FAQ - EMCview V1 - 8
FAQ - EMCview V1 - 8
FAQ - EMCview V1 - 8
Index
QUESTION
Which CISPR standards are relevant to set up projects in EMCview?
ANSWER:
EMCview already contains many pre-defined project. In most cases it is as simple as connecting your DUT
to a LISN and the spectrum analyzer, start EMCview, select the applicable standard / project file and click
the start button to run a measurement.
The spectrum analyzer parameters have been chosen in order to cover many different analyzer brands and
models.
With respect to spectrum analyzer settings, you have to look primarily at CISPR 16-1-1, CISPR 16-2-1,
CISPR 16-2-2, CISPR 16-2-3 and CISPR 25.
With respect to emission limits, you need to figure out which of the standards CISPR 11, 13, 14, 15, 22, 25,
32 are applicable for your DUT. Most limit lines for conducted and radiated noise are already predefined in
EMCview. EMCview segment files are specified to be on the safe side, rather than to achieve the fastest
measurement speed. However, with good knowledge of CISPR 16 and depending on the properties of your
DUT and on the capabilities of your analyzer model, you can optimize segment files to speed measurement
time.
Below a few screenshots of relevant CISPR 16 and CISPR 25 details. CISPR xx and EN 550xx standards
are equivalent. The full standards need to be purchased.
Standards can be obtained in the web shops of national standardization institutes and various other sellers
in the internet.
There are also free sources as some countries provide free access to national standards which often are
adopted CISPR standards. They may however not always match the latest edition of the relevant base
standards. For example:
https://law.resource.org/pub/in/bis/manifest.litd.9.html
CISPR 16-2-1 / D.2.1 Scan rates and measurement times for AVG detector
2 Measurement time
QUESTION 1
Measurements take very long. Can´t it be accelerated?
ANSWER 1:
In short: Yes, measurement times can be accelerated. However it requires good knowledge of the CISPR
standards and of the technical specifications of your spectrum analyzer. If the sweep time is too short, the
measurement result may not be correct. You can change the sweep time easily by modifying the segment
files of your project.
The implementation of sweep times of EMCview is conservative to put you on the safe side and to find a
common denominator for all the supported spectrum analyzer models.
Things you need to consider:
CISPR 16 standards
DUT properties – are its emissions mainly CW or are there also slowly fluctuating or short sporadic
emissions?
Spectrum analyzer properties such as the number of display points / frequency steps per sweep, sweep
settings for its Quasi-Peak detector
The measurement time is dominated by sweep time settings. Next, the number of segments could be
reduced by matching it with the number of display points of your analyzer. Less segments means less
sweeps and consequently less time for initializing sweeps and read out measurement data via SCPI.
However the gain in speed is relatively small, compared to what you can achieve with faster sweep times.
Detailed answer:
We were considering following points when choosing the sweep time settings in the segment files:
1) CISPR 16 which specifies measurement equipment, measurement equipment settings and test set
ups is done for measurement receivers mainly. There are also some specifications for spectrum
analyzers, but for really compliant measurements you must use a measurement receiver. FFT
based measurement / real time spectrum analysis can considerably speed up measurement times.
Optimizing measurement speed for FFT based analyzers is not considered, as the documented
technical properties of various analyzers may not be detailed enough.
2) Measurement equipment settings also depend on the properties of the DUT. A DUT which
predominantly produces CW emissions can be measured with faster sweep times than a DUT
which produces random disturbances. E.g. a device with a microcontroller that carries out various
tasks and turns peripherals on/off depending on firmware tasks; a fast measurement could miss
noisy events. In practice, the engineer in the test house would observe the DUT emissions for
fluctuations first and then determine measurement time.
3) It must be clear, it is not a target of EMCview to carry out 100% CISPR compliant measurements.
First of all, we use a low cost spectrum analyzer and not an expensive measurement receiver. We
are carrying out pre-compliance measurements to get an idea, if our DUT behaves well and most
likely would pass type approval in the test house. Else, we would improve and re-test the design of
the DUT, until we are confident to carry it to the test house.
4) We want to make pre-compliance measurements as simple as possible for.
5) EMCview has to cover many different models of low cost spectrum analyzers.
6) We want to find a “common denominator” for the settings in order to cover analyzer models with
different specifications
7) We want to make the settings to cover as many possible noise scenarios
In order to accomplish the points above, we tried to make a general table for measurement settings based
on CISPR 16 specs. These meant, that we had to find the most applicable settings and consequently the
sweep time is not optimized for every single case. You can measure too fast to be compliant, but you can´t
measure too slow.
