Glossary

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

A

ATTENUATOR, (ATTN)
In the context of spectral analysis this term mostly denotes an input attenuator. "Attenuator Setting" describes the setting of the "input attenuator".

Decreasing the attenuation by 10 dB results in a
noise floor decreased by 10 dB

AUTO, AUTOMATIC
Automatic coupling of device parameters: AUTO. In order to obtain optimum measurement results the measurement parameters need to be set in a specific relation to one another.

• Resolution bandwidth (RBW), video bandwidth (VBW) and sweep time (SWT) coupled to frequency span (SPAN).
  Background: When parameters are changed, care must be taken not to reduce the ratios of SWT x RBW2/SPAN and SWT x RBW x VBW/SPAN, as this would increase the measurement error.
• Input attenuator (ATTN) coupled to reference level (REF).

This feature allows exact measurements to be made in the majority of spectrum analysis applications within a relatively short time (sweep time) under optimum drive level conditions (mixer level). Automatic coupling should be disabled for special applications such as sensitivity or intermodulation measurements.

AVERAGE NOISE LEVEL
> Noise floor

AVERAGING
A method of smoothing the results of several measurements by combining them and taking the average. Two methods are used:

Noise level with analog averaging:
a lower VBW results in a lower noise figure
(VBW = RBW, VBW = RBW/10, VBW = RBW/100, VBW = RBW/1000)

• Analog (or time-constant) averaging, which is the weighting of the level display by means of the video filter time constant.
  Background: This averaging will occur as soon as VBW is set to a smaller value than RBW. In order not to increase the measurement error, the sweep time must be increased.
• Digital averaging, which is the formation of the average of several sweeps from the stored level values for each measurement frequency.
  Background: This method does not affect the sweep time. However, this measurement principle bears the risk of major measurement errors.

Spectrum without averaging

Spectrum with digital averaging

^ top

B

BAND
Frequency band

BRIDGE
Directional bridge

BROADBAND DETECTION
A method of detection used in scalar network analysis which makes use of broadband power level measurement.

Advantages:

• Simple, inexpensive and robust equipment
• Frequency offset between input and output of device under test possible (e.g. for testing mixer stages)
• Fast

Disadvantages:

• Dynamic range limited to approximately 60 dB
• Measurement errors caused by generator harmonics (e.g. when measuring high- or lowpass filters).

BROADBAND DISPLAY
By measurement of pulse spectra a spectrum analyzer is said to operate in this broadband display mode when the resolution bandwidth is greater than the spacing between the individual spectral lines (pulse repetition frequency) of the pulse spectrum.
Background: With this method the dynamic properties of the selection filter, the video filter and the rectifier will affect the measurement result so that a correction factor has to be taken into consideration (impulse bandwidth).

^ top

C

CALIBRATION
Adjustment of the spectrum analyzer to ensure maximum accuracy of level and frequency measurements. The RF input section, IF amplifier and the resolution filters are the main sources of measurement error and should be calibrated before making measurements. Many spectrum analyzers are equipped with automatic calibration facilities (autocal) which make use of builtin reference sources and thus frequently do not need any calibration signal. The Willtek 9101 with its innovating digital IF has no need for calibration.

CENTER FREQUENCY
The frequency which corresponds to the center point of the frequency span.

COMPRESSION
Nonlinear behavior caused by overdriving (overloading) the analyzer. This results in measurement error, since the output level from the input or IF stages can no longer linearly track the input signal level. The 1 dB compression point is normally specified; this is the point at which the analyzer indicates a level of 1 dB less than it would if it still worked in the linear mode. Background: The analyzer shows a difference of just 9 dB between a signal whose level corresponds to the "1 dB compression point" and a signal which is smaller by 10 dB.

Compression

CONTINUOUS WAVE
This term describes a signal which is continuous as opposed to a pulse signal. This is the most common type of signal encountered in analog telecommunications, particularly with respect to amplitude (AM) and frequency modulation (FM) techniques.

CROSSTALK
Unwanted coupling of signals between the various sections of the device circuitry. In network analysis, this term is generally used to describe internal and external coupling between the generator and detector of the analyzer.

^ top

D

DECIBEL, (dB)
If two signals have the power level P1 and P2, their level difference is 10 x log(P2/P1) dB. An ideal way to express amplification and attenuation is just to offset the power levels P1 and P2 at the input and output of a quadripole against one another.

