HP 54750A
High-Bandwidth
Digitizing Oscilloscope
Product Overview
The HP 54750A is a modular
system designed to meet the
stringent demands of today’s
engineers. Up to four highbandwidth channels and time
domain reflectometry (TDR) are
supported by the mainframe. The
modular design allows for expansion to meet changing needs.
The HP 54750A high-bandwidth
digitizing oscilloscope offers the
highest throughput and richest
feature set in the industry. An
engineer’s time is maximized by
more than 50 built-in parametric
measurements. Fast acquisition
and processing provides exceptional speed in both front panel
operation and in automated test
systems controlled by computers.
The 20-GHz or 50-GHz bandwidth
and the low-noise specifications
allow very precise measurements
on low-level, high-speed signals.
Timebase stability, accuracy, and
resolution allow characterization
of jitter in the most demanding
applications.
The Features and
Speed You Expect in a
High-Bandwidth
Digitizing Oscilloscope
Time domain reflectometry
(TDR) measurement
_____________________________________________________________________________
Time Base (Horizontal)
_____________________________________________________________________________
Scale Factor (Fullscale is 10 divisions)
Minimum 10 ps/division
Maximum 1 s/division
_____________________________________________________________________________
Delay (Time offset relative to trigger)
Minimum 22 ns
Maximum 1000 screen diameters or 10 seconds,
whichever is smaller
_____________________________________________________________________________
Time Interval ≤ 10 ps ±0.1% of reading
Accuracy (Dual marker measurement)
_____________________________________________________________________________
Time Interval ≤ (screen diameter) / (record length)
Resolution or 62.5 fs, whichever is larger
_____________________________________________________________________________
Trigger-External Input Only
_____________________________________________________________________________
Sensitivity
dc to 100 MHz 40 mV peak-to-peak
100 MHz to 2.5 GHz Increasing linearly from 40 mV at 100 MHz
to 200 mV at 2.5 GHz
_____________________________________________________________________________
Pulse Width Required >200 ps
_____________________________________________________________________________
Pulse Amplitude Required >200 mV
_____________________________________________________________________________
Bandwidth Trigger bandwidth reduced to
Limit approximately 100 MHz
_____________________________________________________________________________
Jitter
(trigger and ≤ 2.5 ps + 5E-5 x delay setting
time base (Tested using a 2.5-GHz synthesized source
combined) at 200 mV)
(one standard deviation)
_____________________________________________________________________________
Trigger Input:
Maximum safe ±2 Vdc
Input Voltage
Nominal 50 Ω
Impedance
Percent Reflection ≤ 10% for 100-ps rise time [1]
Connector 3.5mm (m)
_____________________________________________________________________________
Calibrator Adjustable Output range: –2.5 V to +2.5 V
when terminated into 50 Ω
Output Delta Voltage Accuracy:
±(0.2% of settings)
_____________________________________________________________________________
[1] With non TDR plug-ins, ≤10% for 200 ps rise time with TDR plug-ins (HP 54754A and 54755A).
HP 54750A Mainframe Specifications
• 62.5-fs Resolution
• 10-ps Accuracy
______________________________
Channels (Vertical)
______________________________
Scale Factors: Adjustable from 1 mV/div to
100 mV/div in a 1-2-5-10 sequence from the
front panel knob or the INC/DEC keys. Also
adjustable over the range in 0.1-mV increments
from the numeric keypad.
Attenuation Factors: Factors may be entered to
scale the oscilloscope for external attenuators
connected to the channel inputs. The range is
from 0.0001:1 to 1,000,000:1.
Noise: Averaging reduces noise by 1/(n)
1/2
,
where nis the number of averages, until a
system limitation of approximately 25 µV (low
BW mode) or 50 µV (high BW mode) is reached.
Channel-to-channel isolation: >60 dB
Bandwidth: You may select between a high
(20 GHz/50 GHz) or low (12.4 GHz/26.5 GHz)
bandwidth mode. The sampler is biased
differently for the two modes. Typically high
and low bandwidth noise is ~1/2 the specified
maximum noise.