QP detector measurement times are rather influenced by SA model performance than by CISPR specs.
Typically, according to CISPR 16 we could measure faster, but we were looking at the sweep time auto
settings of the analyzers, based on RBW and sweep frequency range. Using CISPR 16 parameters, would
make the UNCAL message pop up at many spectrum analyzers.
(easier to write down, copy and paste, etc. than 150 kHz – 2.7 MHz, 2.7 MHz – 5.4 MHz, 5.4 MHz – 8.1 MHz,
8.1 MHz – 10.8 MHz .....)
Optimizing sweep range to every analyzer would not significantly increase measurement time. The time
reduction would come from less segments requiring less time to set up the sweep and read out the
measurement results via SCPI.
Our sweep time selection is based on the following table:
It is conservatively taking values from various CISPR tables and selecting the slowest sweep times as common
denominator. Anyhow, every user can easily modify it upon his own considerations.
QUESTION 2
EMCview uses the default settings for measurement points per sweep. Can´t I accelerate the measurement
time by changing the settings to a higher number of measurement points per sweep?
ANSWER 2:
Any manual settings of the spectrum analyzer will be overwritten by an initialization sequence of EMCview
to ensure a defined state prior to starting any measurement.
Furthermore, the dominating factor is the sweep speed in [MHz/s] specified by the standards. Changing the
number of measurement points per sweep would enable larger sweep segments, but the overall
measurement time would remain unchanged, as the sweep speed in [MHz/s] cannot be changed. There
would also hardly be any gain in speed with respect to the time needed in between segments to download
the measurement data from the analyzer. Though fewer segments would mean less breaks in between the
sweeps, the time to download data would increase due to an increased number of measurement points or
in other words – fewer breaks, but longer breaks. A disadvantage of segments with a wider frequency span
is increased measurement time, when measuring single segments.
To summarize – increasing the number of measurement points will not result in reduced overall
measurement time.
3 Segment files
QUESTION
In the software, if I want to select a different STEP frequency, e.g. 5 KHz (inside the CN RBW) to get more
information or 20 KHz (more than CN RBW) to speed up the measuring, how can I do?
I think that I must change the low and high limit for each SEG1 band that compose the test configuration
file right? But how many points for each band to increase or decrease the frequency step? Can you explain
better with a little example?
ANSWER:
Inside EMCview segment files you specify a series of single sweeps.
Example:
.....
FRQa__1=150.000
Author: Mayerhofer 6 of 33 TekBox Digital Solutions 2-Jun-21
FAQ_EMCview V1_8.docx V1.8
FRQb__1=2.500.000
BW____1=9000
Sweep_1=25000
Att___1=0
PreAmp1=off
Detector1=VAV
FRQa__2=2.500.000
FRQb__2=5.000.000
BW____2=9000
Sweep_2=25000
Att___2=0
PreAmp2=off
Detector2=VAV
.....
In the above example, each segment has a start and stop frequency. With exception of the very first
sweep, we always use 2.5 MHz sweep bandwidth between 150kHz and 30 MHz. Reason is that CISPR16
specifies two adjacent frequency points shall not be separated more than half of the RBW. Now CISPR 16
specifies in parallel the RBW for the frequency ranges 9kHz to 150kHz (RBW 200Hz), 150kHz to 30MHz
(RBW 9kHz), 30 MHz to 1GHz (RBW 120kHz) and above 1 GHz (RBW 1 MHz).
Coming back to the above example, the step size shall be 9 kHz / 2 = 4.5 kHz. Given the default number of
measurement points of Rigol analyzers per sweep, 601 points, you can make a single sweep not larger than
(601-1)*4.5 kHz = 2.7 MHz. For easier segment file creation, I use 2.5 MHz per sweep. Siglent has a larger
display and hence 710 points per sweep. Means, in order to be compliant with CISPR 16, you can make
sweeps of (710-1)*4.5 kHz = 3.2 MHz. You could make segments of 150 kHz – 3.35 MHz, 3.35 MHz – 6.55
MHz, 6.55 MHz to 9.75 MHz and so on all the way up to 30 MHz. This would be perfectly compliant with
CISPR 16.
In our segment files we use 2.5 MHz segments, because the segment file is easy to read and it is CISPR 16
compliant for all analyzers supported by EMCview plus some margin.
