Science
© Agilent Technologies, Inc. 2009
• Converts Deterministic Values to Statistical Distributions
• Imposes Cost Tradeoffs: Accuracy, Speed, Repeatability, Yield, €/£/¥/$, “False Alarms”
• Financial Impacts: Circuit Cost, Size, Complexity, Development Time, Mfg.
![Page 4: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/4.jpg)
Noise is the Enemy (cont)
• Phase Noise
– Increases noise floor close-in
– Creates same problems as broadband noise
– Imposes similar tradeoffs, costs
• Solutions for Analysis are Similar
© Agilent Technologies, Inc. 2009
• Solutions for Analysis are Similar
– Optimize measurement technique
– Choose a higher performance analyzer
– Use special techniques, processing
Use “Moore’s Law” where it does not
normally apply
![Page 5: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/5.jpg)
Exceptions: Desirable Noise
• Noise Figure Measurements
• Accurately Generate to Test for Margins, Real World Performance
– Generate known signal/noise, phase noise
• Broadband/Narrowband Noise as Stimulus
© Agilent Technologies, Inc. 2009
• Broadband/Narrowband Noise as Stimulus Signal, with Different Statistics
• Add to Simulations to Avoid Non-Representative Behavior
• Noise Floor Appearance as a Clue or Hint
![Page 6: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/6.jpg)
Noise Can Frequently be IgnoredBut it is Important to be Alert
• Good Signal/Noise Ratio
• Measurement BW Wider than Signal
• Alert for Clues
– Measurement variance large compared to
© Agilent Technologies, Inc. 2009
– Measurement variance large compared to desired accuracy
– Measurements change with analyzer settings
– Decreasing SNR
– Comparing signals with different statistics
![Page 7: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/7.jpg)
Optimize Noise vs. Other Measurement Priorities
• Attenuation = 0 dB Optimal for Sensitivity
• Attenuation = 10+ dB for Optimal Accuracy
• Attenuation = 4-6 dB a Compromise
© Agilent Technologies, Inc. 2009
• Attenuation = 4-6 dB a Compromise
• Problems with Impedance Match
• Problems with Distortion, Compression
• Chance of Analyzer Input Damage
![Page 8: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/8.jpg)
Trade Away Extra Noise Performance for other Benefits
• Better Overload Performance
• Lower Distortion Measurements
• Faster Measurements
© Agilent Technologies, Inc. 2009
![Page 9: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/9.jpg)
Bad Measurements are Easy to Make
Spur appears
3 dB above
noise level
© Agilent Technologies, Inc. 2009
![Page 10: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/10.jpg)
Error in Noise Level Measurements, Measured Near Analyzer Noise floor
Actual Noise Ratio Measurement Error
0 dB 3.01 dB
1 dB 2.54 dB
3 dB 1.76 dB
5 dB 1.19 dB
© Agilent Technologies, Inc. 2009
Note that error is always positive
5 dB 1.19 dB
10 dB 0.41 dB
15 dB 0.14 dB
20 dB 0.04 dB
![Page 11: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/11.jpg)
Multiple Tones Example
© Agilent Technologies, Inc. 2009
“No” error for best S/N, 3 dB for signal at analyzer DANL
![Page 12: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/12.jpg)
Noise Converts ~Deterministic Value to Statistical Distribution
GSM (MSK)SNR 60+ dB
CCDF1 dB/div
1 MHz Span*
© Agilent Technologies, Inc. 2009
SNR10 dB
* CCDF varies significantly with span at low SNRs
SNR20 dB
AWGN Ref
![Page 13: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/13.jpg)
Clues to Two Meas ProblemsSignal Close to Noise, Wider than RBW
© Agilent Technologies, Inc. 2009
![Page 14: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/14.jpg)
Narrower RBW Improves S/N
© Agilent Technologies, Inc. 2009
• Residual FM, signal is wider than RBW filter
• Amplitude measured low in error
![Page 15: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/15.jpg)
Band Power Marker, Averaging
© Agilent Technologies, Inc. 2009
• Band power marker compensates for signal spread
