GW Instek MDO-2072EX MSO / MDO Oscilloscope - Review

Table of contents

RoadTest: GW Instek MDO-2072EX MSO / MDO Oscilloscope

Author: fmilburn

Creation date:

Evaluation Type: Test Equipment

Did you receive all parts the manufacturer stated would be included in the package?: True

What other parts do you consider comparable to this product?: Several other oscilloscopes in this class have a waveform generator but not the spectrum analyzer, DMM, and power supply. The Keysight DSOX1102G will be used as a comparison where capabilities overlap.

What were the biggest problems encountered?: None of consequence.

Detailed Review:

Introduction

 

The GW Instek MDO-2072 EX is a mixed domain oscilloscope with a spectrum analyzer, dual channel 25 MHz arbitrary waveform generator, 5000 count digital multimeter, and dual 5V / 1A DC power supply.  Bandwidths of 70 / 100/ 200 MHz and 2 or 4 channel versions are available.   This makes it well suited for educational facilities and users who's bench space is at a premium.  The review starts with an unboxing, overview of features, and some overall impressions followed by the main body of the RoadTest that:

 

  • describes the main features of the instrument
  • outlines a test procedure or method that demonstrates the feature
  • describes and summarizes the results of the test or procedure

 

Unboxing

 

The instrument came well packaged in a single box and without damage.

image

 

Including in the box are:

  • region dependent power cords (mine came with European and British versions instead of North American)
  • two 70 MHz passive probes with accessories
  • two leads for power supply
  • two probes for DMM
  • two BNC to BNC leads for arbitrary waveform generator
  • CD with user manual and software (also available for download online)
  • calibration certificate

 

Summary:  While all parts were received well packaged and in good condition the power cords were for a European customer rather than North American.

 

Features and Specifications

 

For feature and specification comparison the Keysight DSOX1102G is included in the abbreviated table below since it is a well equipped oscilloscope in the same price range and I am familiar with it.  There is a comparison of the Keysight to entry level Rigol and Siglent oscilloscopes here.  See the datasheet and user manual for additional information on individual instruments.

 

Feature

GW Instek MDO-2072 EX

Keysight DSOX1102G
Bandwidth70 MHz (100 and 200 MHz available)100 MHz hackable to 200 MHz
Sample Rate1 GSa/s2 GSa/s
Memory Depth10 Mpts per channel1 Mpts
Segmented MemoryYesYes
Waveform Update Rate120 k/s max50 k/s
Analog Channels

2 + 1 Ext Trigger

2 + 1 Ext Trigger

Ext Trigger can be digital input

Triggering

Edge

Pulse Width

Video

Pulse Runt

Rise / Fall Slope

Timeout

Alternate

Event-Delay

Time-Delay

Bus

Edge

Pulse Width

Video

Pattern

Rise / Fall Time

Setup / Hold

Time Base Range1 ns/div to 100 s/div (1-2-5 increments)5 ns/div to 50 s/div
Vertical Sensitivity1 mV to 10V div5 mV to 100 V/div
Data LoggingUp to 1000 hoursNo
Mask TestingYesYes
Math

FFT

Add

Subtract

Multiply

Divide

User Expressions with numerous options

FFT

Add

Subtract

Multiply

Divide

Display Quality

8" WVGA

Low Glare

Medium Font

7" WVGA

Low Glare

Medium Font

Serial Decoding

I2C

UART

SPI (4 channel only)

CAN

LIN

no additional cost

I2C

UART / RS232

SPI

CAN

LIN

option with additional cost

Wave Generator

25 MHz dual channel

14 bit, 200 MSa/s sample rate

Sine, Square, Ramp, Pulse, DC,

Noise, Lorentz, Exponential, Rise,

Exponential Fall, Haversine, Cardiac

20 MHz single channel

Sine, Square, Ramp, Pulse, DC,

Noise

Digital Multimeter

5000 count

DC Voltage 6 ranges 50mV - 1000V

DC Current 3 ranges 50mA  - 10A

AC Voltage 5 ranges 50mV - 700V

AC Current 3 ranges 50mA - 10A

Resistance 5 ranges 500Ω - 5MΩ

Diode Test max 1.5Vf

Temperature -50C to 1000C

Continuity Beeper

3 digit Voltage
Digital Power Supply

2 Channel

1.0V to 5V range

1 A max output

0.1V increment adjustable

No
Spectrum Analyzer

Frequency Range DC - 500 MHz

Span 1kHz - 500 MHz

Res Bandwidth 1 Hz - 500 MHz

Reference Level -50 dBm to + 40 dBm

Vertical Scale 1 dB/div to 20 dB/div (1-2-5)

100V max input

No

Frequency Response Analysis

(Bode Plot)

Yes, to 25 MHzYes, to 10 MHz
ConnectivityUSB, Ethernet, Go/NoGo BNCUSB
Input Impedance

1 MΩ

16 pF

1 MΩ

16 pF

Probes

70 MHz

1x / 10x

200 MHz

1x / 10x

Fan NoiseModerateHigh
Training Signals and EducationSeparate board available extra costBuilt-in
Documentation QualityGoodGood
StatusCurrent ModelCurrent Model
Approximate Cost, USD$1050 ($1300 for 100 MHz Model)$1100 fully optioned 100 MHz

 

There are a number of features and specifications deserving further discussion.

