3GHz Spectrum Analyzer - R&S® FPC1000 - Review

Table of contents

RoadTest: 3GHz Spectrum Analyzer - R&S® FPC1000

Author: gpolder

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?: Below is a list of comparable spectrum analysers.

What were the biggest problems encountered?: A few firmware bugs are detailed below.

Detailed Review:

document version 1.0 - Mar 17, 2018 - initial release

document version 1.1 - April 3, 2018 - bug fixes with firmware update to 1.3

document version 1.2 - April 30, 2018 - added figure to explain the detector

document version 1.3 - June 20, 2018 - added DDS-60 calibration for WSPR

document version 1.4 - Januari 31, 2020 - after my 3 Series MDO Mixed Domain Oscilloscope - Review I was convinced that I needed to increase my score of the FPC1000.

 

Introduction

In March 2013 I already opted for testing a spectrum analyzer, when Nicole Fusz asked what product(s) we would love to get our hands on? (http://www.element14.com/community/message/71435/l/re-hey-road-testers-what-products-would-you-love-to-get-your-hands-on#71435)

So I'm very happy that I was selected by element14 and R&S for reviewing the FPC1000. Thanks a lot.

 

The review is organised as follows: After the unboxing I will give some first impressions an explain how to setup the device. After explaining a little bit of spectrum analyser background, I will focus on how the FPC1000 compares to its competitors. The largest remaining part of this review will focus on experiments and projects with the spectrum analyser.

 

Unboxing

{gallery} Unboxing

image

Box: whats in there?

image

image

Additional small box

image

Additional small box

image

Power cables, USB cable and documentation.

image

Safety instructions, Getting started guide and instructions for installation of options.

image

......................

image

There he is!

image

................

image

Isn't he nice?

 

 

First impressions and setup

On the bottom the instrument is equipped with 4 steady rubber feet, from which the front feet can be folded out in order to place the device in a slanted position:

image

image

On the back are the power input (AC 100-240V), the power switch, USB, ethernet and trigger input (BNC):

image

There is no fan in the instrument, so it operates very quietly. To me this is a big advantage. image

 

After unpacking and placing it on my bench, I have powered it up (it boots up quickly!):

image

Within 20 seconds it comes up with a full span (0-3GHz) spectrum display:

image

 

As can be seen from the unboxing images there is not much documentation provided on paper. All documentation can be downloaded from:

https://www.rohde-schwarz.com/manual/fpc1000/ and firmware updates from: https://www.rohde-schwarz.com/firmware/fpc1000/ .

The latter learned that the latest firmware was not installed on the unit. Mine has v1.10 and the latest is v1.20. So before doing anything else, I upgraded the unit to the latest version. The firmware comes as a self extracting .exe file, which, although I don't have a windows PC I was able to extract using macOS.

 

After the upgrade I had a look at the system information HW/SW info and installed options which could be found in the setup menu:

{gallery} System information

image

HW/SW info

image

Installed options

image

Configuration overview

 

As you can see the 2GHz, 3GHz, FPC-B22 Preamp, FPC-B200 WiFi and FPC-K55 Advanced Measurements are installed, but not the FPC-K43 receiver mode and FPC-K7 modulation analysis which were mentioned in the RoadTest description. By the way, when playing with the device I found that after using it for a while the 'Controller Version' number changes to Vff.ff.ff.ff.

I have no idea what triggers the errornous display, also don't know whether it has any consequences.

image

During my tests I found another very strange issue. I changed the time in the setup to my local time, but unfortunately this change is not persistent. When powering the device of and on with the front button, it changes back to the original time. Also the WiFi connection settings are not persistent, I will come to that later.

 

The input socket is N type, since I mainly work with BNC or SMA, I bought a N-BNC adapter. I resisted the temptation to buy a cheap Chinese adapter. Such a high quality spectrum analyser deserves a high quality adapter, so I ordered this one (Rosenberger):

image

Another must have when using a spectrum analyser is an attenuator, which makes it possible to look at signals whith higher power as the max rating of the device (30dBm, or 1W in this case). I have to admit that in this case, I bought one in China. I'v choosen for a 20dB 25W attenuator, which makes it possible to measure signals up to 25W (43.98dBm). The maximum rating of 30 dBm is not reached in that case.

image

 

 

Summary: The FPC1000 is a very nice looking spectrum analyser with a large screen. It boots up very fast. It is whisper quiet since it doesn't have a fan. It looks like there are some bugs in the firmware.