• Longer sweep, average detector reduce variance
![Page 16: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/16.jpg)
Detector Types, Averaging Scales Understanding Enough
• Processing, Data Reduction Always Done
• Signal Characteristics Affect Results
• Basic Settings (RBW, Span, Sweep Time) Affect Results, Even With Same Detector
© Agilent Technologies, Inc. 2009
Affect Results, Even With Same Detector
• Know Enough To:
– Use your knowledge of your signal
– Understand results, know what you are getting
– Make good measurements
– Identify signal problems, measurement errors
![Page 17: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/17.jpg)
The IF Envelope Detector
Envelope Detector
Time
VoltsIF Signal
© Agilent Technologies, Inc. 2009
• Input is the IF Signal (RBW), Output is Baseband
• Output Voltage is Proportional to Input Envelope
• Response is Different for CW Signals and Noise
![Page 18: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/18.jpg)
Display DetectorsPeak, Negative Peak, Sample
Volts
Peak
Sample
Display points or buckets
© Agilent Technologies, Inc. 2009
• Input is the IF Envelope Detector Output (log or linear scaled)
• Peak and Neg Peak Detectors Bias the Output Value, Perform Some Data Reduction
Time
Neg Peak
![Page 19: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/19.jpg)
Averaging Detectors
Volts
Average valuefor this bucket
Samples of envelope detector output
© Agilent Technologies, Inc. 2009
• Each Output Represents 2 or More IF Envelope Values
• Reduces Variance Faster Than Single-Value Detectors; Reduce Variance Further by Increasing Sweep Time
• Good Choice for Channel Power or ACP Measurements
Time
![Page 20: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/20.jpg)
Signal-and-Noise Detectors
Time
Volts
© Agilent Technologies, Inc. 2009
• Better Represent Combination of Signals and Noise
• Show Distribution of Noise and Noise-Like Signals
• Show Discrete Signals as a Single Value
• More Intuitive, but not Required for Accurate Measurements
![Page 21: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/21.jpg)
Biased & Unbiased Detectors
• Sample Detector-Unbiased
– Data discarded, not good at reducing variance
• Average Detector-Unbiased
– Efficient for reducing variance
• Peak, Negative Peak Detectors-Biased
© Agilent Technologies, Inc. 2009
• Peak, Negative Peak Detectors-Biased
– Bias is desirable, special-purpose use
• “Normal” Detector
– Bias depending on signals, implementation
– Designed for combined signals and noise
![Page 22: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/22.jpg)
Effects of Slower Sweep Speed
• Normal Detector: Wider band of results
• Average Detector: Narrower band of results
– Variance decreases with square root of number of independent samples in average
© Agilent Technologies, Inc. 2009
of independent samples in average
• Sample Detector: Results band unchanged
• Peak Detector: Band is narrower, higher
• Neg Peak Detector: Band is narrower, lower
![Page 23: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/23.jpg)
Averaging Error ExampleAveraging Power
• Average of the Log is Not Equal to a Log of the
0 dBm
6 dBm
1 mW
4 mW Average of the log of the power = 3 dBm(0 dBm + 6 dBm)/2
Log of the signal’s average power = 3.98 dBm(1 mW + 4 mW)/2 = 2.5 mW
© Agilent Technologies, Inc. 2009
• Average of the Log is Not Equal to a Log of the Average
• Narrow VBW or Trace Averaging Performs an Average of the Log, an Error in Measuring Time-Varying Signals
• Instead, Average the Power or Report the RMS Value of the Voltage of the Signal
![Page 24: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/24.jpg)
Log Scale Bias for CW Meas.Log scale better for single CW signal near noise
© Agilent Technologies, Inc. 2009
Video BW filtering or averaging dB
See Agilent AN1303
![Page 25: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/25.jpg)
Detectors/Averaging Example