 

  • Bandwidth on the GW Instek model tested is relatively low for a scope these days at 70 MHz but models up to 200 MHz are available.  The sample rate is adequate for a 70 or 100 MHz scope.
  • MDO-2072EX memory depth is better than the Keysight model and when combined with segmented memory increases the chance of capturing important events in a single capture.
  • The higher waveform update rate on the MDO-2072EX is useful for spotting infrequent events and glitches
  • Data Logging for extended periods is a feature on the MDO-2072EX not found on all of the competitors
  • The display is noticeably bigger and easier to read than the Keysight (which is easier to read than the other entry level scopes I have used)
  • Serial decoding of SPI is not possible with 2 channel versions of the MDO-2072EX (it is possible with the Keysight 2 channel scopes which can use the Ext trigger as a digital input)
  • The waveform generator has a relatively high spec, has two channels, and can be used to make Bode plots all the way up to 25 MHz
  • The Digital Multimeter is full featured with the exception of capacitance measurement
  • The two channel power supply is unique and adequate for many low voltage, low power experiments
  • The spectrum analyzer is unique and a great addition although not a replacement for a full featured instrument

 

Costs are indicative only, see retailer for actual cost.  Detailed discussion and findings are covered in the sections below.

 

Summary: The instrument being tested is feature rich and well positioned for the educational market and might also be suited for a home electronics bench with limited space.

 

Setup and First Impressions

 

The GW Instek MDO-2072EX appears physically large (380mm x 208mm x 127mm) in the comparison photo with the Keysight DSOX1102G but is shallow in depth and would not take excessive space on a bench especially given the features.

image

 

The top is such that other instruments could not be stacked on it.  I don't find the size off putting and on the 4 channel model each channel retains it's own vertical controls which I prefer.  The control locations were easy to get used to and the selection buttons / tabs at the bottom worked well with the user interface.

 

The instrument takes about 28 seconds to boot and then displays a message that all power on tests passed.  The firmware was updated as recommended by the user manual to version 1.41 by downloading a file and transferring it via a USB memory stick without issue.  It was also necessary to download and install the frequency response app and this was also done without issue.

 

The probes supplied are rated at 70 MHz.  The grounding clips are stiffer and a bit more difficult to open than some.  Ground springs that fit over the probes are not provided.  Probe compensation is straight forward with the adjustment located at the BNC connection.

 

Summary: Setup was uneventful and it was possible to find the way around the oscilloscope without referring to the manual too much.  Menus aren't too deep and for the most part are intuitive.

 

Layout

 

The front panel is laid out in a fairly conventional manner with the addition of the bottom menu keys working nicely with the user interface.

image

Source:  GW Instek User Manual

 

In addition to what one normally expects to find on the front panel there are ports for the digital multimeter placed bottom left.  The option key located near the bottom center is used to select the spectrum analyzer, AWG, DMM, and power supply menus.  The writing above the hardcopy key, menu key, and option key is small and difficult to read except in very good light.

 

image

Source:  GW Instek User Manual

 

The two power supply outputs and the two arbitrary wave generator outputs are located top right on the rear panel.

 

Summary:  In general the physical interface was easy to adapt to and doesn't get in the way of the user.

 

Basic Sine Wave

 

In this section a sine wave will be generated with the AWG and displayed with cursors to demonstrate basic operation.  Setup consists of running one of the BNC leads from GEN1 on the rear panel to channel 1 and setting the sine wave to 1V p-p and 100 kHz.

 

More detail on the AWG will be given below but it is easy to set up.  Cursors are different than what I am used to but were quickly figured out without referring to the manual.  Probe Voltage needs to be set to 1X for proper readings when using BNC to BNC connections between the AWG and scope so beware when changing back to 10X probes.  The horizontal time and vertical voltage settings were done manually and this was quick and easy.  Auto setting is also quick.  Screen shots can be recorded with a single button press.

image

 

Summary: Basic setup is quick and easy.  The base display gives the necessary information and is laid out neatly in an easy to read fashion.