Note: If you calculate with voltages (U1 and U2) and not with power levels, the level difference is 20 x log(U2/U1).

It is also possible to indicate absolute levels comparing the actual level to a reference level. Variants on the decibel (dB) measure used in spectrum and network analysis are:

• dB
  spectrum analysis: absolute voltage level referred to 775 mV
  network analysis: level difference (gain/loss) referred to the reference level (0 dB = 775 mV)
• dBµV
  absolute voltage level referred to 1 µV (0 dBµV = 1 µV)
• dBm
  absolute power level referred to 1 mW (0 dBm = 1 mW)
• dBc
  level difference referred to carrier level (c)
• dBc/Hz
  noise level difference referred to carrier level calculated for a measurement bandwidth of 1 Hz.

DIRECTIONAL BRIDGE
Circuit based on the principle of the Wheatstone Bridge; used for measurement of reflection coefficient. Can also be used for low-reactive injection of signals.

DIRECTIVITY
A measure for the performance of the directional bridge or directional coupler. It specifies its ability to separate the forward and backward (reflected) components of a signal. A finite directivity, D, gives the following measurement error for the reflection coefficient r:

|r_measured| = |r ± D|

DISCRETE SPURIOUS RESPONSES
Spurious products

DISTANCE TO FAULT (DTF)

DTF measurements are a performance verification and failure analysis tool for any kind of transmission line or antenna. It uses the Frequency Domain Reflectometry. DTF measurements indicate either the return loss or the standing wave ratio versus distance. The velocity factor of the cable measured must be known. The following formula shows the mathematic coherence:

Dmax = (f * c * V ) / 2

where Dmax = maximum distance , c = speed of light, f = frequency of the standing wave, V = velocity factor

The maximum distance to be measured should also be known in advance. The entered distance affects the frequency range in which the measurement is performed; the formula below shows the relationship between the maximum distance measured and the frequency span being used.

D = (FACTOR * Vrel * Xpoints) / (F2 - F1)

where D = distance to fault, FACTOR = system specific factor, Vrel = relative propagation velocity, Xpoints = number of frequency data points, F2 = stop frequency , F1 = start frequency

The return loss measurement (in dB) shows
a problem in a distance of 21.45 m.

DRIFT
Relative local oscillator (LO) frequency change measured over a fixed period of from one hour to one year. Caused by thermal response and aging effects. Drift influences frequency accuracy and reproducibility at narrow resolution bandwidths. Innovative designs such as the Willtek 9101 Handheld Spectrum Analyzer with LO synthesizer and digital IF reduce the drift into the ppm range.

DRIVE LEVEL
Mixer level

DYNAMIC ACCURACY
Also known as amplitude accuracy, scale fidelity; a measure of the accuracy of measured levels which are not equal to the reference level. Inaccuracies are caused by nonlinear behavior of variious analyzer circuits, in particular:

• Compression effects in the mixer and IF stages
• Errors in the logarithmic amplifier characteristic
• Detector nonlinearities
• Calibration uncertainty
  Innovative designs with digital IF eliminate errors in logarithmic amplifier characteristic and detector linearity and part of calibration uncertainty as well.

DYNAMIC RANGE
Network analysis: The greatest difference in level which can be measured to a given degree of accuracy; normally the level difference between the analyzer noise floor and the onset of compression.

Spectrum analysis: The greatest difference in level between two signals applied simultaneously to the analyzer input which can be measured to a given degree of accuracy. Three main interpretations of the term are distinguished, depending on the source of the measurement error; they cannot be interconverted.

The following diagram illustrates the relationships between the dynamic ranges:

The three types of dynamic range are

• Intermodulation-free dynamic range (A)
  This describes the maximum difference between the noise floor and the level at which the intermodulation of the analyzer just do not exceed the level of the noise floor. Other spurious responses than intermodulation are ignored.
• Interference-free dynamic range (B)
  This describes the maximum difference between the noise floor and the level at which the spurious responses of the analyzer (intermodulation, distortion) just do not exceed the level of the noise floor.
• 1 dB compression dynamic range (C)
  This describes the maximum level difference between the noise floor and the level at which the measurement error of the analyzer due to compression is 1 dB. The spurious responses of the analyzer are ignored.