If you want any other step size, e.g. 1 kHz and you use a Siglent with 710 measurement points, simply
change the segment file using a text editor to:
......
FRQa__1=150.000
FRQb__1=860.000
BW____1=9000
Sweep_1=25000
Att___1=0
PreAmp1=off
Detector1=VAV
FRQa__2=860.000
FRQb__2=1.570.000
BW____2=9000
Sweep_2=25000
Att___2=0
PreAmp2=off
Detector2=VAV
......
However, CISPR 16 specifies one more parameter – the sweep time. For the band 150 kHz to 30 MHz,
CISPR 16 specifies a sweep time of 100 kHz/second. This is why for 2.5MHz sweep segments, the segment
file sets the sweep time to 25000 (=25000ms = 25 sec). Means that you could change the above segment
file to:
......
FRQa__1=150.000
Author: Mayerhofer 7 of 33 TekBox Digital Solutions 2-Jun-21
FAQ_EMCview V1_8.docx V1.8
FRQb__1=860.000
BW____1=9000
Sweep_1=7100
Att___1=0
PreAmp1=off
Detector1=VAV
FRQa__2=860.000
FRQb__2=1.570.000
BW____2=9000
Sweep_2=7100
Att___2=0
PreAmp2=off
Detector2=VAV
......
The sweep is faster now, but you have more segments. Neglecting the time needed to read out the results
of a sweep and starting a new sweep, the total sweep time remains the same. It would be no different, if
you would increase the step size. Your sweep bandwidth becomes larger, but since you have to adjust the
sweep time, the time would remain the same. At least unless you want to operate outside CISPR 16 specs,
which you already would do by increasing the step size above RBW/2 anyway.
I have seen that is it possible into the SEG file to change the time sweep (expressed in ms) but they are very
high value (500000ms for QP Detector or 200000ms for AVG Detector). I have tried some tests and
checked that if more long is the value and more detailed are the result of final sweep. From you side what
are the best value to use (compared to EMI Receiver) for Peak, Qpeak and Avg detectors?
ANSWER:
A part of this question is already covered by the answer to the previous question. Of course CISPR 16
sweep times differ depending on frequency range and detector type. Have a look at our CN application
notes for more details.
When it comes to QP sweep times, we could use any sweep time equal or higher specified in CISPR 16.
Now the sweep times specified for QP detectors specified by CISPR 16 would be too fast for the Rigol and
Siglent analyzers. Using CISPR 16 QP sweep time would make an UNCALIBRATED message or icon pop up
on the Siglent or Rigol analyzer. Uncalibrated means, that the sweep time is too fast for the chosen
resolution bandwidth and sweep bandwidth. The filter would not have sufficient time to settle and the
displayed amplitude would be lower than in reality. You would also see a frequency shift due to the delay
of the filter response. Of course, you can change any settings in the segment file according to your will,
however watch for the UNCALIBRATED message/icon on your analyzer.
Question:
There is a term called “SPAN” which I do not understand. Normally span is regarded as the total frequency
coverage from start to stop and is often used when a centre frequency is selected and the span then prescribes
the frequency range which will operate about that centre. I do not see this as the case in the software so
maybe, as is often the case with the EMC View software the terminology is not correct and it applies to
something else.
Answer:
I assume your question refers to the entry box below
This is explained in chapter 10 of the manual. The “span” refers to the span of the final scan/peak re-
measurement. It has no influence on standard measurement mode. When re-measuring peaks, it may
happen that the frequency of the peak has drifted away in the meantime. This is what we observed with
other EMC tools, which simply make a single frequency peak measurement. When we measured power
supplies, we saw peaks disappearing due to the temperature related frequency drift of the switched mode
regulators free-running oscillators. Consequently we developed EMCview and specified a “span” where
EMCview sweeps around the frequency of the peak frequencies in order to find the maximum of potentially
drifted peaks.
Question:
I want to manually set the video bandwidth upfront to a measurement, however it always jumps to VBW “auto”,
Answer:
We can´t enable setting VBW manually upfront to the measurement. As we need to have a defined state for
the initialization of the measurement, we issue a preset command to the analyzer first. Consequently the VBW
BW gets preset to auto and once in remote mode, there is no manual access anymore, except by pressing
the escape button which would abort communication between EMCview and the analyzer.
With EMCview REV 5.xxx you will be able to choose between VBW = Auto and VBW = 3*RBW.