Time Base (Horizontal)
______________________________
Delay Between Channels: The difference in
delay between channels can be nulled out with
1-ps resolution to compensate for differences
in input cables or probe length. Up to 100 µs of
skew can be nulled out.
Reference Location: The reference point can
be located at the left edge or center of the display. The reference point is the point where the
time is offset from the trigger by the delay time.
The reference point is also the point that the
time base sensitivity expands and contracts
around as the time base is changed.
Triggered Mode: Causes the scope to trigger
synchronously to the trigger input signal.
Freerun: Causes the scope to generate its
own triggers.
HP 54750A
System Characteristics
2
Electrical Channels Optical Channels (Available with HP 83480K Firmware Upgrade)
Sensitivity
Unfiltered BW Number Data Rates Fiber (smallest power
Number Bandwidth GHz GHz (typical) of Filters Mbit/s Input Type Wavelength for mask test)
HP 54751A 2 12 or 20
HP 54752A 2 26.5 or 50
HP 54752B 1 26.5 or 50
HP 54753A 1 TDR/1 Electrical 12 or 18
1 Electrical 12 or 20
HP 54754A 2 TDR/2 Electrical 12 or 18
HP 83481A 1 12 or 20 2.5 (>3.0) 2 155 and 622 9/125 µm 1000 to 1600 nm –17 dBm
HP 83482A 1 18 or 40 30 9/125 µm 1000 to 1600 nm
HP 83485A 1 12 or 20 20 1 155/622 or 2488 9/125 µm 1000 to 1600 nm –10 dBm
HP 83485B 1 18 or 40 1 9953 9/125 µm 1000 to 1600 nm –8 dBm
HP 83486A 1 12 or 20 2.5 (≥2.7) 2 155/622 or 62.5/125 µm 1000 to 1600 nm –19 dBm
1063/1250
HP 83487A 1 12 or 20 2.5 (>2.7) 2 1063/1250 62.5/125 µm 750 to 870 nm –17 dBm
Plug-In Modules for the HP 54750A
Measurements
Vamptd freq
Vbase Tfall
Vtop Trise
preshoot Tmax
overshoot Tmin
Vp-p Tvolt
Vtime Vavg
Vmin Vupper
Vmax Vmiddle
Vrms Vlower
+width FFTfreq
-width FFTmag
duty cycle FFT delta freq
delta time FFT delta mag
period TDR min reflection
TDR max relection
TDT prop delay
TDT gain
Histograms
p-p
median
mean
std dev
µ ± 1 sigma
µ ± 2 sigma
µ ± 3 sigma
hits
peak
Limit Test
Signals can be tested by up to four automatic
parametric measurements and compared to
user-defined test boundaries. Failures tolerances can be selected independently for each
of the parametric tests.
On failure actions:
1) Save channel data to memory,
disk or printer
2) Save screen to pixel memory,
disk or printer
3) Save a text log summary of all
failures with time tagging to disk or
printer.
Limit test can be set to run continuously for a
user-selected number of waveforms, or for a
defined number of failures.
Mask Test
Acquired signals are tested for fit outside
areas defined by up to eight polygons. Any
samples that fall within the polygon boundaries result in test failures. Masks can be
loaded from disk, HP-IB, or created
automatically or manually.
HP 54750A
System Characteristics (cont’d)
Typical Timing Accuracy: The time base uses a
series of 4-ns blocks. Time base linearity and
small discontinuities across these blocks contribute to the 8-ps accuracy specification.
When operating within 4 ns blocks, the typical
accuracy is shown by the following graph. The
graph below is a result of many measurements
on multiple instruments.
Because averaging implies single-valued
waveforms, the Best Flatness control takes
advantage of this to further improve flatness.
This is done by taking a sample with the samplers turned on and then with samplers turned
off. The two results are then subtracted, thereby removing the residual nonflatness.