There is an optimum analyzer drive level (mixer level) for each set of conditions which maximizes each type of dynamic range.

The automatic setting facilities of modern analyzers are normally programmed to give the best possible interference-free dynamic range. To achieve optimum conditions for the other types of dynamic range, the analyzer may be set manually.

^ top

E

ERROR LIMITS
The limits for possible measurement errors.

The frequency error limits are basically dependent on the following factors:

• Accuracy of the crystal-controlled timebase
• Long-term stability (drift)

The level error limits are basically dependent on the following factors:

• Device adjustment and calibration
• Thermal characteristics
• Aging
• Measurement port mismatch
• Frequency response

EXTERNAL ATTENUATOR, EXTERNAL ATTENUATION
This parameter offers advanced spectral analyzers the possibility to correct the measurement results automatically by the attenuation of a preconnected adapter (attenuator, probe, coupler, preamplifier, impedance matching pads, etc.). Example: When displaying the amplitude results of the markers the corresponding value will be automatically taken into consideration.

^ top

F

FREQUENCY ACCURACY
Drift
Modern analyzers are tuned by a synthesized LO. All frequencies are derived from one crystal oscillator the so-called reference oscillator with frequency accuracy in the lower ppm range and even lower.

FREQUENCY BAND
Depending on the technical concept the frequency measurement range of the analyzer can be subdivided into "frequency bands". This applies in particular to units having a high upper frequency limit (> 26.5 GHz). This concept can mostly be found with older units. Modern analyzers practically use bands only with harmonic mixers. Example: With a harmonic mixer the frequency bands are produced by mixing each input signal with one harmonic of the local oscillator.
Only down conversion is used; IF conversion is according to the following relationship:

f_IF = |n x f_LO - f_RF|
(n > 1, f_IF < f_RF, f_IF < f_LO)

FREQUENCY DOMAIN REFLECTOMETRY (FDR)
FDR is the measurement of signal reflections through a device under test (DUT) across a certain frequency range. A portion of the test signal transmitted into the DUT will be reflected back to the transmitter if either the impedance inside the DUT or the load on the end of the DUT does not match the correct impedance (typically 50 ohms).

FREQUENCY RESPONSE
Describes the frequency-dependant level error over a given frequency range. Any variation is undesirable.

FREQUENCY RESPONSE CORRECTION
Automatic correction of the error in the measured level due to the frequency response of the analyzer. Frequency response correction often occurs only at the set center frequency (CENT). High-performing analyzers offer frequency response correction which covers the whole frequency range displayed.

FREQUENCY SPAN
Span

FULL SPAN
A sweep across the entire frequency range of the analyzer including automatic switching between the frequency bands.

FUNDAMENTAL MIXER
A mixer stage which only makes use of the local oscillator fundamental frequency f_LO for heterodyning. IF conversion is according to the following relationship:

f_IF = |f_LO - f_RF|

If f_IF, f_LO are > f_RF => up conversion
If f_IF, f_LO are < f_RF => down conversion

^ top

G

GAIN COMPRESSION
Compression

GAUSSIAN FILTER
A filter, the passband characteristic of which is described by a Gaussian curve ("bell-curve"). This type of filter has the following advantages when used as a selection filter:

• Short settling time
• Minimal signal distortion

Gauss-shaped filter curve centered at 4 kHz

^ top

H

HARMONIC(S)
A periodic signal of frequency f can be described as the sum of sinusoidal signals of frequencies f, 2 x f, 3 x f etc. The sinusoidal signal of frequency f is called fundamental wave. The sinusoidal signals of frequencies 2 x f, 3 x f etc. are called harmonics.

HARMONIC DISTORTION
A real active circuit will always generate harmonics of the input signal. The level of these harmonic products is a growing function of the input signal level. The 2nd and 3rd order harmonic distortion ratios of the input stage of a spectrum analyzer are determining factors for the dynamic range of the analyzer.

HARMONIC MIXER
A mixer which utilizes the 2nd order harmonic distortion of the local oscillator frequency to convert the RF input signal to the IF level:

fIF = |f_RF - n x f_LO| where n = 2, 3, 4 etc. The harmonics are either present in the LO signal or are generated in the mixer from the fundamental of the LO signal.