Question:
I changed the detector type and the sweep time, but it did not have any effect on the settings of the
analyser when running a measurement
Answer 1:
You most likely changed the settings for peak re-measurement and then ran a “Set 1” or “Set 2”
measurement.
These settings are for re-measuring peaks (or doing a final scan as it is called sometimes). There is a
chapter in the manual, explaining peak re-measurements.
In order to change the settings for “set 1” or “set 2” measurements, you need to change the settings in the
segment files. This can either be done with any text editor or with the built in editor in the SETTINGS menu
of EMCview.
Answer 2:
If you change a segment file, while EMCview is open, you have to press the “Reload Project Files” button in
the right upper corner of the window, after saving the modified segment file.
Answer 3:
If you set a spectrum analyzer parameter manually, before starting EMCview, it will be reset to default when
EMCview starts. Means, manual spectrum analyzer settings will not have any effect. EMCview needs to
start from a known configuration of parameters to ensure consistent measurements.
7 Graph / Traces
Question:
If I want to compare some traces in the same windows, I need to save each sweep in CHART mode? And is it
possible to load the previous sweep and make a traces compare?
Answer:
Go to the FILE/UTILITIES menu and click SAVE CHART. You then can load up to three charts as reference
traces in FILE/UTILITIES/LOAD REFERENCE TRACE/TRACE1,2,3
You would then see your latest chart plus 3 loaded charts (reference traces) in the graph window. In total
you could see up to 8 curves – 4 charts composed of two graphs (set1/set2). The selected traces will then
be listed on top of the graph window. By clicking on either the chart name or trace name, you can move it
into the foreground. You can also compare curves against each other using the math function. E.g. subtract
two curves from each other to explicitly show the difference.
Question 1:
When I make measurements, the analyzer beeps and displays the warning ADC overload or IF overload. What
is the reason?
Answer 1:
ADC overload or IF overload warnings report that the input of the spectrum analyzer is overdriven. Overdriven
does not necessarily mean that the maximum permitted RF input power is exceeded.
If the internal attenuator setting is 0 and/or the pre-amplifier is on, strong input signals can drive the IF or ADC
section into saturation. Increase the internal attenuation and / or turn off the pre-amplifier in the EMCview
segment file.
Question 2:
When making conducted noise measurements with a Tekbox RF-current probe I get ADC / IF overload
warnings? However I do not see any noise with excessive amplitude.
Answer 2:
The useful frequency range of Tekbox RF-current probes starts in the very low kHz range. You may
measure an automotive switched mode regulator, starting conductive noise testing at 150 kHz according to
CISPR 25. Have a look at the frequency range below 150 kHz. Your switched mode regulator may produce
sub – harmonics, which you are not aware of. Sub-harmonics typically have high amplitude. The RF current
probe picks it up and the amplitudes are high enough to saturate the spectrum analyzer. Use a 150 kHz high
pass filter to suppress the sub harmonics, without losing dynamic range in the main frequency range of your
measurement.
Question:
When I make measurements using a pre-configured project, even with the DUT not powered, I see that the
base noise at very low frequencies is higher and then decreases towards higher frequencies. Is this normal?
Answer:
There are two reasons, why you see higher noise at very low frequencies. The architecture of spectrum
analyzers causes “0” Hz falling into the pass-band of the IF filter. As the filter curve does not have a perfectly
rectangular shape, sideband noise appears at low frequencies, decreasing with increasing frequency. This
noise behavior can also be observed without using EMCview. Second, when applying frequency response
correction for LISN or RF-current probes, there is increased transmission loss towards low frequencies. Means
conducted emission levels need to be compensated for the frequency response of the LISN or RF-current
probe. This compensation also shifts the base noise of the analyzer to higher levels, similar as you would see
on your analyzer in manual mode, when increasing the internal attenuator value.
Question 1:
I want to add some spurious which are close to the limits to the list of peaks?
Answer 1:
Set a margin line, so that these spurious cross the margin line. Next press display peaks and all peaks
crossing the margin line will be listed for consecutive peak measurement.
Question 2:
I want to manually add frequencies to the peak list?
Answer 2:
Press the MARKER button in the upper right corner of the EMCview screen. Select the measurement graph
to which you want to add a marker (Set1/M1 or Set2/M2). Place a marker at the frequency, which you want
to get added to the peak list. Press the COPY MARKERS button and the frequency will be copied into the
peak list for consecutive peak measurement.