 

 

Some Spectrum Analyser background

Before I continue the review, I will briefly explain some spectrum analyser basics, in order to give the reader some background knowledge.

With a spectrum analyser you can visualise signals, like with a oscilloscope.

  • Where an oscilloscope shows the signals as function of time, the spectrum analyser shows the signals as function of frequency. Note that all signals can be decomposed in a number of sine waves of different frequency and amplitude.
  • The amplitude of the signals is displayed on a logarithmic scale.
  • The signals are compared to a reference level, which is the horizontal line on the top of the screen.
  • The spectral resolution is determined by the resolution bandwidth filter (RBW).
  • The video bandwidth filter (VBW) filter determines the capability to discriminate between two different power levels.
  • The sweep time is determined by the total frequency range and the RBW.

 

More in depth information  can be found here:

 

Summary: Some spectrum analyser basics explained in order to better understand the rest of the review.

 

FPC1000 Options: Investment protection

Like many more companies, R&S uses a trick to be able to mass produce their equipment and still let the user decide what he wants to buy and pay for.

All possible features are built in the device hardware and firmware and can be unlocked by entering a license key for validation.

 

They have a name and a document on this: Investment protection –buy only what you need when you need it.

(170218_FPC_Option_Sheet_5214-9180-32.pdf)

 

Here is a full list of possible options and their prices in € (ex VAT). Options provided to the roadtesters are noted in bold.

Optiondescriptionprice (€ ex VAT)
FPC1000FPC1000Spectrum analyser 1GHz1530
FPC-B2FPC-B2Upgrade 2 GHz840
FPC-B3FPC-B3Upgrade 3 GHz840
FPC-B22FPC-B22Preamplifier update350
FPC-K7FPC-K7Modulation analysis upgrade690
FPC-K43FPC-K43Receiver mode upgrade490
FPC-K55FPC-K55Advanced Measurement upgrade490
FPC-B200FPC-B200WIFI Connection support upgrade 290

 

So for € 1530 you already can buy a SA, which is very competitive. Its about the same as other 'budget' devices (see also next section). When asking around most people prefer R&S above the Chinese brands. When a larger frequency range is needed, or some options, like for instance the preamplifier, the unit will quickly become more and more expensive.

 

Summary: The FPC1000 is rich of features, the user decides which ones he needs and likes to pay for.

 

 

 

FPC1000 compared to its competitors

R&S positions the FPC1000 as an entry level spectrum analyser. They advertise it as:

Outstanding quality and innovation does not have to come with a high price tag. The R&S®FPC1000 spectrum analyzer delivers unexpected performance at a budget-friendly price.

It looks like R&S wants to compete with Chinese spectrum analyser which nowadays are very popular with hobbyists and ham radio operators.

On the R&S website you can find comparisons between the FPC1000 and the Rigol DSA815 and Rigol DSA832E.

Here is a small table showing the key features of current budget spectrum analysers. I have to admit that I didn't have experience with most of them. I just took the numbers from the published specs. As I'm in Europe, I noted the european prices (€) ex VAT.

BrandSARange
Noise floor RBW rangeScreen sizeTG (***)Price (€)
R&SFPC1000FPC10005kHz-1GHz-146dBm1Hz-3MHz10.1"no1530
R&SFPC1000FPC10005kHz-2GHz-146dBm1Hz-3MHz10.1"no2370
R&SFPC1000FPC10005kHz-3GHz-146dBm1Hz-3MHz10.1"no3210
R&SFPC1000FPC1000 (PA)(PA) (*)5kHz-3GHz-165dBm1Hz-3MHz10.1"no3560
R&SFPC1000FPC1000 (full) (**)5kHz-3GHz-165dBm1Hz-3MHz10.1"no5520
RigolDSA8329kHz-3.2GHz-158dBm10Hz-1MHz8"no1962
RigolDSA832-TG9kHz-3.2GHz-158dBm10Hz-1MHz8"yes2454
RigolDSA8159kHz-1.5GHz-155dBm100Hz-1MHz8"no1775 ??
RigolDSA815-TG9kHz-1.5GHz-155dBm100Hz-1MHz8"yes1372
RigolDSA705100kHz-500MHz-130dBm100Hz-1MHz8"no628
RigolDSA710100kHz-1GHz-130dBm100Hz-1MHz8"no888
GW InstekGSP-730GSP-730150kHz-3Ghz-100dBm30kHz-1MHz5.6"no829
SiglentSSA30219kHz-2.1GHz-161dBm1Hz10.1"yes1440
SiglentSSA30329kHz-3.2GHz-161dBm1Hz10.1"yes2260
OwonXSA1015-TG9kHz-1.5GHz-150dBm10Hz-3MHz10.4"yes975