Display Colors: You may choose a default color
selection, or select your own colors from the
front panel, or via HP-IB. Different colors are
used for display background, channels, functions, background text, highlighted text, advisories, markers, overlapping waveforms, and
memories.
Documentation Aids
_____________________________________
Waveforms, scaling information, and measurement results can be transferred directly to
HP-IB or Centronics graphics printers.
Waveforms may also be stored on the internal
MS-DOS®compatible disk in PCX or TIFF
format. This allows moving screen data into
word processors for documentation.
Programmability
_____________________________________
Instrument settings and operating modes,
including automatic measurements, may be
remotely programmed via HP-IB (IEEE 488.2).
HP-IB programming complies with the recommendations of the IEEE 488.2 standards. The
HP 54750A can be programmed to take data
only at specified time points, or to return only
measurement results (such as, tr, tf, and frequency) to speed up data acquisition.
Data Transfer Rate: 550 Kbytes/s typical.
Measurement Times: 25 automatic measure-
ments per second typical.
Data Record Length: 4k points maximum per
channel.
Measurement Aids
_____________________________________
Markers: Dual voltage or time markers can
be used for a variety of time and voltage
measurements. Voltage markers can be
assigned to channel data, measurements, functions, FFTs, histograms, color-graded displays,
and memories.
Automatic Pulse Parameter Measurements
Functions
magnify differentiate
invert min
add max
subtract FFTmag
multiply histograms
divide measurement limit testing
versus waveform mask testing
integrate color-graded display
bw limit
Histogram of 274 Measuremets of
50 ps Time Interval that Crossed
the 4 ns Sweep Disontinuity
Mode on Five HP 54750 Systems
all at 10 ps/div
Average Error = 0.4 ps
Standard Deviation = 1.82 ps
Histogram of 236 Measurement
on Four kHP 54750 Systems of
50 ps time Interval that did not
Cross the 4 ns Discontinuity.
All at 10 ps/Div.
Average Error = 0.2 ps
Standard Deviation = 0.15 ps
Time Interval Error
10 8 6 4 2 0 -2 -4 -6 -8 -10
Trigger
_____________________________
Edge Trigger: Triggers on the positive or nega-
tive edge of the trigger input.
Hysteresis: The trigger hysteresis can be set to
two modes:
Normal— the trigger hysteresis is set so
the instrument meets the trigger sensitivity
specification.
High Sensitivity — hysteresis is turned
off to allow a best sensitivity to highfrequency signals. This mode should not
be used for noisy lower frequency
signals that may mistrigger without
hysteresis.
BW Limit: Puts the trigger signal through a
low-pass filter of approximately 100 MHz.
Display
_____________________________________
Data Display Resolution: 451 points
horizontally x 256 points vertically.
Graticules: You may choose full grid, axes with
tic marks, frame with tic marks, or no graticule.
Display Modes
_____________________________________
Averaging: The number of averages can be
specified between 1 and 4096 using the numeric keypad. The INCrement/ DECrement keys or
the knob will set averaging to 1, 2, 4, 16, 64, 256,
1024, 4096. On each acquisition, 1/ntimes the
new data is added to (n-1)/nof the previous
value at each time coordinate.
Best Flatness: The samplers used in the
HP 54750A system have a certain amount of
passive feedthrough. This feedthrough is
corrected by a compensation circuit. A small
amount of feedthrough does remain after
compensaton and this causes slight
nonflatness in the step response.
3
Data collected during test:
Total number of waveforms examined
Number of failed waveforms
Total number of samples taken
Number of failed samples
Number of hits within each polygon
boundary
Auto Mask Creation
Masks are created automatically for singlevalued voltage signals. Both delta X and delta
Y tolerances can be specified.
The failure actions are identical to those
of Limit Testing. Both mask testing and
limit testing may be used independently or
simultaneously.