Question 1:
I carry out conducted noise measurements on Line and Neutral, or on the positive and negative supply line.
How can I merge the two graphs to a worst case graph, showing the maximum spurious levels of both
measurements in a single graph?
Answer 1:
Carry out the measurement runs and save the results as chart. Click the MATH button in the right upper
corner of the EMCview screen. In the File menu, load both graphs as traces, e.g. the measurement result of
the positive supply line as Trace 1 and the measurement result of the negative supply line as Trace 2. Select
the two traces, using the mathematical operations box at the right side of the EMCview screen. Select MAX
as mathematical operation. Press the CALC button and a graph with the merged result will be displayed.
Question 1:
I figured out, that my measurements were done choosing wrong correction files. Can I avoid re-measuring
the set up and instead recalculate the graph with the right correction files?
Answer 1:
Yes, this can be done by using the DISPL. PEAKS feature. Searching peaks also re-calculates and re-
displays the graph. A chart file is composed of the graph data, plus limit lines and correction lines. However,
the chart file stores limit lines and correction lines as data only, not containing their file names. In order to
avoid overlaying newly chosen correction files with previous correction data, proceed as follows:
Start EMCview and load the chart that you want to re-calculate with new correction files. If you loaded the
chart, without loading a project before, all configuration file boxes will be empty. Next, click the RELOAD
PROJECT FILES button in the right upper corner of EMCview. Next click the DISPL.PEAKS button. This
action stripped any correction files from the chart and displays the raw data as received from the spectrum
analyzer.
Next, load the desired correction file and click the DISPL. PEAKS button. The resulting graph will then be re-
calculated using the right correction file.
Answer 2:
Starting with EMCview V5.44, you have the option to load the raw data of any previously saved chart. The
FILE / UTILITIES / LOAD CHART menu was updated to load RAW DATA only. The chart raw data can then
be combined with any correction file and re-calculated accordingly, using the DISPL. PEAKS feature. This
way, you don´t need to strip existing correction data using the process described in Answer 1.
Additional comment: You can load raw data of any chart, and then save this chart. The resulting chart file
will not contain any correction data then. Note, that such files can only be re-loaded as chart then. If you
would want to load it as trace, you had to re-calculate it with the DISPL.PEAKS feature while all correction
files are set to NONE and then save as chart. Such charts, despite not containing real correction data, can
also be loaded as trace.
13 Display Issues
Question:
I run EMCview on a tiny laptop and don´t get the complete EMCview window displayed on the screen. Is
there any solution?
Answer:
Create a desktop shortcut of EMCview.exe, then right click the EMCview icon, click properties, enter the
parameter [space]s to the link and click the Apply button.
Upon starting EMCview, scrollbars will added to the EMCview windows and you can scroll around to get
access to the complete EMCview window
Question:
I would like to automatically switch between antennas during an EMCview measurement run. Is this
possible?
Answer:
With EMCview V5.48 or higher, commands for ASCII controllable devices can be added to the segment
files. You can use a device such as KMTronic 4UCRB Relay Controller to switch coaxial antenna relays or
external attenuators.
Next select the COM port and set up the parameters for the serial interface of the relay controller or any
other serial device. Establish the connection and optionally transmit a string for testing/validation:
In order to automatically transmit test strings, they need to be entered into the desired segment file.
You can add two serial command strings to each segment. Command SCPI_A will be executed before the
sweep of the corresponding segment. Command SCPI_B will be executed after the sweep of the
corresponding segment.
Example: Segment file with commands for the KMTronic 4UCRB Relay Controller
Explanation:
SCPI_A_x with x being the segment number; command sent at the beginning of the segment
SCPI_B_x with x being the segment number; command sent at the end of the segment
SCPI_B_x is optional. A single command at the beginning of the segments would be sufficient to turn
ON/OFF a relay.
/$FF/$01/$01: serial string to turn on output 1 of the KMTronic 4UCRB Relay Controller. It will send the
characters FF 01 01. This string is not EMCview specific. Its syntax has to be chosen according to the
manual of the corresponding device. FF 01 01 means turn ON Output 1 of the 4UCRB. EMCview will send
whatever string is entered in this place.
/d1000: wait 1000 ms before starting the sweep. Maximum value: 9999 ms
Special control commands can be entered as /$XX (upper case, always 2 characters)
Example:
/$0A LF, Line Feed (not /$0a)
/$0D CR, Carriage Return (not /$cr)
Question:
How can I make EMCview remember the path where to save charts or reports?