 

Triggering on Noisy Signal / Noise Rejection

 

A noisy signal can sometimes cause an edge trigger to occur outside of where it is expected.  In this demonstration a 1 kHz Sine wave is fed to the GW Instek scope from the Keysight.   The default trigger settings being used result in an undesired trigger on the noise as shown below.

image

 

The setting for noise rejection and HF / LF rejection are in the coupling tab which required a search in the manual to find.  Selecting Noise Reject On causes the scope to trigger where desired.  Since the noise is high frequency, choosing HF Reject also allows proper triggering.

image

 

Summary: Noise rejection works as expected.

 

Saving and Recalling Information

 

The MDO-2072EX can store *.bmp or *png 800x480 pixel images.  It can also store up to 20 waveform files in internal memory in a *.lsf format.  Stored waveforms can be transferred to up to 4 reference waveforms for direct display.  Waveforms can also be stored in comma separated value (*.csv) format and opened in spreadsheets.  Settings are stored in a proprietary *.set format.

 

Individual labels are available for input channels, reference waveforms, and file names.  The letters are selected by turning the "Variable" knob and then pushing the "Select" key.

 

There are numerous options, for example below is a sine wave using saved ink saver mode.

image

Displaying a reference waveform alongside the current real time display can be useful.  The stored waveform must be moved into one of the 4 reference waveform locations.  Here the 1 MHz waveform from above has been stored and then displayed in grey alongside a 999 kHz waveform.

image

 

Saving settings would be useful in a classroom situation where multiple students are using the same equipment.  It is also useful where an experiment or test is repeated frequently.  Removing reference waveforms is done with the small R button in the vertical control section which I had to hunt for on first use.

 

Summary:  The file saving and retrieval features are adequate and easy to use.

 

Arbitrary Waveform Generator (AWG)

 

The 25 MHz Dual Channel Arbitrary Waveform Generator has 14 bits vertical resolution and a sample rate of 200 MSa/s with 16k memory length.  Thirteen pre-built waveforms are provided as seen in the screenshot below.

image

 

The waveform settings include the ability to set frequency, amplitude, and DC offset.

 

Modulation options include AM, FM, and FSK.

 

Frequency can be swept in a linear or log fashion and there are options available for Start, Stop, Sweep Time, Span and Center

 

The load impedance can be set to 50 Ohms or High Z.  Phase can be set for the channel 1 generator.  Other options such as duty cycle can be set for pulse waves.

 

There is a useful summary view of the state of the waveforms.  In the screenshot below both channels can be seen with a sine wave defined on the left and a square wave on the right.

image

 

  The resultant waves can be seen in the screenshot below.

image

 

I find the display clean and informative.  On the left can be seen in white font the AWG output.  Just below it are the values for the horizontal and vertical grid.

 

It is possible to load arbitrary waveforms as CSV files and it is also possible to edit files on the instrument as shown by the glitch I've added to the sine wave below.

image

 

Summary: The Arbitrary Waveform Generator is flexible, goes to 25 MHz and has two channels.  It integrates well into the oscilloscope and will be used to demonstrate other features of the instrument as the RoadTest progresses.  By using the math function more complicated waveforms could be generated such as adding noise from one channel to the other.

 

Advanced Triggering

 

The MDO-2072EX offers a number of triggering options beyond edge including holdoff, delay, pulse width, video, pulse & runt, timeout, and single shot.  Edge triggers can be rising, falling, or either and are demonstrated elsewhere in this RoadTest.  Holdoff, pulse width, and single shot will be demonstrated here.

 

Holdoff is the waiting period before the oscilloscope starts triggering again after a previous trigger point.  This can be helpful when bursts are occurring and triggering is desired only on the first pulse in the burst.

image

Source:  GW Instek MDO-2000E_2000EA User Manual

 

In the contrived example below an arbitrary pulse waveform was generated manually on channel 1 and holdoff is set at the default 4 ns.  Perhaps not apparent immediately is that triggering is occurring on different rising slopes and that multiple overlapping bursts are being displayed.

image

 

The major divisions horizontally are 5us.  By changing the holdoff to something greater than 5us (here 10us) assurance can be obtained that the start of the burst is caught.  The horizontal scale can then be changed to better view the burst as shown below.

image

 

Pulse Width Triggering is used to catch the following 50 kHz clock signal with a rare intermittent glitch being purposely generated by the Keysight.  The glitch can be seen as a brief flash with normal settings but isn't visible most of the time or in the screen capture below.

image

 

By turning the persistence up to infinite a glitch becomes visible in the waveform on the left.

image

 

It can be seen that the glitch has a pulse width less than 2us.  By setting the trigger to pulse width to < 1.6us the glitch is isolated as shown below.

image

 

Single Shot was also used in the screen shot above to stop the capture after the glitch was triggered.

 

Summary: All of the triggering functions tested work as expected.  The pulse width triggering test performed above could not be done with some scopes I've tested in the past and can be tricky but worked without issue here.