FFT
Up to three fast Fourier transforms can be
run simultaneously. The three built-in filters
(Hanning, rectangular, and flattop) allow optimization of frequency resolution, transients,
and amplitude accuracy. Automatic measurements can be made on frequency, delta
frequency, magnitude, and delta magnitude.
Frequency Span = Sample rate /2 = record
length/(2 * time base range)
Frequency Resolution =
Time base range/record length
Color Graded Display
Infinite persistence display mode where color
differentiates the number of times any individual pixel has been acquired. All points
acquired are added to a database and then
displayed as one of eight colors depending
upon the frequency of acquisition.
Automatic parametric measurements may be
taken on the Color-Graded Display allowing
parametric evaluation of multivalued (eye
diagram) waveforms.
Autoscale: Can find repetitive signals:
> 50 Hz
duty cycle > 1%
amplitude > 10 mV p-p vertical
50 mV p-p trigger
Waveform Math: Two functions can be
specified and displayed. Functions may be
defined as:
magnify versus
invert integrate
add differentiate
subtract minimum
multiply maximum
divide fft magnitude
Waveform Save: Four waveforms may be
stored in four nonvolatile memories. They
may also be stored to the internal MS-DOS
compatible 1.44 Mbyte disk.
Setup Aids
_____________________________________
Autoscale: Pressing the Autoscale key
automatically adjusts the vertical of all channels, the horizontal scale factors, and the
trigger level for a display appropriate to the
applied signals.
Channel Autoscale: Autoscale can be
performed on individual channels. This mode
only sets the vertical for the channel selected,
therefore saving time and keeping the
automatic feature from changing other userselected settings.
Save/Recall: Up to 10 complete instrument
setups may be stored in the internal nonvolatile memory. Additionally, many more
setups can be stored on the internal MS-DOS
compatible disk.
Vertical Software Calibration: Changes in the
environmental conditions can be accommodated by performing a software calibration on
the plug-ins. The calibration resets the plug-in
for the current mainframe and plug-in operating temperature. Software vertical calibration
is recommended prior to taking measurements
requiring the best possible accuracy.
Digitizer Converter: 12-bit successive
approximation A/D converter.
Resolution: Up to 15 bits with averaging.
Variable IF gain assures that resolution is
≥9 bits on all ranges (11 on most).
Digitizing Rate: The signal is sampled and digitized at a rate dictated by the trigger repetition
rate and the time base range. If data acquisition is not trigger rate limited, the maximum
sample rate is 40 KHz
.
Reflection Measurements
_____________________________________
Source: Measurements are made using the
Channel step source or a user supplied
external source.
Calibration: A reference plane is defined by
calibrating the reflection channel with a short
placed at the point where the device under
test (DUT) will be connected. The short
calibration is followed with a 50-Ω calibration.
These calibrations derive the normalization
filter.
Cursor: Reads out the percent reflection,
impedance, time, and distance from the reference plane to the cursor.
Percent Reflection: Automatic measurements
provided to calculate the maximum positive
and negative percent reflections of the waveform shown onscreen.
Normalization Filter: Applies a firmware
digital filter to the measured data. The rise
time of the filter may be varied to allow the
user to simulate the edge speeds that would
be seen by the device under actual operating
conditions. Normalization uses the Bracewell
transform, which is under license from
Stanford University. See TDR output specifications for allowable rise time values.
Variable Bandwidth Limit
Changing the rise time of the normalization
filter requires that a short and 50-Ω load be
placed at the launch point. When this is not
possible or differential TDR is used then a
variable bandwidth limit filter can be used.
The variable bandwidth limit function is
located in the waveform math menu.
Percent Reflection Measurements: Used to
quantify reactive peaks and valleys of the TDR
display. Impedance measurements are valid
only for resistive, horizontal flat-line TDR
displays. Because the accuracy depends on
the measurement being made, percent reflection and impedance accuracies are not specified. Percent reflection and impedance measurements are ratios of voltage measurements
whose accuracies are specified.