© Agilent Technologies, Inc. 2009
• Default: normal detector, log-power averaging
![Page 26: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/26.jpg)
Detectors/Averaging Example
© Agilent Technologies, Inc. 2009
• Normal detector, log-power avg. longer sweep
![Page 27: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/27.jpg)
Detectors/Averaging Example
© Agilent Technologies, Inc. 2009
• Average detector, RMS power
![Page 28: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/28.jpg)
Detectors/Averaging Example
© Agilent Technologies, Inc. 2009
• Average detector, RMS power, longer sweep
![Page 29: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/29.jpg)
Reducing Measurement Noise
• Filtering
– RBW (not VBW or averaging)
– Preselection
– External
© Agilent Technologies, Inc. 2009
– External
• Time or Synchronous Averaging
• Subtraction
• Best Practices
![Page 30: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/30.jpg)
Costs of Sensitivity
• Time
– Sweep time (narrower RBW, VBW)
– Averaging
• Money
© Agilent Technologies, Inc. 2009
• Money
– Better instruments
– Better cables, connectors
– Preamplifiers
– External filtering
![Page 31: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/31.jpg)
Traditional Approaches
• Narrower RBW
– Reduces noise floor (ENBW)
– May reduce measurement variance
• Video Filtering, Trace Averaging
– Does not reduce noise floor
© Agilent Technologies, Inc. 2009
– Does not reduce noise floor
– Reduces measurement variance
– Average of log of meas, may distort results
– Effect varies with signal statistics
• Detector Choices
– Select for accuracy, fast reduction of variance
![Page 32: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/32.jpg)
Traditional Approaches (cont)
• Improve Measurement Noise Figure
– Better connection to DUT
– Preamplification - internal
– Preamplification-external
– Preselector (internal YIG-tuned filter)
© Agilent Technologies, Inc. 2009
– Preselector (internal YIG-tuned filter)
– External filtering
• Intelligent Operator Filtering
– “If it looks good it is good” (?)
• Define Results, Objectives Better
– Account correctly for noise effects
![Page 33: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/33.jpg)
Cables, Connectors, AdaptersImportance Increases with Frequency
• Shortest, Best Connection to Circuit
– Shorter cable, move analyzer or DUT
– Choose best cable
– Minimize adapters
© Agilent Technologies, Inc. 2009
– Minimize adapters
– Cable/connector hygiene
– Periodically check cable assemblies
• Visual inspection is not enough
• 1 dB Gained/Lost = 1 dB Noise Figure
![Page 34: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/34.jpg)
Example: Different Cables
© Agilent Technologies, Inc. 2009
• 0.9 m and 1.6 m cables
• Clue in signal/noise ratios
![Page 35: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/35.jpg)
Preamplifiers
• NF of Preamp must be Less than NF of Analyzer
• Benefits
– Improved sensitivity
– May be built-in, internally switchable
– Gain may be included in analyzer calibration
© Agilent Technologies, Inc. 2009
– Gain may be included in analyzer calibration
• Drawbacks
– Best when input is small signals only
– May limit TOI, increase distortion
– May have limited bandwidth
– Calibration less convenient if external
![Page 36: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/36.jpg)
Less-Common Approaches
• Measure and Subtract the Noise Power
• Vector Average to Remove Noise
– Coherent (time-triggered) averaging
– Requires repeated signal, triggering
QQQQ
© Agilent Technologies, Inc. 2009
Vector average of repeated, synchronized
measurementsconverges to thenoiseless value
QQQQ
IIII
![Page 37: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/37.jpg)
CW Signal Measured Near Analyzer Noise Floor
Actual S/N
DisplayedS/N
ApparentSignal
© Agilent Technologies, Inc. 2009
Actual S/NS/N
CW Signal
Amplitude & Frequency Axes Expanded