 

Cursors and Automatic Parametric Measurement

 

To demonstrate cursors and parametric measurement the AWG was set to generate a 10 kHz 1V p-p Gaussian (just to be different) waveform with 1.5V offset on channel 1.

image

 

The horizontal and vertical cursors were then set to the peak and mid-valley by eye, and the automatic measurements for frequency, period, peak to peak voltage, and low voltage added to the bottom display pane.  Up to 8 automatic measurements can be displayed at any one time.

image

It is also possible to add statistics for the automatic measurements.

image

A measurement summary can also be displayed in tabular form overlaying the waveform.

image

 

Summary: The automated measurements one expects are present and the cursors are easily set manually.  Up to 8 automated measurements can be selected for display simultaneously and statistics are also available.  The measurement summary is displayed in an easy to read tabular form.

 

Gate and Zoom

 

The Zoom function is handy for capturing and scrolling through waveforms.  A small window at the top gives a global view with a close-up in the window below.  The main features are demonstrated in the short video below.

 

Summary:  Zoom works as expected and has a useful play / pause feature.

 

Math Functions

 

While FFT is grouped with the Math Functions in the MDO-2072EX, it will be evaluated with the Spectrum Analyzer in this RoadTest.

 

The scope has basic and advanced tabs for doing math on the channel.  The basic tab allows one channel to be added, subtracted, multiplied, or divided by the other.  The advanced tab allows about any reasonable operation desired.

 

The AWG was set up first with a 1kHz 500mV p-p Sine wave on channel 1 and a 100Hz 2V p-p Sine wave on channel 2.

image

 

For basic math the source can be either of the two oscilloscope input channels or reference waveforms that have been previously stored.  The user then picks the one operator to be applied.  The position and units per division can be modified.  In this example channel 2 is added to channel 1 and displayed on the red trace with the same vertical units per division as channel 1 and channel 2.

image

 

Advanced math offers a significant number of options as shown in the screenshot below.

image

 

Let's say we want to integrate the difference between channel 1 and channel 2.  Then the following expression can be entered.

image

 

After modifying the time scale the oscilloscope shows the following where the red trace is the integrated result.

image

 

Summary:  The math and advanced math functions are powerful and easy to use.

 

Peak Detect

 

Infrequent glitches can go undetected on the oscilloscope when in normal sampling modes.  In the waveform below there is an unseen glitch that appears infrequently and would only appear in a screenshot with luck.

image

 

The MDO-2072EX has a peak detect mode where minimum and maximum value pairs are displayed.  The red arrow  inserted in the screenshot below points to one of the detections from the waveform above.

image

 

Summary:  Peak Detect works as expected.

 

Segmented Memory

 

Segmented memory can be used to capture increased information / number of bursts separated by long intervals as only the bursts with associated timing are stored and not the intervening interval.

image

Source: GW Instek User Manual

 

A "RF Burst" was set up on the Keysight and monitored on channel 1 of the MDO-2072EX as shown below.  The number of segments to be captured and other information can then be entered behind the acquire button.

image

 

Here 600 segments were captured over a period of more than 2 seconds.  The screen capture shows the 590th segment out of 600 captured.

image

 

It is also possible to use the play / pause buttons to scan through segmented memory.

 

Summary: Segmented memory on the MDO-2072EX is easy to use, works well, and due to the increased memory can capture more segments than  the Keysight DSOX1102G.  The Rigol and Siglent used in my previous oscilloscope comparison did not have segmented memory.

 

Masking

 

Masking works by creating a template of a defined shape which is compared to an input signal for excursions.  It is installed as an application and is accessed through the APP key.  Both auto masking and user defined masking are available along with a number of options.  Only a high level examination of auto masking was done.

 

Below a mask defined by Auto Mask was created around a purposely noisy sine wave from the Keysight and monitored.

image

 

Simple statistics are available in red on the left bottom of the trace indicating that 100 excursions have occurred out of 13,983 waveforms examined.  On the Keysight, excursions are marked with a red dot on the trace so that they are more readily apparent to the user.

 

Summary: Masks behaved as advertised.

 

Serial Protocol Decoding

 

While serial protocol decoding is included in the base cost, SPI is not available on the 2 channel scopes.  A quick check of I2C was done to get an idea of what it looks like on the screen using an Infineon XMC 2Go microcontroller development board communicating with an Infineon 3DSense Shield2Go .  After configuration with the Bus key the following output is obtained.

image

 

Summary: The feature would be useful in an educational setting for introducing students to serial decoding but SPI is missing on the 2 channel oscilloscope models.

 

Power Supply and Digital Multimeter (DMM) DC Voltage Measurement

 

The DC voltage measurement capabilities of the DMM and and the power supply were tested together and compared to a Tenma 72-1020 bench multimeter known to be accurate by incrementing the power supply from it's minimum to maximum voltage without a load.  Several loaded cases were then checked. The RMS ripple and noise of the power supply were checked using the oscilloscope.  Other capabilities of the DMM were then tested separately against the Tenma 72-1020.