Percent Reflection (ρ) = (Vcursor - Vtop)
_____________
(Vtop - Vbase)
Impedance (Z) = 50 Ω x (1 + ρ)
________
(1 - ρ)
Where Vcursor = voltage at the cursor
Vtop = high level of incident step
Vbase = low level of incident step
and is determined during the
reflection calibration
Distance measurements are subject to the
accuracy of the velocity factor or dielectric
constant that you enter. Because the
HP 54753A and 54754A have no control over
the accuracy of these numbers, distance
accuracy is not specified. Distance is derived
from time interval measurements whose accuracies are specified.
4
Distance (d) = 1/2 x ∆t x Velocity Constant,
where ∆t = time from the reference plane to
the cursor.
Dielectric constant = (3 x 108m/s)
2
__________________
(Velocity Constant)
2
where you enter either a relative Dielectric
Constant or a Velocity constant.
The TDR’s ability to resolve the distance
between two discontinuities is limited to 1/2
the system rise time. Without normalization,
this is approximately 1/2 x 45 ps or 7 mm in air.
For the distance resolution in your media,
divide 7 mm by the square root of eff of your
media. With normalization, the system rise
time can be 10 ps yielding 1.5 mm of resolution
in air.
The maximum length the TDR can measure is
subject to media loss. For a lossless vacuum,
and using a 50-Hz TDR repetition rate, the
system can measure 1500 km. Actual maximum lengths will generally be limited by the
losses of the media under test.
Excess L/C
Once a calibration plane is established, the
excess L/C feature will compute the series L
or shunt C equivalent to the area between
cursors. The result is the equivalent L or C
that causes a discontinuity with equal area to
that measured.
Scaling
The vertical scaling allows scaling in either
percent reflection or ohms. Cursors will also
read in voltage or ohms. A reference plane
calibration must be performed to utilize these
scales.
Differential TDR
The differential and common stimulus are
generated by staggered rising edges from two
independent TDR step generators. Hardware
setup remains fixed and therefore the skew for
both differential and common measurements.
The response to either differential or common
mode stimulus may be viewed simultaneously
as differential or common mode.
All waveform math functions are automatic.
Both the differential and common mode
responses are computed without user
intervention.
Differential TDR Timing Deskew
For accurate differential TDR measurements it
is essential that the TDR steps are coincident
at the reference plane and the reflected steps
are coincident at the samplers. Ideally, this is
accomplished by using electrically matched
launch cables. When this is not possible, the
TDR channels must be deskewed.
To accomplish this, it is necessary to have both
TDR step time skew and channel skew capabilities. Each of the TDR steps may be moved
± 400 ps. When used in conjunction with the
channel skew control, then 1.6 ns of timing
delta can be removed. Depending upon the
cable, this equates to approximately 1 foot of
cable length difference that can be electrically
removed.
Transmission Measurements
Source: Measurements are made using the
channel step source or a user-supplied external source.
Calibration: A calibration with a straightthrough path or through your standard device
determines reference amplitude levels and
reference time and distances of the signal
path. These reference levels are used for gain
and propagation delay measurements.
Cursor: Reads out time referenced to the calibration edge and gain referenced to the transmission calibration results. (See Note 1)
Propagation Delay and Gain: Automatically
calculates the difference in time and distance
between the calibration signal path the test
signal path. Also calculates the ratio of the
test signal amplitude to the calibration signal
amplitude. (See Note 1)
Normalization Filter: Applies a firmware
digital filter to the measured data. The rise
time of the filter may be varied to allow you to
simulate the edge speeds which would be
seen by the device under actual operation.
See TDR output specifications for allowable
rise time values.
Note 1:
∆t = Time of the cursor (50%) - Time of
reference edge (50%).
Gain = (Vtop - Vbase)signal
___________________ ,
(Vtop - Vbase)reference
Prop Delay = Time of test edge (50%) - Time of
reference edge (50%).
Distance (d) = Prop delay x Velocity Constant.
where Vtop = High level of waveform and
Vbase = Low level of waveform.
5
Loading...