With the EZRadioPro series, Silicon Labs offers a range of attractive receiver and transceiver chips in a QFN20 package for telemetry applications in the sub-GHz range. There are also complete modules equipped with this IC for little money (including on AliExpress). For home automation applications on 433 or 866 MHz I usually use the comparable RFM69 from HopeRF, but what makes the Si4463 special is the enormous continuous frequency range (142 - 1050 MHz), the fact that the VHF band is part of this and especially the maritime AIS-channels (161.975 and 162.025 MHz) come into view.
The examples in the application note (AN643) show how with 3 or 4 components an LC balun can be assembled to adjust the differential input of the receiver to 50 Ω. Because I found the receiver sensitivity disappointing in practice, I became increasingly suspicious of this network. It is difficult to experiment with fixed capacitors and inductances in a 603 footprint. Instead, I copied a 169 MHz example from the application note in LTspice to see how critical everything was. And then things became a little more clear to me. This article discusses the possible cause of the poor reception sensitivity and tries to provide some direction towards a possible solution.
Above the recommended matching network from the application note for 169 MHz. The differential input of the receiver at that frequency has a nominal input impedance of 536 Ω with a parallel capacitance of 0.98 pF. The network is an LC balun and has two tasks. The first is the impedance transformation from 50 Ω to 536 Ω. The second task is to perform a 'single-ended' to 'differential' conversion. Next to the schematic is a Bode plot of sum and difference of both input signals. It is immediately clear that the plus and minus 90 degrees phase shift and the impedance transformation are particularly critical. The example is indeed neat at 169 MHz but with the closest E12 v alues (5.6 pF, 12 pF, 150 nH and 220 nH) it is already completely lost. But that won't be the only problem!
The Bode plot above shows that the two output signals at 169 MHz are neatly in phase opposition and the voltage gain is approximately 2 x 4.3 = 8.6 db. This plot runs from 160 to 180 MHz. However, the situation looks quite different if you look a bit more broadly.
Pictured still the same circuit but now from 50 to 200 MHz. What becomes visible is a huge peak at about 90 MHz where the two output signals are signals after transformation resp. 17.5 and 21 db above the input signal. That's right in the FM broadcast band. It is true that the two signals are in phase with each other in this area and should cancel each other out. The common mode rejection of the differential amplifier in the SI4463 is not specified, but there will be little left of it anyway once you get into the regions where blocking and intermodulation occurs.
To see what I find in the radio spectrum at my location I connected the AIS antenna, a simple vertical dipole, directly to the spectrum analyzer.
Only the strongest AIS signals come out just above the noise on the set bandwidth, but a local radio station at 90Mhz is about 50 dB above that and a DAB+ signal at 182 MHz about 35 dB. The unfortunate behavior of the LC-Balun makes the receiver particularly vulnerable to strong signals precisely in an FM band where the most danger is to be feared. A preamplifier will obviously only make the situation worse. It appears that the receiver is completely overloaded while the vast majority of AIS stations are significantly weaker than the few visible on the spectrum analyzer. Two possible solutions are : 1. Additional preselection and 2. A better alternative for the LC balun. I assume both are necessary.Silab's recommendation for an LC balun will mainly be motivated by offering a simple / uniform / reproducible solution over the entire frequency range and in particular for the countless applications in the 433 and 868 MHz band. The VHF range is a bit of an outsider in that light. At that frequency it is much more obvious to choose a transformer on a ferrite core. The impedance transformation is roughly 1 : 10. With a winding ratio of 1 : 3.3 (or 1 : 1.66 + 1.66) we come a long way. The 1 pF input capacitance is not dramatic and can be compensated if necessary.
To investigate further, I performed a simulation with a 1 : 9 impedance transformer, the Ciolcraft, the WBC9-L_. I chose this one because the inductances could be found in a datasheet and I couldn't find any Spice models of similar transformers from Murata, Amidon etc
This all looks a lot more soothing. 180 degrees phase shift and approximately 2 x 6.5 dB voltage gain over a large range.
Perhaps this produces a somewhat too ideal picture, but we have in any case dealt with the resonance phenomena in the FM-band and the critical component choice. The next step is to put it into practice.
Since I wish to receive the two AIS channels simultaneously, I have also chosen to have a Splitter/combiner to be included in the design. It usually consists of two transformers. The first is an auto transformer that halves the input impedance. The second is bifilar wound with the impedance for each port doubling again. Because we want to go in the direction of 500 Ω we can drop the first transformer. The 50 Ω at the input will end up at two ports of 100 Ω each after the splitter. That is already a step in the right direction. A simple trifilar transformer will then bring it to 2 x 200 = 400 Ω for each AIS receiver. The idea is to pass both secondary windings one more time through one of the holes of a double aperture core (pig's nose). With this extra half wrap we will get pretty close to the final target of 532 Ω. The test setup looks like this :
To measure the most critical part I placed a ferrite core between two "matching pi attenuators". At the input a 20 db attenuator with an input impedance of 50 Ω and an output impedance of 100 Ω. Behind the transformer a second matching pi network with input impedance of 266 Ω and output impedance of 50 Ω, now at the price of another 10 dB attenuation. The unused secondary winding of the transformer is simply terminated with a resistance of 266 Ω.
The used calculator can be found at href="/https : //chemandy.com/calculators/matching-pi-attenuator-calculator.htm"
A first measurement was performed with a trifilar winding of 2 + 2.5 + 2.5 turns (d = 0.2 mm) on a FairRite 2643000301 ferrite core 3.5 x 1.6 x 6 mm #43 material on the PCB below. Impedance ratio should then be about 1 : 5, so that we arrive at 500 Ω.
The outcome of the measurement
The measured attenuation at 162 MHz is 34.25 – 30 = 4.25db. It has to be taken into account that half of the power is now dissipated in the dummy resistance of 266 Ω but ultimately gives its share to the receiver. The actual loss is 4.25 - 3 = 1.25 db, which seems very reasonable to me.
The proposed complete schematic for a dual channel AIS receiver then becomes as below:
And below the provisional mounting of the splitter and input transformers on an AIS receiver. They replace the LC baluns the board was designed for. This time performed with miniature double aperture cores, in #61 material from FairRite part no. 2861002302. The findings were identical.
An experiment was also performed with an Aliexpress Si4463 module with a transformer balun.
Preliminary Findings
Extra pre-selection with an LC or SAW bandpass filter is certainly necessary and I may come back to it later, but the balun with transformer already provides a considerable advantage when it comes to preventing blocking from the FM band. Moreover, it is easier to realize than the LC variant with capacitors and coils with exotic values. These are not only difficult to obtain in a reasonable quality with a 603 footprint, but there is also nothing to verify or adjust once they are on the PCB.
It also appears that careful layout and complete shielding and decoupling is of paramount importance. Due to its high input impedance, the Si4463 is extremely sensitive to any kind of noise and interference signals on the input circuit. See Silabs' recommendations in this regard.
Additional aplification must be handled with care. On top of the 3 db loss due to the splitter, maybe 2 dB occurs through transformation. Anything more to compensate for that loss only compromises the (moderate) strong signal properties of the receiver chip.