 

(*) Power Amplifier need to be ordered separately, see section above.

(**) A lot of other options can be ordered, here I give the price for all options installed, see section above.

(***) Tracking generator, a very valuable add-on for measuring filter or antenna characteristics.

 

As already said in the previous section, the entry level FPC1000 is very competitive, but with all options installed the price is quite high compared to its competitors. The screen size is larger than the Rigols and equal to Siglent and Owon. I expected a bit a touch sensitive screen, unfortunately this is not the case. Menu options needs to be selected from the buttons on the right side of the screen. Furthermore the menu on the right side is always visible, which makes the area where the spectrum is displayed smaller.

To my opinion there is one important feature missing namely a tracking generator. A tracking generator is indispensable when measuring filter characteristics, or antenna performance.

As you can see in the list most competitors do have a tracking generator. Sometimes it also is a feature which can be unlocked, but that is for the FPC1000 not the case. As there is no second RF connector on the front pannel It is quite clear that the necessary hardware is missing.

 

Summary: the entry level FPC1000 is very competitive. With all options installed the price is quite high compared to its competitors. A tracking generator is missing.

 

 

 

Display received signals

After connecting a small dipole antenna, tuned to 150 MHz, the spectrum analyser shows the received signals over the whole frequency range.

Here are some screenshots, first the whole range from 0-3GHz, than focusing on 0-1GHz, and finally the FM broadcast band 80-110 MHz.

 

{gallery} Received signals

image

image
image

 

 

A great feature of this spectrum analyser is that you can demodulate and play AM or FM signals at the marker position.

Below is a short video that demonstrates this feature. First a FM broadcast signal, followed by a HAM radio transmission in the 2 meter band on a normal radio, and then on the FPC1000. Note that the spectrum analyser can not show a spectrum and demodulate a signal at the same time. Therefore there is this time option in the menu, which defaults to 0.5s, meaning that after each 0.5s of audio a spectrum is drawn. In the video I increase this value to 10s, which makes the sound much better, at the expense of the interval of displayed spectra. Also noticeable is that the decoding of the small band amateur FM signal sounds much worse than the same signal on the radio.

 

 

A spectrum analyzer usually shows signal levels in the unit dBm, which is an abbreviation for the power ratio in decibels (dB) of the measured power referenced to one milliwatt (mW). If necessary, the FPC1000 can also select other units. In that case, the analyzer automatically converts the results into the selected unit, as described on page 90 of the manual.

Unfortunately this didn't work for Voltage (V) and Watt (W) as can be seen from the following screenshots:

 

{gallery} Different units on the Y axis

image

image
image
image
image

 

 

Summary: When connecting an antenna to the input all RF signals which surrounds us are displayed, it is even possible to listen to AM and FM modulated signals. The Y axis for W and V is not shown.

 

 

 

Measuring harmonics of HAM radio transmitters

One of the first applications in which I tested the FPC1000 is the measurement of unwanted signals from my HAM radio transmitters.

I have two handhelds around, the first one is a Teleport 9 from AEG. This is a solid and very robust portable transceiver for the 2 m band. Visitors of the main Dutch airport (Schiphol in Amsterdam) may remember these brick-size blue handies being used by all airport service personel. The AEG Teleport 9 was part of a trunking network used by all services at the airport and has now been replaced by a more modern (and smaller) alternative. The old units ended up in the HAM radio community and have been converted by Rob Spijker, PE1RJY, into amateur transceivers covering the entire 2 m band.