Answer:
When saving reports or charts, use the windows file dialog to change the path, if required. However,
EMCview always first points to the default \EMCview\out directory. If you want to change the default
directory, enter the SETUP / OPTIONS menu and set whatever path you want:
Question:
Can I change the sweep step size from Hz to %, in Generator mode?
Answer:
By default, the step size of Frequency Sweeps is set to Hz:
In order to change the step size to %, enter the SETUP / OPTIONS menu. In MISC, tick the % sweep box:
Question:
I am not sure, which sign I have to enter for the values in the correction/compensation files. How does
EMCview calculate the amplitude values of the graph?
Answer:
Graph amplitude = Raw amplitude [dBµV] – cable file [dB] – LISN file [dB] + ant file [dB] – amp file [dB]
Answer:
ADC / IF overload:
An ADC overload event is only reported via buzzer beep and a message on the spectrum analyzer display.
None of the spectrum analyzers supported by EMCview provides any means to remotely query the ADC/IF
overload status.
And here the zoomed view. The amplitude values obtained from the analyzer are already compensated with
respect to the internal attenuation.
And here the zoomed view. The amplitude values (raw values) obtained from the analyzer are not
compensated with respect to the internal attenuation. EMCview compensates the raw values based on the
content of the amplitude file.
Question:
How shall I carry out pre-compliance radiated emission measurements?
Answer:
The majority of standards require radiated emission measurements using the quasi peak detector.
However, quasi peak measurements are extremely slow. It would take hours to carry out any radiated
emission measurement.
Let´s have a look at two measurement plots from a test-house to see how they proceed:
In the legend of the screenshot you can see QP limits, but the measurement was carried out using the peak
detector. Peak detector measurement results are far faster than QP measurements and Peak detector
Author: Mayerhofer 20 of 33 TekBox Digital Solutions 2-Jun-21
FAQ_EMCview V1_8.docx V1.8
measurements always deliver a higher value than QP measurements. As in the above case, the Peak result
is not crossing the QP limits, you know that the radiated noise test is passed and there is no need to make a
QP measurement.
Again QP limits and a peak detector measurement. Now, several spurious cross the QP limits and as QP
detector amplitudes are always lower than peak detector amplitudes, there is a chance that the spurious are
still below the QP limit line. Consequently, at the frequencies with the highest spurious levels, QP
measurements are carried out. You can see that the emissions in the 100 MHz and 250 MHz region still pass
QP limits. However, at frequencies from 150 MHz to 250 MHz, both peak and QP amplitudes are above limits
and the DUT does not pass the test. The procedure of first making a peak measurement and thereafter a QP
re-measurement at the critical frequencies saves a lot of time.
EMCview contains radiated emission measurement projects with QP limits and corresponding segment files
using QP detector scans. In order to create projects most suitable for your spectrum analyzer, take any of the
segment files, change the QP detector entry to Peak detector and reduce the sweep time.
EMCview radiated emission segment files are also split into various frequency bands, as changing antennas
according to CISPR 16 would be necessary. Furthermore, project files are split, considering analyzers may
cover 1.5 GHz, 3 GHz or 6 GHz frequency ranges.
Below a segment file example for a setup, using a TEM cell or a hybrid antenna, covering a frequency range
of 30 MHz to 1 GHz. Segment size is 25 MHz, considering spectrum analyzers with ≥ 600 measurement points
per sweep and CISPR 16 requirements for the maximum spacing between adjacent frequency points. RBW
is 120 kHz, sweep time is 10 seconds. Peak detector.
[Application]
Software=TekBox RP-W32-D7
Version=Demo
Date=30/01/2017 6:56:06 PM
[General]
Name=RN_30M_1G_PK.seg
[Data]
FRQa__1=30.000.000
FRQb__1=50.000.000
Author: Mayerhofer 21 of 33 TekBox Digital Solutions 2-Jun-21
FAQ_EMCview V1_8.docx V1.8
BW____1=120000
Sweep_1=10000
Att___1=0
PreAmp1=on
Detector1=POSPEAK
FRQa__2=50.000.000
FRQb__2=75.000.000
BW____2=120000
Sweep_2=10000
Att___2=0
PreAmp2=on
Detector2=POSPEAK
FRQa__3=75.000.000
FRQb__3=100.000.000
BW____3=120000
Sweep_3=10000
Att___3=0
PreAmp3=on
Detector3=POSPEAK
FRQa__4=100.000.000
FRQb__4=125.000.000
BW____4=120000
Sweep_4=10000
Att___4=0
PreAmp4=on
Detector4=POSPEAK
.