 

The power supplies can output from 1V to 5V in 0.1V increments.  Output is limited to 1 Amp and is protected by a resettable fuse.  The screen below pops up if current is exceeded which occurred in an unplanned test due to operator error.

image

 

The unloaded power supply output voltage was checked on the MDO-2072EX DMM by comparing to a Tenma 72-1020 for both channels and recorded in the table below.

image

image

A screenshot of the MDO-2072EX DMM in DC voltage measuring mode is shown below.

image

The power supply was then loaded with a nominal 10 ohm resistor measured to be 10.19 ohms using a LCR meter with Kelvin 4-wire measurement and tested on both channels using the Tenma 72-1020 to check output voltage.

image


image

Both power supply channels are well within the datasheet specification of +/- 3%.  The DMM DC voltage is also within the datasheet specification of +/-(0.1% + 5 digits).

 

A limited number of readings of ripple and noise were made with the MDO-2072EX and found to be under the 50 mV RMS datasheet spec.

 

Summary: Both power supply channels met the specifications from the datasheet in the areas checked.  The DMM also met DC voltage specifications in the areas checked.  The power supply has limited adjustment resolution and 1V to 5V range with both channels voltage relative to the same ground.  The high voltage range of the DMM was not tested.

 

Additional Digital Multimeter (DMM) Functions

 

The 5,000 count DMM has the following features.

FeatureRangeAccuracyComment
DC Voltage6 ranges (50 mV to 1000V)+/- (0.1% + 5 digits)10 MOhm Input Impedance
DC Current3 ranges (50 mA to 10A)+/- (0.5% + 50 mA)Accuracy in 50 mA and 500 mA range
AC Voltage5 ranges (50 mV to 700V)+/- (1.5% +  15 digits)Accuracy in 50 Hz to 1 kHz range
AC Current3 ranges (50 mA to 10A)+/- (3% reading + 50 mA)Accuracy in 50 Hz to 1 kHz range
Resistance5 ranges (500 Ohms to 5 MOhms)+/- (0.3% reading + 3 digits)Accuracy up to 500 kOhms
Diode TestMax forward voltage 1.5VOpen voltage 2.8V
Temperature-50 C to 1000 C0.1 C resolution
Continuity15 Ohms

 

DC voltage is tested in the section above using the instrument power supply.  DC current was not tested.  AC voltage and current were not tested.

 

Resistance was tested on a range of resistors and compared against a Multicomp Pro Handheld LCR meter using Kelvin 4-wire measurement and known to be reasonably accurate.  Probes provided with the instrument were used for measurement on the MDO-2072EX.

image

The meters agree within the datasheet specification range.  It is not surprising that the 10 Ohm measurement is slightly high given that 2-wire measurement was used for the MDO-2072EX.  A screenshot of the DMM in resistance measuring mode is shown below.

 

image

 

Diode measuring mode worked as expected.

 

image

 

Temperature measurement was not tested other than at normal room temperature but worked as expected.

 

image

 

In general the instrument is slow measuring resistance and the continuity mode is slow and care must be taken to make good contact. 

 

Capacitance measurement is not provided by the instrument.

 

Summary: The DMM met the specifications given in the datasheet.  The DMM has more features than the voltage measurement available on most oscilloscopes and provides a useful range.  The continuity beep is slow to respond compared to other multimeters in my possession.  Current measurement was not tested.  Capacitance measurement is not provided.  It is expected that many labs will want to purchase a handheld DMM for the times a second meter is needed as well as to supplement the features on the MDO-2072EX.

 

Frequency Response Analysis (FRA)

 

The FRA application uses the oscilloscope and wave generator to assist in frequency response analysis.  The simple RLC band pass circuit illustrated in the schematic below was created on a breadboard.  The values shown for the components were measured with a RLC meter at a frequency of 40kHz.

image

 

An online filter analysis tool was used to generate the Bode Plot for the filter for comparison to the experimental value.

image

The FRA application on the MDO-2072EX was then connected to the breadboarded filter per the diagram on the oscilloscope, setup and run.

image

image

 

The analysis tab leads to a screen that provides cursors for measurement.

image

 

The measured center frequency of the filter is 28kHz experimentally versus 28.2kHz calculated.

 

Summary: The Frequency Response Analysis application is easy to setup / run and provides useful output for analysis.

 

PC Software and SCPI

 

The MDO-2072EX comes with a CD but I no longer have a computer with a CD drive so I won't be evaluating whatever is on it.  OpenWave software is available for download online but was not tested.  The instrument has SCPI commands built in and can communicate with a PC over USB or Ethernet.  recently published a tutorial on using SCPI with pyvisa here and a quick demonstration will be done here using some of the techniques he covers.