The second one is a cheap Chinese dual band handheld Baofeng UV-5B.

image

The measurements were very easy to perform with the FPC1000. I first selected the frequency range, put the RBW at 300kHz and the VBW at 100 kHz. I used the radios in low power mode (0.5W) using their own antenna and just picked up the signal with a small wire in the input of the SA.

Then by pressing the marker menu 'Mkr' and pressing the marker 1, marker 2 and marker 3 successively, the three strongest signals will automatically be selected. I started with the Baofeng at 145.475 MHz, where the main signal and the second and third harmonic were selected. I kept the markers the same for the AEG. Finally I again tested the Baofeng, but now at 430.475.

 

Using the measurements menu of the FPC1000, it is possible to very quickly measure the harmonics of a signal. When you start a measurement, the R&S FPC looks for the first harmonic of the signal (= the highest signal) in the frequency range you have defined. The number of harmonics it looks for depends on the number of harmonics you have defined. It then adjusts the frequency axis so that all harmonics are visible.

In addition, the R&S FPC also calculates the total harmonic distortion (THD). The THD is the root mean square of all harmonics in relation to the power of the fundamental frequency. Details can be found on page 65 of the manual. The last two screenshots below shows the harmonic analysis of my Baofeng radio, on 145 and 430 MHz.

 

Here are the screenshots:

{gallery} Measuring Harmonic distortion

image

Baofeng 145MHz: second and third harmonic too strong

image

Teleport 9: Nice clean signal

image

Baofeng 430 MHz: better performance than at 145 MHz

image

Baofeng 145 MHz: harmonic distortion measurement THD = -28dB

image

Baofeng 430 MHz: harmonic distortion measurement THD = -41dB

 

The pictures speaks for themselves, the Teleport 9 has a much cleaner signal than the Baofeng at 145 MHz. To be honest, the Baofeng is not compliant with the regulations. In the Netherlands regulations state that unwanted signals must be  -60 dBc (60 dB below carrier signal). As you see in the screenshots, for Marker 2 and 3 the FPC1000 nicely shows the frequency distance to Marker 1 as well as the difference in dB, dBc so to speak. The last screenshot above shows some other signals which are picked up by the SA. They are not transmitted by the radio. A future challenge might be to detect the source of these signals.

Another feature worthwhile to mention is that the FPC1000 detects overloads on the input signal. In that case  the upper right corner shows 'IF Ovl':

image

When configured, the "Beep On Power Overload" menu item in the "Audio" Setup, the R&S FPC even beeps when it detects an overload at one of its inputs. (Manual Page 39)

To be honest the overload detection is not well documented, Only chapter 14.14.9 (Remote Control) refers to IFOVL.

 

The harmonics tests above are not completely reliable, since they are affected by the frequency responce of both the transmitter antenna on the radio, as well as the receiver antenna (piece of wire) on the spectrum analyser. Reason is that I didn't have the attenuator available yet when starting the road test. Luckilly a few weeks before the deadline the postman brought the nice 25W 20dB attenuator which I have described above. That made it possible to repeat harmonics measurements by connecting the radios directly to the FPC1000.

imageimage

As can be seen in the screenshots below the result for the Baofeng is even worse than with the antenna. For the AEG Teleport 9 it is about the same.

 

{gallery} Harmonics tests with attenuator

image

Baofeng 145 MHz: harmonic distortion measurement THD = -36dB

image

Teleport 9 145 MHz: harmonic distortion measurement THD = -65dB

Now we are measuring with an attenuator, the output level on the main frequency also can be precisely determined.

For the Baofeng it is 12.3dBm + 20dB, is 32.3dBm, or 1.7W.

For the Teleport 7dBm + 20dB = 27dBm, or 0.5W.

image

 

Summary: Harmonics from transmitters can easily be measured with the FPC1000.

 

 

 

Exploring home automation signals on 433 MHz.

In this section I will describe some other useful features of the FPC1000. For this I did some measurement on wireless home automation transmitters. These transmitters work in the 433MHz range, and codes are sent by keying the transmitter on and of in the sequence of the encoded data.

