It seems like it was ages ago (September 2021) that Randall had posted about a Special Project to demonstrate using a Rohde & Schwartz FPC1500 Spectrum Analyzer to accomplish specific tasks such as EMI debugging, low power design, or IoT Application.
Example tasks included:
- EMI debugging a switched-mode power supply
- EMC pre-compliance testing
- IOT: testing Bluetooth Low Energy (BLE) transmit signals
- IOT: Antenna matching in IoT and low power devices
I've wanted to get a spectrum analyzer for a while and this seemed like a good opportunity to get a high quality instrument and to use it to do some testing of the RF capabilities of my development boards.
I noticed in the FPC1500 documentation that "the BLE transmit signal can easily be evaluated using an R&S
FPC with the R&S FPC-K7 modulation analysis option". So, I proposed testing BLE transmit signals of the various BLE capable development boards that I have to see how BLE PHY performance correlates relative to power consumed and with the transmitter and antenna designs.
I was selected to participate in the Special Project at the end of October 2021, but things were delayed due to the lack of an available instrument. I finally received an instrument on May 4. It is a used demo instrument, but it is in good condition and has a full set instrument options enabled. Included with the instrument were a power cable, USB cable and a soft-sided case.
Here is my initial setup to start evaluating the instrument capabilities before finalizing my project plan. I am going to try using a 2.4GHz 9dBi omnidirectional antenna as my reference input to the FPC1500.
Here is the configuration of the instrument as I received it. There is a newer version of software (firmware), V1.70 - so one of my first tasks will be to update it.
Here is the list of installed options.
Unfortunately the WiFi support option has expired, but I can use Ethernet for remote operation. Just can't use it outdoors where I don't have wired access.
For the purposes of this project, I need the B2/B3 Frequency Upgrades to cover the 2.4GHz ISM band used by BLE and the K7 Modulation Analysis Option for demodulation and analysis of the received signal.
The FPC1500 is basically the FPC1000 Spectrum Analyzer with the addition of a Tracking Generator and VNA capabilities. Both the FPC1000 and the FPC1500 have been roadtested and reviewed previously on Element14, so for this Special Project I will only be focused on demonstrating the capabilities required for BLE transmitter PHY testing.
Here are some of the key specifications of the FPC1500:
|
Key specifications |
|
|
Frequency range |
5 kHz to 1 GHz, with upgrades up to 3 GHz |
|
Max. input power |
up to +30 dBm |
|
Low noise floor |
down to typ. –165 dBm (with preamplifier) |
|
Connectivity |
LAN, USB, Wi-Fi (optional) |
|
Measurement options |
modulation analysis, receiver mode, advanced measurements |
|
Tracking generator |
5 kHz to 1/2/3 GHz |
|
Network analyzer |
Internal VSWR bridge |
BLE RF Signaling overview
The RFPHY characteristics are shown in the table below.
(values are relative to LE introduction in Bluetooth 4.2 - I have devices that span BLE 4.2 to BLE 5.2 and some of the libraries that I use don't allow using enhanced performance)
The specific devices that I'm using also have limits in terms of the Max Tx power. BLE was intended for lower power so typical operating power might be 0 dBm or less with the device max at 8 or 10 dBm. Recent spec enhancements allow up to 20 dBm for longer range (like standard Bluetooth), but I don't think any of my boards have that capability - I'll have to check some of the newer ones.
|
Characteristic |
|
|
Radio Frequency |
2.4GHz ISM (2400MHz ~ 2483.5MHz) |
|
RF Channels |
40 (f=2402+k*2 MHz, k=0, … ,39) 2 MHz separation |
|
Range |
10m~50m |
|
Modulation |
GFSK |
|
Modulation Index |
0.5 |
|
Bandwidth Period |
0.5 |
|
Max Tx Pavg |
+10 dBm |
|
Symbol Rate |
1 MS/s |
|
Application Throughput |
0.27 Mbit/s |
|
Access Scheme |
FDMA, TDMA |
|
Packets |
2 types |
|
Data Rate |
1 Mbps |
|
Security |
128-bit AES and applications layer user defined |
The frequency spectrum occupied by Bluetooth BR/EDR with 79 channels @ 1MHz spacing and BLE with 40 channels @ 2MHz spacing.
Spectrum analyzer test setup
I am going to start with the instrument configuration described in The Smarter Way to Test BLE Transmit Signals Application Card.
|
R&S®FPC setup configuration |
|
|
PRESET |
|
|
Mode |
Digital Demodulation |
|
Frequency |
2402 MHz |
|
Ampt |
Reference Level ▷ –20 dBm |
|
Sweep |
Trigger ▷ I/Q Power ▷ –30 dBm |
|
Meas |
FSK ▷ Standard ▷ Bluetooth® LE |
|
Ampt |
Deviation per ▷ 100 kHz Division |
|
Meas |
Demod Parameters ▷ Burst Processing |
|
|
Number of Symbols ▷ 400 |
Ambient RF Environment
I was happy to see a nice eye diagram displayed when I did a quick sanity check that I showed in the previous blog. It did bother me a little that Carrier Power was shown as -48.0 dBm since this was a close proximity test I had expected a higher value, but I'm still trying to get a feel for what values are reasonable.
It was later when I powered down the dev board that I realized that I had gotten a spurious capture from some other BLE device in my workroom environment -> and I was still getting captures 
. I had initially not gotten any triggers, so I had set the trigger level to a low value, -50 dBm. I have a lot of BLE devices in my house and dozens active in my workroom and the adjacent computer room (desktop computers, laptops, tablets, cellphones, headphones, earbuds, and a bunch of peripheral devices - mice, keyboards, remotes, and an Apple pencil, and a bunch of dev boards). I thought that I had been careful to shut down all of the devices in the two rooms, but apparently I had missed at least one. For doing some limited range testing I was hoping to get background "noise" to under -70 dBm.
After some head scratching, I decided it was best to see if I could try to identify it using the Nordic nRF BLE tool that I have installed as part of nRF Connect for Desktop on one of my computers.
I attached an nRF52840 USB dongle and ran a scan - and there it was, plus a couple others that are probably in another room.
Looking at the Details provided a clue, but I could not identify the device immediately- I thought that I might try to look on my router to associate the MAC address.
Luckily, there is an index of Service UUIDs - 16-bit UUID Numbers
And FE78 is an HP device, so it's probably my multi-function printer/scanner.
It's funny, because I run that unit with Ethernet, and I've never used BLE with it - did not even know it had BLE... But, the MAC address correlated, so I found BLE in the settings and turned it off.
And the problem is gone (although the lower level ones are different - I've run longer duration scans to make sure that there aren't any larger amplitude ones showing up). Now to get on to trying some real testing.