 

The SCPI documentation for the MDO-2072 is detailed (over 300 pages) and appears to include about everything the instrument is capable of doing.  This demonstration is being done on a Windows 10 machine with NI-VISA installed and Python 3.8 over USB. 

 

The first thing I normally do with a new oscilloscope is set up a 1000 Hz 1V p-p sine wave  and see what it looks like.  Rather than do it with the knobs we will do that with SCPI.  Here is a python script that prints out the oscilloscope ID, generates a 1kHz 1V sine wave on AWG channel 1, auto sets oscilloscope channel 1, and then save a screen shot to a USB memory stick.

 

# Test pyvisa on GW Instek MDO-2072EX
# 1 Print oscilloscope ID to serial monitor
# 2 Generate 1kHz 1V sine wave on AWG channel 1
# 3 Autoset oscilloscope channel 1
# 4 Store screenshot to USB memory stick

import pyvisa
import time

resource_manager = pyvisa.ResourceManager()
mdo2072ex = resource_manager.open_resource("ASRL6::INSTR")

# ID
print(mdo2072ex.query("*IDN?"))

# AWG
mdo2072ex.write("AWG1:FUNCtion SINE")
mdo2072ex.write("AWG1:FREQuency 1000")
mdo2072ex.write("AWG1:AMPlitude 1")

# Autoset
mdo2072ex.write("AUTOSet")

# Screenshot to USB
time.sleep(5)
mdo2072ex.write("HARDcopy:START")

mdo2072ex.close()

 

This is what it looks like in the PyCharm IDE after running.  The serial monitor in the IDE shows the instrument information - i.e. GW,MDO-2072EX, GES190912,V1.41

image

 

Shown below is the screenshot from the oscilloscope sent to USB by SCPI.

image

 

Summary: The MDO-2072EX can communicate with a PC over ethernet or USB.  The SCPI commands are thoroughly documented and worked well in this brief examination.

 

Logging Data

 

The Data Log application is not a conventional data logger but instead logs waveform data or image captures along with the date and time.  Up to 1000 hours can be captured.  The minimum interval for waveform data is 2 seconds and the minimum interval for screen captures is 5 seconds.  In the example below a potentiometer connected to the MDO-2072EX power supply is varied by hand and the data logging app set up to capture screens at 10 second intervals for one minute.

 

 

Data Logging

image

0 Seconds

image

10 Seconds

image

20 Seconds

image

30 Seconds

image

40 Seconds

image

50 Seconds

 

 

Summary: The Data Logger makes screen captures or records oscilloscope waves / traces in data format at minimum intervals in the 2 to 5 second range for up to 1000 hours.  In a short test it performed as expected.

 

Spectrum Analyzer and FFT

 

EDIT 30 March 2021: The FFT section has been modified to include recommendations from the manufacturer on using FFT on the oscilloscope.

 

In addition to the FFT math function found on most oscilloscopes in this class, the MDO-2072EX has a software / firmware spectrum analyzer interface which is one of the reasons I was attracted to this RoadTest.  The differences in the time domain and frequency domain will be briefly demonstrated as well as a discussion of the FFT and spectrum analysis capabilities of the instrument.

 

Conventional FFT calculations in an oscilloscope are done over the entire signal bandwidth up to half the sampling rate.  The MDO-2000E series can analyze a defined region of the spectrum with greater frequency resolution.  The center frequency, span, and start/stop frequency can be entered in a similar manner to a traditional spectrum analyzer.  Since the oscilloscope frontend is utilized, it can accept DC components that would damage traditional spectrum analyzers and frequency sweeping is faster.  However, the upper frequency range is limited to 500MHz.

 

The specifications given below are from the GW Instek User Manual.


image

Source:  GW Instek MDO-2000E_2000EA User Manual

 

Time and Frequency Domains

 

The video below is a quick demonstration on use of time domain, FFT, and spectrum analysis on a signal generated by the arbitrary wave generator in the instrument.

 

AM Signal

 

Several demonstration waveforms for the spectrum analyzer are available by selecting APP->DEMO->SA on the front panel and will be used here for AM, FM, and FSK waveforms.

 

In the time domain it can be seen that there is something at 1 MHz signal with varying amplitude.

image

 

Turning on FFT the following can be seen.  With cursors the sidebands can be picked out 50 kHz away.

image

Note:  This screenshot has been modified from the original post to improve the display. 

 

For comparison, here is the FFT from the same signal displayed on the Keysight DSOX1102G.

image

 

The MDO-2072EX can capture more points (1M) and when adjusted properly displays with better resolution.