Here are some pictures of the transmitters used. Two cheap handheld transmitters, the third one is a small transmitter used in my raspberry pi based home automation system. The latter is used to automatically control the lights in my house.

imageimageimage

 

Before showing the signals of the different transmitters, lets first explain how the spectrum is displayed. You can read this in detail in chapter 10 of the manual. The number of samples collected in a single sweep usually is very large, especially if the span is large. However, the display of the R&S FPC can display only a limited number of results on the y-axis, because it is limited by the display resolution (one pixel usually combines a large quantity of measurement points). Therefore, it has to reduce the data and combine measurement results to fit them on the display.

The detector determines the way the FPC1000 combines and displays the results for one pixel.

 

The following detectors are supported:

  • "Auto Peak" The auto peak detector displays both the highest and the lowest power levels that were measured in the frequency range covered by a pixel. The auto peak detector loses no information. If a signal's power level fluctuates (for example in case of noise), the width of the trace depends on the magnitude of the signal fluctuation.
  • "Max Peak" The max peak detector displays only the highest level that was measured in the frequency range covered by a pixel. The max peak detector is useful for measurements on pulses or FM signals, for example.
  • "Min Peak" The min peak detector displays only the lowest level that was measured in the frequency range covered by a pixel. The min peak detector displays sine signals with the correct level and suppresses noise. Therefore it is useful to find sine signals in the vicinity of noise.
  • "RMS" The RMS detector measures the spectral power over one pixel. In case of power measurements, the RMS detector always shows the true power of a signal, regardless of the shape of the signal. The RMS detector is best for measurements on digitally modulated signals because it provides stable and true power readings. In combination with a high sweep time, you can increase the display stability even more because the measurement time for each pixel increases. Noise measurements also provide stable results if you apply the RMS detector in combination with a high sweep time.
  • "Sample " The sample detector shows the last level that was measured in the frequency range covered by a pixel. The sample detector is useful for measurements in the time domain (span = 0 Hz) as it provides the only way to represent the timing of the video signal correctly.
  • "Auto Detector" The auto detector selects one of the detectors automatically, depending on the selected trace mode.

 

Figure 19 from the R&S Spectrum Analyzer Fundamentals – Theory and Operation of Modern Spectrum Analyzers Primer, nicely explains the different detectors, although in this figure the sample detector takes the first level instead of the last one as explained in the FPC1000 manual.

image

 

The screenshot below shows the signal of one of the handheld transmitters. Center frequency at 434MHz, span is 38.4 MHz. The default (auto) detector is the 'Auto Peak' which shows  both the highest and the lowest power levels per pixel.

image

As described above the 'RMS' detector is best for digitally modulated signals, so lets use that:

image

That gives a much beter view on the signal. With this very short sweep time you don't see the full occupied spectrum. And even when the sweep time is increased to 1s you still see the gaps caused by digital modulation.

Before continuing I need to mention a very useful feature of the FPC1000, namely its ability to show a second trace. It is not a second channel as with a dual channel oscilloscope. Both traces are based on the same settings, except the trace settings like the trace mode or the detector. You can use the second trace to compare, for example, two different detector settings. And thats exactly what I will show here. From the trace menu I selected the second trace and selected the 'Max Hold' as detector. Now the display clearly shows the bandwidth used by the different home automation transmitters:

 

{gallery} Bandwidth of home automation transmitters

image

EURODOMEST transmitter: worst (broadest) signal, also unwanted signal 15 MHz higher.

image

IMPULS transmitter: better signal than EURODOMEST.

image

Raspberry Pi system: Best signal of the three. ( the signal 5 MHz below the transmitter was from a different source)

 

For these signals the RBW I used above is to wide, so I repeated the measurements with RBW set at 100kHz. Furthermore I would like to show two other measurement functions which are very usefull for exploring these transmitters. First measurement of the ocupied bandwith, which shows how much bandwith is used for a certain percentage (default to 99%) of the output signal. Second function is the so called spectrogram, which shows an image of the spectrum intesity in time. In software defined this is called the waterfall. It is even possible to hold the specgrogram and look at the signal at specific timesteps using the cursor. The next video shows how this looks like:

 

 

 

Below are the screenshots. it is again clear that the signal of the IMPULS transmittor is superior to the EURODOMEST.