^ top

I

IF
Intermediate Frequency

IF INTERFERENCE
Spurious responses of the analyzer caused by the use of superheterodyning. Crosstalk between circuit modules may lead to a signal value being displayed when an input signal with a frequency equal to one of the intermediate frequencies is applied, even if the analyzer is not tuned to this frequency.

IF REJECTION
A measure of the degree to which IF interference is suppressed by the analyzer.

IMAGE FREQUENCY
A displayed interference signal caused by ambiguity in mixing. Since conversion to the IF level takes place according to the relationship:

f_IF = |f_RF - f_LO|

There will be exactly two receive frequencies which generate the same IF for a fixed setting of the local oscillator frequency:

f₁ = |f_LO - f_IF|
f₂ = |f_LO + f_IF|

f₂ is called „image frequency“.

If the intermediate frequency is above the maximum RF signal frequency, the image frequency can be rejected by first passing the signal through a low-pass filter. If the intermediate frequency is below that of the RF signal, then the image frequency can only be removed by using a preselector.

IMPEDANCE
Usually refers to the internal resistance of the analyzer at the input port. Z = 50 Ω is most often used as an impedance value in the RF sector. However, also Z = 75 Ω can be found (e.g. TV). If measurements are made on a system whose impedance differs from the analyzer’s impedance, an impedance transformer or impedance matching adapter is to be used.
Referring this value to a characteristic impedance value gives the !reflection coefficient or the standing wave ratio.

Note: If connectors and cables of different impedance values are connected with one another, severe measurement errors can result and the connectors might even be destroyed! (Example: N connectors 50 Ω & 75 Ω differ considerably in the diameter of their center conductor.)

IMPULSE BANDWIDTH
The bandwidth of a hypothetical filter with rectangular passband characteristic having the same area as the resolution filter under the characteristic curve (voltage linear) (i.e. the same voltage transfer function). For a Gaussian filter, the following relationship applies:

IBW = 1.5 RBW

(RBW = 3 dB resolution bandwidth)

Impulse bandwidth of a Gaussian filter

INPUT ATTENUATOR
A variable attenuator placed between the input and the preamplifier stage or first mixer stage of the analyzer. For automatic operation, the attenuator is normally set to give maximum interference-free dynamic range. It can be set manually to cope with other requirements (sensitivity, intermodulation). A manually set input attenuator has the advantage of allowing the operation of the unit with a low-distortion or low-noise setting. In this way the optimum load characteristics of the unit can be selected for the measurement task. At the same time it is possible to e.g. assign distortion-based spectral portions to the DUT or the analyzer by changing the attenuation value thus avoiding any erroneous measurements.

INTERCEPT POINT
Descriptive measure for intermodulation: a theoretical (extrapolated) level at which an intermodulation product is as high as the fundamental signal. Various orders of intercept point are distinguished; the most important factor in the dynamic range of a spectrum analyzer is the 3rd order intercept point (TOI). At the TOI, the 3rd order intermodulation products have the same level value as the fundamental of the test signal (this is a theoretical value, since compression already occurs at much lower values).
If the TOI point is known, it is possible to calculate the 3rd order signal to intermodulation noise ratio (a) for any mixer level.

a_d3/dBc = 2 x (IP₃ - L_in)

where:

a_d3 = 3rd order signal-to-intermodulation noise ratio
IP₃ = 3rd order intercept point (TOI)
L_in = input level to 1st mixer

Determining the intercept point

INTERMEDIATE FREQUENCY (IF)
The spectrum analyzer input signal is converted and selected into fixed frequencies, known as the intermediate frequencies, by heterodyning (mixing) with the aid of a tunable oscillator signal (local oscillator).
Converter and filter components used for converting or signal filtering are also called IF.

INTERMODULATION
If a non-linear circuit is fed simultaneously with several different signals of frequencies f₁, f₂, f₃ ... etc., the output signal will contain signals of frequencies given by the relationship n x f₁, ñ m x f_k, in addition to the original frequencies.
In the datasheet of a spectral analyzer under Intermodulation only the special case of a 3rd order intermodulation of a 2-tone signal is indicated.
The 3rd order intermodulation products of the 1st mixer stage and the IF stages are of particular importance in spectrum analysis as they govern the measurement error and dynamic range of the instrument.