 

Turning on the GW Instek MDO-2072EX spectrum analyzer allows easier set up for the frequency domain and improved information display compared to other scopes I have used.  With just a few adjustments  it can be seen that a 1MHz AM carrier is present and being modulated with a 50kHz signal. 

image

 

FM Signal

 

Below is an FM signal with a center frequency of 15 MHz as seen with the FFT (adjustment not optimized).

image

 

With the spectrum analyzer it is easier to set up a good view with more information.

image

 

FSK

 

An FSK signal as seen by FFT (adjustment not optimized).

image

 

And here with the spectrum analyzer.

image

 

Audio Frequencies

 

To explore the resolution possible in the audio range a 1V p-p square wave was set to sweep from10Hz to 10kHz in one second.  The 100Hz resolution band width obtained with 40kHz center frequency and 100kHz span is shown in the short 10 second video that follows.  The max holding feature is turned on - can you guess what the trace will look like before watching the video?

 

Summary: While not the equal of high end traditional spectrum analyzers the MDO-2072EX spectrum analyzer is easier to use in the frequency domain than FFT, gives a better picture, and has improved information display.  It has an advantage over traditional spectrum analyzers for inexperienced users in that it can safely handle higher DC voltages due to the oscilloscope frontend.  The controls are similar to a traditional spectrum analyzer and work well.  The user can start with a wide frequency range for an unknown signal, move the area of interest to the center, set the span, and using the cursors quickly analyze the spectrum of interest.

 

Conclusion

 

This was my first opportunity to use a GW Instek instrument and I was impressed.  The documentation is very good, the hardware worked as expected, and the user interface easily adapted to.  I especially enjoyed learning my way around the spectrum analyzer.  Thank you to element14 and GW Instek for selecting me to perform this RoadTest.

 

Some of the promotional material for the instrument seems marketed towards an educational setting and it does seem well suited to undergraduate labs.  It takes up little room on the bench and has a :

  • 70, 100, or 200MHz oscilloscope,
  • relatively large display which would be visible to several students working together,
  • Two power supply channels up to 5V,
  • Two arbitrary wave generator channels,
  • DMM,
  • Spectrum Analyzer,
  • and the software features expected in a modern instrument.

 

The scope has an optional  GDB-03GDB-03 oscilloscope training kit that must be purchased separately which might be useful in an education setting but was not tested here.

 

While some of the hardware is not as full featured as stand-alone instruments, a lab course could be modified to work within their constraints.  For example a lab on op-amps could be set up with modern rail to rail chips designed to work on battery power to meet the restrictions of the power supply.  An inexpensive but rugged handheld DMM could be added to the bench for current measurement, capacitor measurement, and those labs where more than one meter is needed.

 

The MDO-2072EX is well suited to educational labs and would also be a good fit for home use with limited bench space.  Those with more room have opportunity to look at standalone instruments (including those from GW Instek) tailored to their specific needs.

 

Thanks for reading.  This is a full featured instrument so coverage is necessarily reduced in some areas to fit things in this space.  As always your questions, comments, and suggestions are appreciated.

 

EDIT 30 March 2021: The FFT section of the RoadTest has been modified to include recommendations from the manufacturer on use of FFT on the oscilloscope.

Anonymous
  • Dear Frank,

     

    Thanks your understanding.

    For hardware of spectrum analysis on the MDO-2000E, there are no difference between normal DSO.

    There are no additional hardware, we are using digital circuit to achievement.

     

    For the bandwidth, its fixed on each model. Not  available to upgrade.

  • Hi Frank,

     

    I think it's acceptable, to have the measurement close like this (it could be slightly inaccurate load causing it, or connections).

    Regarding quality, unfortunately low-cost loads are not purely resistive, but will have capacitance/inductance, but for measurements up a say 100 MHz it might be fine. As you say, it's possible to make them, one method is to take two to four resistors  (100 ohm resistors if you're using two, otherwise four 200 ohm resistors (surface mount)) and solder them from the centre pin to body. The multiple resistors provide lower inductance. But too many provides too much capacitance, so there's a sweet spot. Like the photo below basically.

    Image source: tt030100, Improved Construction Technique for a 50 Ohm Termination

    There the author determined four 1206 were the sweet spot. However this is hard to assemble I think, it would be much easier to get just a connector rather than an inline adapter, and cut the center solder point flush, and then solder across the resistors. And then use a  BNC T-piece. Not as effective as inline, but I think good enough. The author's inline version as shown in the photo, was good enough to 1 GHz.

    image

  • Hello

     

    Thank you for the clear and detailed discussion on using FFT and also congratulations on a great instrument.  I will try the changes you recommend and update my review. Would you please give additional explanation of the hardware difference between conventional oscilloscope FFT and Spectrum Analysis on the MDO-2000E?  I am also curious if it is possible to buy a software key to update the 70 MHz oscilloscope to 100MHz or 200MHz.

    Regards,

    Frank

  • Thanks your great review.  

    Regarding FFT comparison, the MDO-2000E’s FFT points can be set up to 1M pts (the  FFT points refer to the number of sampling data points used to calculate the FFT), and the Keysight DSOX1000A’s FFT points are up to 65536 pts only.