 

{gallery} Occupied Bandwidth Measurements

image

EURODOMEST

image

IMPULS

image

Spectrogram

 

 

Finally I like to show another neat feature of the device. When selecting the span menu you will have the choice between full span (0-3GHz), last span (previous one used) and zero span. In the zero span mode the device switches from the frequency domain, in which signals are displayed as function of frequency, to time domain, in which the signal is displayed as function of time, just like a normal oscilloscope, but at the selected base frequency instead. Here is the signal of the IMPULS and the EURODOMEST transmitter at different sweep times clearly showing the digital transmitted signal. As can be seen the slope of the EURODOMEST is steeper than the slope of the IMPULS transmitter, which explains the broader bandwidth.

 

 

{gallery} 434 MHz Zero Span signal display

image

IMPULS transmitter: sweep time 20 ms

image

IMPULS transmitter: sweep time 50 ms

image

IMPULS transmitter: sweep time 2 ms

image

EURODOMO transmitter: sweep time 2 ms, steeper slopes.

Summary: By using additional traces, different detectors, and time domain display options, the FPC1000 is a very convenient spectrum analyser for exploring 433 MHz home automation signals.

 

 

 

Exploring a LoRa sensor node

Next I looked at the signals coming from a LoRa sensor node. I got this one at the main Dutch Electronics trade show (https://github.com/YourproductSmarter/KISSLoRa-demo ).

The unit operates between 863 to 870 MHz (LoRa RF Bands | LoRa Frequency Bands | RF Wireless World ).

 

imageimage

Same as with the 433 MHz modules in the previous paragraph, I'v put the second trace on 'Max Hold' with a RBW of 10 kHz the minimal sweep time is 386 ms. Since the LoRa module is programmed to send a signal each 30 seconds, it took some while before I got the picture below, in which clearly can be seen which channels are used.

image

Summary: Monitoring the band for a large period using the Max Hold option,  gives an impression of signals transmitted by LoRa devices.

 

 

Measuring filter characteristics

A spectrum analyser is a great tool to measure filter characteristics, or antenna performance, but the spectrum analyser needs to be equipped with a tracking generator to do so. By linking the sweep of the spectrum analyzer to the tracking generator, the output of the tracking generator is on the same frequency as the spectrum analyzer, and the two units track the same frequency. More information can be found e.g.: https://www.electronics-notes.com/articles/test-methods/spectrum-analyzer/spectrum-analyser-tracking-generator.php.

As already stated earlier, unfortunately the FPC1000 doesn't have one, so end of the story.

Luckily not, another solution to show behaviour of a certain filter is by using a wideband noise source. This noise source generates wideband noise, ideally at a steady signal level. This can be used as a cheap substitude of a tracking generator.

BG7TBL designed a rather cheap wideband noise source, called "NOSE SOURCE" as mentioned on the PCB. This product can be bought easily on eBay. I bought one (version 2016-03-06) and tested it in combination with the FPC1000.

image

image

Lets have a look at the signals, using the RMS detector, and a Min Hold on the second trace clearly shows the difference between the noise level of the FPC1000 and the output of the noise generator. As can be seen the signal is reasonable flat until 1.5 GHz, but also above that frequency usable. I have to admit that I'v put the device on 10V in stead of 12. On 12V the PCB gets terribly hot presumably since the Ub voltage of the MMICs is 7V compared to the max rating of 5.5V in the specs. By putting the device on 10V the Ub voltage is in accordance with the specs.

The total power output on 10V is slightly less than on 12V but still very usefull.

image

 

When looking at the very beginning of the spectrum, 0-200 kHz, there are some peaks, probably from the switching power supply on the PCB.

The level is still below 10 dBm, so for the FPC1000 with its max input value of 30 dBm not an issue. For other spectrum analysers I would advice to use an attenuator.

image

Here is the noise floor of the FPC1000, for comparison.

image

When looking at a small part of the spectrum, in this case 100-200 MHz, the spectrum looks rather nice and flat.

image

So my conclusion is that this NOSE SOURCE is a very useful addition to the FPC1000.

 

The first thing I'm very curious about is the frequency responce of my  attenuator. So I connected the noise source directly to the FPC1000, measuring RMS and put the first trace at MAX HOLD. Then I placed the attenuator in between, looking at the second trace. The second trace reasonably folows the first at a 20dB lower level, indicating that the attenuator works quite well over the whole range.