• Input signal:
  f₁, f₂

• Output signal:
  f₁, f₂ original signal
  2f₁ – f₂, 2f₂ – f₁ 3rd order intermodulation products

The amplitude varies in 3rd order proportion, i.e. if the input signal is reduced by 10 dB, the intermodulation product amplitude is reduced by 30 dB (signal-to-intermodulation ratio increases by 20 dB).

Example of third order intermodulation products

INTRINSIC NOISE
Spurious responses, noise floor

^ top

J

^ top

K

^ top

L

LO
Local Oscillator

LOCAL OSCILLATOR
The oscillator generating the heterodyning frequency for the mixer stage(s). The abbreviation LO is also used to indicate the frequency of the heterodyning signal (e.g. 1st LO). In fundamental mixers, the fundamental of the LO frequency is used; in harmonic mixers, the LO harmonics are used for mixing.

LONG-TERM STABILITY
Drift

^ top

M

MISMATCH
Describes nonideal coupling between two circuits. Not all of the available signal is transferred, part of the signal is reflected at the connector, leading to a measurement error.
The degree of mismatch is specified by the reflection coefficient and the standing wave ratio. The most critical point in spectrum analysis is the connection between the device under test and the input of the analyzer; in network analysis, the connection between the tracking generator output and the device under test is also important.
By means of an external high-precision low-reflection attenuator mismatching errors can frequently be corrected to such an extent that the error contribution resulting from the mismatching can largely be ignored.

MIXER LEVEL
Describes the input level to the 1st mixer of the analyzer. The mixer level is dependent on the input signal level (L_in), the setting of the input attenuator (ATTN) and the insertion loss (a₀) of the input stage or external devices like attenuators:

L_mixer (dBm) = L_in (dBm) — ATTN (dB) — a₀ (dB)

The mixer level influences the harmonic distortion, intermodulation, signal-to-noise ratio (noise floor) and dynamic range of the analyzer. It can be adjusted to optimize these various factors.

Simplification:

L_mixer, nominal = L_in — ATTN

^ top

N

NARROWBAND DISPLAY
A method of displaying a pulse spectrum where the resolution bandwidth is less than the spacing between the lines (pulse repetition frequency) of the pulse spectrum.

NETWORK ANALYSIS
Measurement technique used to characterize the behavior of electrical networks with the aid of a generator and receiver. The most common measurands in RF engineering are the twoport network parameters, known as the S-parameters.

NOISE BANDWIDTH (NBW)
More accurately: equivalent noise bandwidth. Describes the equivalent bandwidth of a filter for broadband (white) noise and is the bandwidth of a rectangular characteristic filter having the same power transfer function as the selection filter (same area under the curve). The relationship between NBW and the resolution bandwidth (RBW) of a Gaussian filter is approximately:

NBW = 1.2 x RBW

NOISE DISTORTION
This term describes the non-linear behavior of circuits or devices when driven with a broadband noise signal. The resultant signal includes both harmonic distortions and intermodulation products of the various components of the noise signal. The distortion noise is the determining factor for the noise power ratio (NPR).

NOISE FIGURE
A measure of the deterioration of the signal-to-noise ratio as a result of generated internally spurious products. Expressed in linear (F) or logarithmic (F/dB) form. The noise figure is directly related to the noise floor:

L_n = L_therm x NBW x F
L_n/dBm = L_therm/dBm + 10 log (NBW/Hz) + F/dB

where:

L_n = noise floor (level)
L_therm = thermal noise level (–174 dBm pro Hz)
NBW = noise bandwidth
F = noise figure

The noise figure is thus a useful measure of sensitivity.

NOISE FLOOR
The level of broadband noise produced internally by the receiver stages of the analyzer. The noise floor determines the sensitivity and hences the lower end of the dynamic range. The noise floor value is always quoted referred to the input port of the analyzer.

NOISE POWER RATIO (NPR)
A parameter to qualify the behavior in presence of an evenly distributed broadband load: the ratio between the density of a noise input signal to the analyzer and the density of the spurious responses of the analyzer ( û ~ noise floor, û ~ noise distortion, intermodulation noise).
In order to measure the NPR, white noise is injected into the analyzer. Appropriate band limitation filters limit the noise bandwidth to the desired frequency range.
An additionally connected blocking filter (band-stop filter) suppresses the noise in a narrow spectral range. The NPR of the DUT (here: the analyzer) results from the ratio between the level in the noise band and the level measured in the blocking range of the noise squelch.