![Page 38: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/38.jpg)
Noise Floor Subtraction
PobsS+N = PobsN + PS
PS = PobsS+N − PobsN
© Agilent Technologies, Inc. 2009
• Analyzer noise adds incoherently to any signal to be measured
• Power calculations are performed on a linear power scale (watts, not dBm) and results typically are shown in dBm
![Page 39: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/39.jpg)
Plotting Ps on a dB Scale
© Agilent Technologies, Inc. 2009
Variance highly multipliedas observed signalapproaches noise level
![Page 40: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/40.jpg)
Noise-Like Signal and Noise
© Agilent Technologies, Inc. 2009
• 1 MSymbol/sec QPSK, 1.9 GHz
• Signal accurately measured, but noise biased higher by analyzer noise power (no NFE)
• Average detector, slower sweep to measure signal & noise, reduce variance
![Page 41: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/41.jpg)
Result from Noise SubtractionImplemented in the Agilent PXA Signal Analyzer
© Agilent Technologies, Inc. 2009
• Blue trace shows more accurate measurement due to removal of analyzer noise power
• Note increased variance of result
![Page 42: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/42.jpg)
Analyzer Noise Floor without NFE
© Agilent Technologies, Inc. 2009
• Source switched off, pink trace shows analyzer noise level, no NFE
• Other measurement conditions unchanged
• PXA DANL (pink) adds to source power (blue) for first meas. result (yellow)
• Note that noise level variance (pink trace) is smaller without NFE
![Page 43: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/43.jpg)
Analyzer Noise Floor with NFE
© Agilent Technologies, Inc. 2009
• Source still off, green trace shows analyzer noise level with NFE
• Other measurement conditions unchanged
• Note high variance result from subtraction of small, noisy numbers
• Analyzer DANL now far enough below source for minimal(0.2 - 0.4 dB) error
![Page 44: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/44.jpg)
A Closer Look
© Agilent Technologies, Inc. 2009
• Pink trace adds to blue trace; result is yellow trace (NFE not used)
• Green trace is included in blue trace but resulting error very small
![Page 45: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/45.jpg)
Noise Power Subtraction vs. Noise Cancellation or Removal
• Power Subtraction Useful for Spectrum & Power (channel power, ACPR) Meas.
• Noise Power from Phase Noise can be Removed
© Agilent Technologies, Inc. 2009
Removed
• Improving Demodulation (EVM, RCE, MER, all vector errors) Requires Lowering Inherent Noise or Vector Subtraction
![Page 46: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/46.jpg)
Conditions and Requirements for Effective Noise Subtraction
• Signal and Analyzer Noise Uncorrelated
• RMS Detection (calculations simpler, more accurate)
• Extensive Averaging (to reduce signal variance)
• Averaging on Power Scale (even if results
© Agilent Technologies, Inc. 2009
• Averaging on Power Scale (even if results displayed on a log scale)
• Need– Method to remove input signal & measure noise
-or-
– Prior knowledge of analyzer noise floor under all measurement conditions
![Page 47: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/47.jpg)
Signal Analyzer Low-Band
© Agilent Technologies, Inc. 2009
• Noise Contribution of Analyzer Elements must be Measured and/or Accurately Modeled for all Operating Settings
![Page 48: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/48.jpg)
Signal Type vs. EffectivenessSignal/Noise Variation with RBW
SNR(dB)
Noise-like
Real signals can be one type or
combination
© Agilent Technologies, Inc. 2009
• Amplitude envelope vs. time
• Best RBW is one matched to signal
• Best ability to separate analyzer noise from signal
RBW (log)
![Page 49: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/49.jpg)
Noise Floor EnhancementCW Example
© Agilent Technologies, Inc. 2009