.
.
FRQa__37=925.000.000
FRQb__37=950.000.000
BW____37=120000
Sweep_37=10000
Att___37=0
PreAmp37=on
Detector37=POSPEAK
FRQa__38=950.000.000
FRQb__38=975.000.000
BW____38=120000
Sweep_38=10000
Att___38=0
PreAmp38=on
Detector38=POSPEAK
FRQa__39=975.000.000
FRQb__39=1000.000.000
BW____39=120000
Sweep_39=10000
Att___39=0
PreAmp39=on
Detector39=POSPEAK
The entire range from 30 MHz is covered with 39 segments with 10s sweep time each. Assuming another 10
seconds per segment for the transmission of the measurement values from spectrum analyzer to PC results
in a measurement time of 13 minutes. Critical frequencies can then be measured with QP detector using the
peak re-measurement feature of EMCview.
Question:
How are QP measurements carried out?
Answer:
Quasi Peak measurements require the QP detector to measure at least 1 second per measurement point.
This would result into sweep times of 601 measurement points * 1 sec per segment for Rigol spectrum
analyzers and 751 seconds per second for Siglent spectrum analyzers. Measuring an entire frequency band
using the QP detector would result in hours of measurement time.
Chapter 19 explains, how QP measurements can be substantially accelerated by making a scan using the
peak detector and then only investigate the frequency points, where the result of the peak detector violates
the Quasi Peak limits. Reducing the measurement to single spurious or frequency points of interest, reduces
measurement time to 1 second per frequency point. This is a net time however, as additional time is
consumed for remote control routines such as configuring the spectrum analyzer and reading out the
measurement result.
Example: Conducted noise measurement of a LED lamp according to CISPR 15; screenshots from
EMCview V8.0
A set up, consisting of a LED lamp, isolation transformer, TBLC08 LISN and a spectrum analyzer with EMI
option is controlled via EMCview over USB.
The lamp is powered, EMCview is started and USB communication with the analyzer is initiated. In
EMCview, the project CN_CISPR15_EN55015_Mains.prj is loaded and the measurement of Set1 is started
by pressing the Play button. After the measurement of all segments is completed, the measurement result of
the Average detector and the peaks violating the Average limits are displayed.
The peaks are displayed in the graph, using marker circles; furthermore the frequencies, amplitudes and
offset to the limit line are displayed in the peak list.
Peak markers can be removed and the peak list can be deleted by pressing the Clr Peaks button. They can
be restored any time by pressing the Displ Peaks button.
In case that a margin line is set by entering a margin in the project set up window, peaks will be referenced
to the margin line instead of being referenced to the limit line.
Also note, that any peaks related operation is always referring to the currently active measurement setup –
Set1 or Set2.
In the next step, a second measurement will be carried out using the Peak detector and applying Quasi-
Peak limits. Set2 is activated, on the top of the screen next to the Play button. Pressing the Play button
starts the measurement. CISPR 15 specifies this measurement to start at 9 kHz, using 200 Hz RBW up to
150 kHz and then switching to 9 kHz RBW all the way up to 30 MHz.
The Peak detector is used, though the standard specifies QP-limits, because of a significant reduction of
measurement time.
The result shows many peaks violating the QP limits. As explained in chapter 19, measurement results of
the Quasi - Peak detector are always lower or equal compared to the results of the Peak detector.
Consequently, there is the chance, that the spurious still may pass the QP-Limits, if measured using the QP
detector.
In the next step, a few peaks will be re-measured using the QP detector. Measuring at discrete frequencies
will save a lot of time compared to measuring the entire frequency band.
In the example below, the first five peaks are selected, using the left mouse button and CTRL key. After
selecting the peaks, press the right mouse button and two measurement options will become available.
Measure (consider drift) will measure a small frequency band around the peak frequency. The span is
configured in the project window under Drift Span. This measurement option deals with potential frequency
drift of the spurious. This may happen, when the spurious originate from switched mode power supplies,
where the switching frequency can drift over time and temperature. There is a time delay between the actual
measurement and the peak re-measurement, where the spurious frequency may have drifted and re-
measuring may miss the spurious or measure it at the shoulder, giving a wrong result.