Measuring the noise power ratio of the analyzer

NORMALIZATION
Digital correction of frequency response error used in network analysis. The measured values of various standards are stored and used as references for the actual measurement.

^ top

O

^ top

P

PHASE NOISE
Single sideband noise

Sideband noise is amplitude noise plus phase noise

PRESELECTOR
Term applied to filtering of the input signal before passing it to the input stages of the analyzer. In case of down conversion and harmonic mixing this ensures the removal of unwanted image frequencies and mixer products of LO harmonics. A preselector is also used to limit the band of the input signal in order to increase measurement dynamics (e.g. for EMI measurement).

^ top

Q

^ top

R

REFERENCE
Value to which all level measurements are referred. A measurement at the reference level has maximum accuracy, since the scaling error (dynamic accuracy) disappears at this point.

REFLECTION COEFFICIENT
A measure of the deviation of an impedance Z from the specified characteristic impedance Z. Can be expressed as a linear (r) or logarithmic (~) quantity. Z û Z,

r = Re(r) + j x Im(r) = Z + Z,

RESIDUAL FM
Measure of the short term stability of the displayed frequency as determined by the instability of the local oscillator. Expressed in terms of peak-to-peak frequency deviation. The residual FM determines the minimum resolution bandwidth which can be used, since at smaller bandwidths, the residual FM becomes visible.

RESIDUAL RESPONSES
Spurious responses of the analyzer in the absence of an input signal, usually given for the input attenuator setting of 0 dB with the input terminated with the characteristic impedance.

RESOLUTION BANDWIDTH RBW
Normally the 3 dB (occasionally the 6 dB) bandwidth of the IF filter used for selecting the signal to be measured. The RBW describes the ability of the spectrum analyzer to discriminate between adjacent signals of similar amplitude. Only signals spaced at a frequency of more than the RBW can be discriminated from one another.
For measurements on signals with closely spaced components, such as two-tone signals or sideband noise, an analyzer with a narrow RBW is required. For measurement of broadband signals such as TV carriers or pulse spectra, a wide RBW is necessary. The RBW indirectly affects the

• noise floor level
• sensitivity
• dynamic range
• sweep time

as it determines the equivalent noise bandwidth of the analyzer.
As described above the correct selection of resolution bandwidth may have a critical influence on the measurement results.

Narrow filters:

• increase the measurement sensitivity of the receiver by reducing noise power.
• allow signals with a small frequency spacing to be separately displayed on the screen.
• increase the sweep time due to significantly longer settling times as compared with broadband filters.
  (A reduction of the resolution bandwidth by a factor of 3 of the original bandwidth will increase the sweep time by a factor of 32, i.e. by a factor of 9 against a 3 times larger bandwidth).
• are hardly suited for signals with rapidly changing amplitudes (e.g. radar, TV) and might lead to significant amplitude errors with such signals.

High resolution bandwidth = 300 kHz

Low resolution bandwidth = 30 kHz

Broadband filters:

• reduce the measurement sensitivity of the analyzer by increasing the detected noise power of the analyzer.
• do not allow signals with a small frequency spacing to be separately displayed on the screen.
• reduce the sweep time due to the significantly shorter settling times as compared with narrow filters.
• are necessary for signals with rapidly changing amplitudes (e.g. radar, TV) in order to prevent the amplitude errors caused by narrow filters with their possibly too long settling times.

In general it can be said that CW signals, i.e. signals whose amplitude or frequency does not or hardly change during the monitoring period, can be measured using narrow filters. Filters need to be sufficiently narrow in particular when measuring signals with a tight frequency spacing (thumb’s rule: filter bandwidth = 1/10 of the frequency spacing) in order to be able to display the separate signals on the screen.
In case of rapidly changing signals (e.g. TV signals) a sufficiently wide resolution bandwidth (and video bandwidth) needs to be selected in order to prevent amplitude errors (settling errors). When measuring unknown signals the resolution bandwidth and the video bandwidth should be changed in steps and the effects on the measurement results monitored.
Any “drop” in signal amplitudes might indicate the selection of too narrow bandwidths.

^ top

S

SCALE FIDELITY
Dynamic accuracy

SCALAR NETWORK ANALYSIS
Measurement technique used to characterize the transmission and reflection behavior of single or multiport networks. The magnitudes of the transmission factor and the reflection coefficient are measured.