 

image

 

Now lets have a look at some filters:

The first filter is a very simple yet powerful notch filter for instance to suppress harmonics from a transmitter. It consists of a single open ended piece of  ordinary coaxial cable at a quarter of the wavelength you like to quash.  These filters will generate sharp, deep notches which can be placed on the harmonics of your transmitter. To test its performance I'v put a BNC T adapter on the output of the noise source and connected a small piece of coax.

imageimageimageimage

The frequency of the notch can be calculated by:

image

 

Where f is he frequency in MHz, Vp the velocity factor (Coax Velocity Factor | Coaxial Cable Velocity Factor | Radio-Electronics.com ) for the coaxial cable (here 0.66) and L the length of the coax in meters.

For the coax stub below the frequency should be about: 0.66*300/(4*0.19) = 260 MHz

image

The result is really close:

image

Now lets make the cable a little bit longer by connecting the SMA socket.

f = 0.66*300/(4*0.20) = 248 MHz

image

image

And yes the frequency is indeed 12 MHz lower than the previous one.

For more information, please have a look at the website on Coaxial Stub Filters from G4SWX.

 

Subsequently I measured 4 bandpass filters from my SDR radio. Below is a picture of the project. It is a Teensy based stand-alone SDR, using an old softrock rx/tx as frontend. There are 4 filters covering mainly all HF ham radio bands.

imageimage

Measurement setup:

imageimage

 

Here are the results for the four different filters, picture of the filter, filter response between 1 MHz and 100 MHz, and filter response between 0 and 30 MHz:

 

{gallery} Filter responses SDR bandpass filters

image

image
image
image
image
image
image
image
image
image
image
image

 

Summary: Unfortunately the FPC1000 doesn't have a tracking generator, but using a broadband noise source still makes it possible to measure filter characteristics.

Calibration a DDS-60 VFO for WSPR transmissions

I created a Weak Signal Propagation Reporter (WSPR) transmitter using a Arduino Mega, a GPS receiver and a DDS-60 VFO with AD9851 DDS.

Moreless similar too: https://www.george-smart.co.uk/arduino/arduino_ad9851/

In its prototype phase it looks like:

image

 

Here is the DDS-60 VFO which is used to generate the RF signal:

image

For WSPR the exact transmit frequency needs to be very precise. Unfortunately due to tolerances in the DDS-60 reference oscillator the frequency can be off for a few kHz's. The reference oscillator is 30 MHz, which is internally multiplied by 6 to 180 Mhz in the AD9851. So a deviation in the reference oscillator is also multiplied by 6. Using the FPC1000 I easily could measure the reference oscillator, just by picking the signal from the air with a small piece of wire. Here is how it looks like:

image

You clearly see that it is off with 2.369 kHz. so the reference oscillator is off with 6 times that value (14 kHz) which is quite a lot.

So I got the DDS-60 back on track by changing the DDS_REF value in the arduino sketch:

 

// DDS Reference Oscilliator Frequency, in Hz. (Remember, the PLL).
//#define DDS_REF 180000000
#define DDS_REF 179985786 // calibrated by measuring Xtal frequency

..........

 

Summary: Got my DDS-60 WSPR station back on track by measuring the reference oscillator and adapting the software accordingly.

 

Remote operation

To use the R&S FPC remotely, you have to establish a connection via the LAN, Wi-Fi or USB interfaces of the R&S FPC.

There are various tools that allow you to operate the  FPC1000 remotely from another computer or tablet:

  • The R&S InstrumentView allows you to remotely control the instrument remotely. It provides printing service, customize report generation and editorial functions on the different datasets used in the instrument. It also provides post-processing features to edit the markers position on the trace measurements. You can download the software, including a user manual, from the Rohde & Schwarz website. Unfortunately InstrumentView is not available for Linux or macOS.
  • R&S MobileView is an app that allows you to remote control the R&S FPC from a tablet or smartphone. It basically works like the remote desktop application built into the R&S InstrumentView software. You can download the app from the Google Play Store (Android based devices) or the AppStore (iOS devices).
  • Remote operation with SCPI commands is a method to automate measurements by writing scripts or programs that contain all instructions required for the measurement. When you start the program, the measurement is done automatically from start to finish. Manual configuration of a measurement with function keys, softkeys or apps is no longer necessary. Remote operation is thus a useful tool for measurements that you perform repeatedly and for which you know the configuration and for which the configuration is always the same. Testing devices during production is an example for such measurement tasks. Each device needs to be tested under the same conditions. There are several tools that allow you to connect to an instrument, send an individual (or more) command(s) and compile automated test sequences.
    • One of these tools is delivered with the R&S VISA library that is available for download on the Rohde & Schwarz website. The VISA library itself provides input and output functions that are required to communicate with the instrument. Unfortunately the macOS version is lagging behind the Windows version (5.5.6 - August 2016 vs 5.8.2 August 2017) a Linux version is not available. I didn't test it, because for my applications I don't need automated test sequences.
    • Another tool with more complex scripting functions is R&S Forum, also available for download on the Rohde & Schwarz website. It allows you to run and edit example script sequences and to write your own script files, as well as on-the-fly remote control of instruments. Script files can range from simple command sequences (Winbatch syntax) to complex programs using the programming language Python. Unfortunately Forum is not available on macOS and Linux.

 

I quickly checked the LAN and WiFi connections. After connecting a network cable to my home network, the device got a ip number assigned which is shown in the setup menu.

For WiFi connection I had to select my home network and after entering the password a connection was established. Password must be entered on the numeric keypad, the same way as we sent text messages from our phone in former days. Not the most convenient way to input text unfortunate.

Even worse is that the WiFi connection information is lost after a power cycle. For a tool that is priced at € 290,- I wouldn't expect that. Possibly this is a firmware bug. This was fixed in firmware release 1.3.

 

Chapter 4.2 from the FPC1000 version 1.2 release notes "Enabling Firmware Options via the R&S ® License Manager" explains how to connect to the device in order to install firmware options by activating licenses.

image

Unfortunately this did not work as you can see the application was not able to connect to the device. This was from Safari on macOS 10.13. Tested it also from Firefox and Google Chrome, no luck. The release notes state that mixed-mode scripts, or  "scripts from unsafe sources" needs to be enabled, but that option is not available any more in modern browsers. In the end I found a solution, by starting Google Chrome in unsafe mode from the terminal window:

 

open /Applications/Google\ Chrome.app --args --allow-running-insecure-content

Lucky me, I succeeded:

image

Not secure anymore as you can see in the address bar.

 

 

Summary: For Windows users there are plenty tools available to control the FPC1000 remotely. macOS and Linux users are a bit out in the cold, since only an old version of VISA is available for macOS. The fact that Wi-Fi connection information is not stored when the machine is powered off is extremely awkward.

 

 

List of firmware bugs

 

During the road test I found a few firmware bugs in version 1.2:

  1. Controller version number sometimes changes to Vff.ff.ff.ff.
  2. WiFi settings can not be saved. (fixed in firmware version 1.3)
  3. Time can not be saved. (fixed in firmware version 1.3)
  4. No vertical axis for Voltage (V) and Watt (W).

 

I'm quite confident that the will be fixed in future releases of the firmware.

 

Update:

On March 26 firmware version 1.3 came out, at the same time the FPC1500 was announced.

In this version the second an the third bug are fixed, the first and the fourth still persists.

 

 

Conclusion

I had a great time reviewing this spectrum analyser and I will continue to work with it in some other applications as well. I'm planning to keep this document updated. If you have questions, please put them in the comments below. Also recently R&S promised us to enable the K43 Receiver mode option, but at the time of this writing we didn't receive anything. When we will get that, I will test it and update the review accordingly.

I'm also planning to compare two VFO's for SDR usage. The first one is a DDS using the Analog Devices AD9851, the other one is based on the very popular Silicon Labs SI570. As I'm traveling next week and also the deadline for the review is by the end of next week, I decided to publish this first version now. Please let me know in the comments below whether you have questions. If possible I will cover them in future versions of this review.

 

 

Bottom line:  The FPC1000 is a high quality spectrum analyser with lots of functions. The base unit is competitive to low-end spectrum analysers, but when you need to buy extended coverage and more measurement features, the device is a bit pricy. With the supplied measurement features you can easily measure all kind of RF signal properties. Unfortunately a tracking generator is lacking, so you need a broadband noice source for measuring filters and antennas.

Anonymous
Parents Comment Children
No Data