• 95% confidence interval, ±2 dB tolerance
• 3.5 dB improvement for CW signal
![Page 50: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/50.jpg)
Noise Floor EnhancementNoise-Like Signal Example
© Agilent Technologies, Inc. 2009
• 95% confidence interval, ±1 dB tolerance
• 9.1 dB improvement for noise-like signal
![Page 51: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/51.jpg)
Noise Floor EnhancementPulsed RF Example
© Agilent Technologies, Inc. 2009
• 95% confidence interval, ±3 dB tolerance
• 10.8 dB improvement for pulsed-RF signal
![Page 52: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/52.jpg)
Subtracting Broadband Noise in a Phase Noise Measurement
B Marker 4 430 590.381 Hz -135.775 dBm/Hz
TRACE A: F1 PSD1/K1
A Marker 4 430 590.381 Hz -123.6 dBm/Hz
-50
dBm/Hz
LogMag
© Agilent Technologies, Inc. 2009
Center: 4.420777881 MHz Span: 50 kHz
-150
dBm/Hz
10
dB
/div
![Page 53: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/53.jpg)
Subtracting Analyzer Phase Noise Using Very Clean Source as Phase Noise Power Reference
© Agilent Technologies, Inc. 2009
Raw measurement Reference measurement Result after subtraction
Noise Subtraction Calculations Increase Variance of the Result
![Page 54: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/54.jpg)
Better Reference Oscillator and/or Higher Performance Analyzer
Fc = 1.8 GHz
© Agilent Technologies, Inc. 2009
![Page 55: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/55.jpg)
Microwave or High Band Architecture Tradeoffs
• Microwave or High Band Section
– Path switching for low band
– Path switching for microwave preamplifier
© Agilent Technologies, Inc. 2009
– Other switching or path losses
• Alternate “Low Noise Path”
– Optimized for high band
– No microwave preamplifier
![Page 56: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/56.jpg)
Low Noise Path
To µWConverters
© Agilent Technologies, Inc. 2009
Converters
To Low Band
![Page 57: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/57.jpg)
Alternate “Low Noise Path”
© Agilent Technologies, Inc. 2009
![Page 58: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/58.jpg)
Combining Noise Floor Extension and Low Noise Path
© Agilent Technologies, Inc. 2009
• 3.6 - 26.5 GHz, preamplifier off
• Low noise path incompatible with preamp
![Page 59: Conquering Noise for Accurate RF and Microwave Signal](https://reader034.vdocuments.mx/reader034/viewer/2022051008/58a2e7b11a28ab41128b8c68/html5/thumbnails/59.jpg)
Resources• “Noise Floor Extension: Improving the Dynamic Range of Spectrum Analysis” Joe Gorin (to
be published soon)
• “Improving the Dynamic Range of Spectrum Measurements by Compensating for the Effects of Phase Noise and Thermal Noise” Wing Mar & Ben Zarlingo, EuMW 2001
• “Spectrum Analyzer Detectors & Averaging for Wireless Measurements” Joe Gorin & Ben Zarlingo http://www.agilent.com/cm/wireless/pdf/detector_averaging.pdf
• “Spectrum Analyzer Measurements and Noise” Joe Gorin, Agilent Technologies Application Note 1303, literature number 5966-4008E
• “Spectrum & Network Measurements” Robert Witte, Noble Publishing 2001 ISBN 978-1884932168
• “Optimizing RF and Microwave Spectrum Analyzer Dynamic Range” Agilent application note
© Agilent Technologies, Inc. 2009
• “Optimizing RF and Microwave Spectrum Analyzer Dynamic Range” Agilent application note AN-1315 literature number 5968-4545E
• “Electronic Test Instruments: Analog and Digital Measurements” Robert Witte, 2002 Prentice Hall ISBN 978-0130668301
• “The Art of Measurement: Theory and Practice” Ronald Potter, 1999 Prentice Hall ISBN 978-0130261748
• “Spectrum Analysis Basics” Agilent application note AN-150 literature number 5952-0292EN
• “Accurate Measurement of Signals Close to the Noise Floor of a Spectrum Analyzer” Andrew Mouthrup and Michael Muha, IEEE Transactions on Microwave Theory and Techniques, Vol. 39, No. 11, November 1991, pg. 1882 – 1885.