SELECTIVE DETECTION
Method used in scalar network analysis which uses a selective measuring receiver tuned to the measurement frequency. This method is mostly used with spectral analyzers equipped with a tracking generator.

Advantages over broadband detection:

• High dynamic range of more than 100 dB
• Insensitive to tracking generator harmonics.

SENSITIVITY
The noise floor of the analyzer determines the smallest level which can still be measured for a given resolution bandwidth. Signal levels below the noise floor cannot be measured, and signal levels just above the noise floor will be measured inaccurately. The sensitivity indicates the signal level required for a display level 3 dB above the average noise level. The relationship between sensitivity, noise figure and equivalent noise bandwidth is given by:

S/dBm = L_therm/dBm + 10 log (NBW/Hz) + F/dB + 3

where:

S = sensitivity
L_therm = thermal noise floor (–174 dBm)
NBW = noise bandwidth
F/dB = noise figure in dB

SETTLING ERROR
An amplitude error caused by sweeping the set frequency range too rapidly for the selection filter to settle properly (sweep time). In addition, a frequency error occurs, when the SWT has been much too short. In the automatic mode the parameters are set in such a way that this error is not recognizable.

Amplitude and frequency are measured incorrectly with too low SWT

SHAPE FACTOR
Defines the selectivity of a filter. Usually the ratio of the 60 dB to the 3 dB (sometimes 6 dB) bandwidth (RBW) of the selection filter is used here. The shape factor indicates how well the spectrum analyzer can discriminate between closely spaced signals of different amplitudes.

The shape factor characterizes the selection filter

SHORT-TERM STABILITY
Variations in the frequency and amplitude of the local oscillator signal measured over short periods of time of a few seconds or less. These variations may be interpreted as modulations. The short-term stability governs the ability of the analyzer to discriminate between closely spaced signals of different amplitudes and affects the reproducibility of measurements made using „narrow” resolution bandwidths. The short-term stability is expressed in terms of residual FM, residual AM and single sideband noise.

SIDEBAND NOISE
Single sideband noise

SIGNAL TRACK
Automatic adjustment of the center frequency to ensure that the displayed signal is centeredon the screen (i.e. center of sweep range) after each sweep.

SINGLE SIDEBAND NOISE
Describes the short term stability of the local oscillator. Noise sidebands are modulated onto the carrier as a result of nonlinear effects in the LO. The power of these sidebands decreases as the spacing from the carrier increases. A distinction is made between amplitude noise (random variation in level stability) and phase noise (random variation in frequency/phase stability).
Phase noise is the dominant factor, and is specified as the absolute power level of a sideband offset from the carrier by a frequency foff referred to a measurement bandwidth of 1Hz and the carrier power level. It is expressed in dBc/Hz (dB). As single sideband noise is converted into the IF level, it can be used as a measure of the ability of the analyzer to discriminate between closely spaced signals of different amplitudes.

SPAN
Also called frequency span. The frequency range swept by the analyzer during a measurement. The span can usually be set either directly via the Span parameter or indirectly via the parameters „Start frequency“ and „Stop frequency“. When setting the span, care should be taken to select a value which is appropriate for the measurement task. If the span is too large, too long sweep times might result (in particular for small resolution/video bandwidths), while too small a span might prevent the detection of spectral portions which then lie outside the monitored range.

SPECTRAL PURITY
Describes the degree to which the local oscillator signal is free of !spurious products, and hence indicating the accuracy of the displayed input signal. The ratio of the peak carrier level to the spurious responses (intrinsic noise) and single sideband noise is usually quoted.

SPECTRUM ANALYSIS
Measurement technique used for displaying the variation of amplitude and/or power level of a signal over the frequency.

SPURIOUS RESPONSES
Refers to signals occurring within the analyzer and which are displayed, but which are not part of the signal to be measured. Such spurious responses are visible in spectrum analysis as spectral lines at the fundamental or harmonics of the spurious signals, or as sidebands on the test signal. Among others, spurious responses are caused by:

• nonlinear effects in the mixer- and IF stages (harmonic distortion, intermodulation)

• intrinsic noise (noise floor)
• LO single sideband noise
• interference from the a.c. line or switch-mode power supply
• IF interference

In network analysis, the most important spurious response is crosstalk.