       The number of FFT points of MDO-2000E in the above Test picture is set at 100k pts (upper left corner of the screen), the waveform sampling rate corresponding to Keysight's FFT screen is 100MSa/s (the lower half of the picture below), and MDO-2000E (the upper half of the picture below) The sampling rate is 500MSa/s, which means that Keysight's setting FFT at that time can see the range is DC~50MHz (half of sampling rate), and MDO-2000E is the range of DC~250MHz (set in a wider range to see the partial zoom).

      Only the partial FFT data is displayed on the screens of both DSO, but MDO-2000E is relatively under the condition of more magnification (as can be seen from the memory bar at the top of the screen).

    We recommended to lower the MDO-2000E horizontal range by two scales. And change the FFT menu of horizontal scale to enlarge for comparison.

    Compare with Keysight’s 65536 pts FFT, MDO-2000E 1Mpts FFT can see the more fine parts of the signal in the frequency domain.

  • Hi Shabaz,

     

    The 50 ohm load arrived (cheap one from Amazon - I imagine I could have cobbled something together as good).

    image

    The user manual says the following:  When the setting unit is dBm, connect a 50 Ohm feed through termination on BNC.

     

    So I did and set up a sine wave 500Hz 1Vpp which after inserting the feed through load gives 0.5Vpp or 0.25Vpeak.  If I am doing the math right I should get approximately -2 dBm.  The SA gives -2.4dBm as shown below so it is off 0.4 dBm from expected.  Does that seem within the realm of expectations?

    image

    Edit:  I don't see in the specifications where it gives this.

  • Thanks - that is a really good explanation of why the FFT has problems and limitations

  • Hi Frank,

     

    Traditional' scope FFT methods are quite poor, because they are reliant on the current sample rate, and the trigger rate. If you're (say) looking at a 10Hz signal, then it may only be triggering (say) 10 times a second, and it will be doing a measurement of the spectrum based in what was captured, so there's a very slow update rate. Also, everything gets squished up usually, because the 'scope displays the FFT for the set sample rate, so if you're sampling at a high rate, then if you're only interested at your 10Hz fundamental frequency (as an example) then it's not visible, it's buried in the left side of the display.  And it still will only trigger at the slow speed, so update rate is still low. It means that overall, very little data is captured, and most of the time is wasted. There are other issues too (like poor dynamic range when sampling infrequently), but in a nutshell everything is stacked against getting a good spectrum that way. There's (almost) zero benefit to the feature apart from showing the difference : ). One minor benefit is that it is easy to show the time domain simultaneously, but the FFT display is so poor is it really worth it : ) With the real-time spectrum analyser, the data is captured all the time continuously, so there's no waiting on a time-domain trigger as such, and so there could be thousands of times faster updates possible, provided the data can be processed. Also, contrasting to a normal RTSA, as you mention, since the GW Instek is using a 'scope input, it can work to audio frequencies and below (whereas a normal RTSA intended for RF would likely have some minimum like 9 kHz or so).

  • Hi Shabaz,

     

    Below is the slide from a GW Instek file found on the internet that got me interested in testing in the audio frequency range:

    image

     

    The explanation below is from marketing material which I find confusing:

     

    Compared with the general spectrum analyzer, the spectrum function of MDO-2000E series can test below~9kHz signals, which is

    applicable to the frequency domain analysis of audio frequency and vibration. MDO-2000E series can also test the frequency domain

    signal with DC component without damaging the instrument. With respect to frequency domain waveform display, MDO-2000E series,

    featuring the same capability of a real-time spectrum analyzer, is faster than the general spectrum analyzer. Why? It is because MDO-

    2000E series utilizes digital circuit and software to calculate FFT. The general spectrum analyzer can only process the signal of a narrow

    frequency bandwidth at a time by frequency sweeping. Each sweeping will take several ms to dozens of ms. Hundreds and thousands of

    frequency sweepings are gathered to form a spectrum. Therefore, the displayed spectrum is not obtained at the same time. MDO-2000E

    series obtains spectrum display at the same time by utilizing digital circuit and software to calculate FFT that is faster than the frequency

    sweeping method. The FFT settings of oscilloscopes are based upon horizontal scale (sample rate) setting, which is totally different from

    the frequency range setting of MDO-2000E series. Most instruments will have insufficient frequency resolution due to insufficient FFT

    points while conducting spectrum measurement by FFT. Compared with the FFT of oscilloscopes, MDO-2000E series satisfies users with

    signal measurement requirements under 9kHz; a better setting interface, measurement resolution and measurements speed.

     

    It seems to say that the SA is a kind of FFT but different than the normal FFT on an oscilloscope in the horizontal settings.

  • Thanks!  Yes, it packs a lot into one package for those tight on space...