The drift span does not specify a sweep, which would consist of 601 or more frequencies, taking 601
seconds or more to measure. EMCview divides the Drift Span into a series of discrete frequency points,
with neighboring frequency points being spaced RBW/2 (4.5 kHz in our example). Consequently, a 200 kHz
Drift Span results in 45 measurement points, rather than 601 or more measurement points. After carrying
out the series of measurements, EMCview displays the maximum peak within the Drift span.
Author: Mayerhofer 24 of 33 TekBox Digital Solutions 2-Jun-21
FAQ_EMCview V1_8.docx V1.8
Measure (ignore drift) will measure at exactly the frequency displayed in the peak list. Only a single
measurement will be carried out.
After the measurement of the five peaks is completed, the results of the quasi peak measurement is
displayed in the peak measurement box.
The QP Levels are furthermore displayed in the graph, as can be seen in the zoomed graph of the
screenshot below. As expected, the QP results are always in between the levels of the peak and average
measurement. The QP re-measurement shows that the first two spurious would pass QP limits with some
tiny margin.
The next screenshots looks at the frequency range around 15 MHz and selects the highest point by
manually placing a marker. First, pressing the Clr Marker button deletes the automatically generated
markers. Next a manual marker is placed at the graph and then transferred into the peak list:
Finally, a QP measurement is carried out, selecting the peak in the peaks list, right mouse button and
selecting Measure (consider drift). After the measurement, the QP measurement result appears in the graph
window and peak measurement list. The result is at the same frequency, which shows that the spurious
maximum has not drifted.
Note that the chapter is based on EMCview V8.0 or higher. This version simplified the UI with respect to the
peak measurement settings. Since the peak re-measurement is typically always carried out with QP-detector
and the attenuator and preamplifier settings should be identical with those of the initial measurements, the
corresponding entry boxes were removed from the project window. Measurement time is always 1 to 2
seconds, depending on analyzer model. Peak re-measurements now always use the QP detector and derive
their settings from the relevant segment files.
Question:
Is it possible to display more than two limit lines in a graph?
Answer:
With EMCview V7.0, every limit file can contain a main limit line plus up to 15 auxiliary limit lines. In
addition, EMCview can display a margin line for each main limit line. The main limit line will be used as
reference for margin lines and to calculate peaks, which exceed the major limit line or the margin line.
Auxiliary limit lines are basically graphic content which cannot be used for any processing purpose.
Main limit lines must be coherent, whereas auxiliary limit lines can be segmented / interrupted.
Example: CISPR22 Class A and Class B as main limit lines and FCC 15 Class A and FCC 15 Class B as auxiliary
limit lines
Colors:
Question:
How can I set peak markers to front/back?
Answer:
use CTRL-O to move circles to front or back
use CTRL-P to move crosses to front or back
use CTRL-M to move markers to front or back
use CTRL-J to move auxiliary limit lines to front or back
With EMCview V7.0, the structure of the SRC directory, which contains limit files, segment files, project files
and hardware related compensation files, such as LISN files, current probe trans-impedance files, antenna
factor files and signal gain/loss files changed to give an easier access to standard related projects.
Furthermore, all project files were updated and many new project files were added. All project files are now
documented.
The installation directory of EMCview 7.0 contains an additional directory with all documentation, such as
manual, FAQ document, application notes and project file documentation.
Question:
The fader does not work as expected. What is wrong?
Answer:
The Fader feature was implemented upon repeated customer requests and is based on the idea of making
an empty (DUT not powered) measurement to record the ambient noise and subsequently subtract the
ambient noise from consecutive measurements, with powered DUT. This idea sounds very tempting,
however it only works reliably with a static ambient noise spectrum. In reality, the ambient noise spectrum
is very dynamic. Main culprits are cellular networks, IOT, remote controls and many more sources.
Shielded set ups are the right things to combat ambient noise, but static signals such as FM broadcast can
be removed using the fader feature with some success.
In fact, below just the worst spurious were selected. These are displayed in blue color (activate the Cor box
in the View settings window in the lower right corner); the black curve shows the resulting zero
compensation curve:
In the File menu, save the zero compensation curve and thereafter load the zero compensation curve.
Disable the Cor box, if you want to make the compensation curve invisible. Then re-measure with the Use
Compensation box under the Play button enabled.
The curve below shows the result. The arrows point at the compensated frequency segments:
Activate the Clr box at the right side of the screen in order to pull windows across frequency segments where
you want to undo fading
25 History