STABILITY
Drift, short term stability

STANDARDS
Standard devices used as references. In spectrum analysis, usually a fixed frequency calibration source; in network analysis, a standard short-circuit and a standard open-circuit used for reflection measurements or a standard thru-connection for transmission measurements.

STANDING WAVE RATIO (SWR)
Sometimes called voltage standing wave ratio. Describes the deviation of an impedance from the characteristic impedance Z. The SWR is calculated from the reflection coefficient r.

SWR = (1 + |r|) / (1 – |r|)

SWEEP TIME
The time taken for the analyzer to sweep the entire frequency span set. The sweep time must be chosen to match the resolution bandwidth, video bandwidth and span, since the IF filters have a finite response time. In spectrum analysis, the following basic relationship should be observed (automatic mode):

SWT > K [SPAN/(RBW)²], where K ~ 2

VBW % RBW

SYNTHESIZER
A signal generator which “synthesizes” the output frequency by digital manipulation of a single timebase. The usual operations are multiplication, division or mixing (heterodyning). The advantage of a synthesizer is that each settable frequency has the same accuracy and stability as the timebase. Thus the frequency-related measurement results, e.g. of the markers, will be much more accurate than with units without synthesizer.
Modern spectral analyzers use a synthesizer as local oscillator to increase frequency accuracy.

^ top

T

TIMEBASE
Frequency-reference source or frequency standard from which the synthesizer output signal and therefore the measurement frequency or output frequency of the analyzer is derived. Since the reference frequency and the output frequency are directly related, the stability of the analyzer is directly attributable to the stability of the timebase.

TRACKING
Identical phase and level behavior of various signal paths in network analysis (e.g. reference and test channels).

TRACKING GENERATOR
A swept signal generator which is synchronous with the receive frequency of the spectrum analyzer. This technique allows a spectrum analyzer to be used for scalar network analysis in order to measure e.g. reflection losses of DUTs such as antennas, amplifiers, filters, etc.
At the same time the tracking generator in conjunction with the spectral analyzer and appropriate software allow impedance mismatches in cables to be localized. Such impedance mismatches are caused e.g. by mechanical damage to the cable. This fault localization is frequently called Distance to Fault (DTF) measurement.

TWO TONE INTERMODULATION
Intermodulation

^ top

U

^ top

V

VECTOR NETWORK ANALYSIS
Measurement technique which allows complete characterization of single or multiport networks. The magnitude and phase of the measurand are determined. The transmission function and the reflection coefficient are usually measured in RF engineering.

VIDEO BANDWIDTH (VBW)
The bandwidth of a low-pass filter connected downstream of the detector circuits. This causes the amplitude or power level to be weighted with a time constant to smooth out the noise components in the measured signal (averaging).
The selected video filter affects the sweep time. Due to their longer settling times, narrow filters require longer sweep times than wide filters. However, wide filters do not average (reduce) the noise as much.
When selecting a video filter you should take into account that rapidly changing signals (e.g. TV signals) might not be detected correctly. In such cases a sufficiently large video bandwidth should be selected. When measuring unknown signals the video bandwidth should be changed in steps and the effects on the measurement results monitored.

High video bandwidth

Low video bandwidth

VOLTAGE STANDING WAVE RATIO (VSWR)
Standing wave ratio

^ top

W

^ top

X

^ top

Y

YIG FILTER
A tunable bandpass filter constructed using yttrium iron garnet (YIG) resonators and often used as preselector. Coils placed at right angles to each other around a YIG crystal in a magnetic field are cross-coupled when energized by the electron spin precession frequency. Altering the magnetic field by means of a control current tunes the filter to a particular frequency.

YIG OSCILLATOR
A broadband tunable oscillator based around a YIG resonator used as a local oscillator.

YTTRIUM IRON GARNET (YIG)
A chemical compound, formula Y Fe O, having ferromagnetic properties and used for producing reciprocal and nonreciprocal electronic components. Its uses are based on the precession of the electron spin when placed in a magnetic field (YIG filter).

^ top

Z

ZERO SPAN
An operating mode of the analyzer where the frequency span is zero. This corresponds to a level measurement at a fixed frequency (receiver mode) and hence a time domain display of the (filtered) input signal.

A pulsed signal in zero span mode