Building Blocks on an FM Band Receiver

As a way to learn more about RF design, I am considering to build an FM Band receiver. I am not looking for a professional level outcome but rather to learn. I have an oscilloscope with 200MHz bandwidth, as well as a low cost VNA. I built the attached block diagram for how the receiver would look like. Cost is not a concern as I am not looking for competing with commercial existing FM receiver but rather to learn.

Any feedback on this block diagram?

I started with the first stage which is the antenna and RF amplifier. I used the VNA to measure the impedance of a short wire antenna (20 cm). I understand that this is far from ideal, but this is what I can practically afford from a physical constraint point of view.

I built as well a common emitter amplifier with a decent gain of 25db from 45-200MHz (measured with VNA) based on the BFP420 transistor.

My first challenge is how to match these 2 together. I can do that for a fixed frequency, but what I found out that both antenna impedance and S11 (input impedance) of the RF amplifier change across the 20MHz FM Band.

Does such matching network exist between a load and generator for which their own impedance also change over frequency. How does commercial FM receivers address this? Do they just design for one frequency in the middle of the band.

Thank you.

Before going further - when you say "FM Band" you seem to be referring to the domestic VHF broadcast band (~87MHz to 108MHz) where relatively wide band transmissions are used but the block diagram is for a narrow band FM receiver. A domestic receiver would not use a second mixer stage and 455KHz IF, it would be too low to cater for the ~180KHz domestic broadcast requirements. Instead more gain at 10.7MHz is used and a demodulator at that frequency.

Brian.

Thank you Brian. Yes, I was referring to the domestic FM broadcast, but as I am planning to only demodulate the mono channel, I thought that I would only care about 15KHz bandwidth. I was considering to use the CD4046 or LM567 which both cannot work at 10.7MHz, hence was considering a second mixer to take the frequency down to this level where these two PLL chips can handle the FM demodulation.

IS there any other PLL chip that would decode at 10.7 MHz?

FM radio receiver bandwidth must be larger than 15 kHz for undistorted reception of the audio baseband. 455 kHz IF doesn't make much sense for it, I doubt you find a suitable ceramic filter.

A possible weak point is the VNA without preselection, can be easily blocked by radio services, e.g GSM.

As per FvM, don't confuse FM and AM bandwidth requirements. The deviation of standard FM broadcasts is far wider than the bandwidth of normal 455KHz filters so you would 'chop off' most of the band edges and at best get distorted audio from it. That has little to do with it being stereo or mono although you obviously won't squeeze the pilot tone through a narrow filter anyway.

You can get PLL demodulators for 10.7MHz but they may not offer any advantage over a quadrature based demodulator. The primary reason is the need for a strong enough signal to enable locking and below that threshold you get nothing at all. There are also some smaller issues such as poor LF response as the loop 'fights back' against slow deviations in frequency and the additional complexity and screening needed.

Consider moving the input BPF to before the RF amp or better still using a tunable filter at the input to improve selectivity.

Brian.

Before amplifier, you should use a filter to clean out the unwanted signals such as VHF , Air Traffic, Amateur Band signals.
Otherwise the Amplifier will try to amplify everything and it will of course mix them up

Thank you for your valuable input. I updated the block diagram accordingly. Regarding the BPF before the RF amplifier, I implemented actually this BPF based on a design I created using Elsie application. I checked its insertion loss using the VNA, and it is around 4-8 db (not flat across the FM band).

Is it OK to get this attenuation and then amplify the received signal? Won't this reduce the sensitivity of the receiver?

For the mixer, I am planning to use an SA612AD that I purchased. Is this a suitable choice?

Regarding the quadrature demodulator, reading about it, seems that it can as well be implemented using a mixer feeding it with the signal, and its 90deg shifted and take the output of the mixer through a LPF. Would it be ok to use the same type of mixer for this, or should I look for a ready made chip to perform the demodulation?

A question in my initial post was regarding impedance matching. Not sure that it was addressed yet. My understanding is that impedance matching is very important in the front end to improve the received weak signal power transfer.

The BPF is design for 50 ohm input impedance. Measuring the antenna impedance in the FM band is showing a Z ranging from (20-j1000) @ 87MHz to (32-j767) @ 108MHz.

How can I match this range to the fixed 50 ohm input impedance of the BPF?

A typical BPF should has 1dB ( more or less ) Insertion Loss at those frequencies, not 10dB.
This Insertion Loss will surely reduce the Sensitivity of the Receiver but it's obligatory..
If the Antenna Impedance varies so, I believe this antenna is not very well designed.Nevertheless, Signal Levels are pretty high at FM Band.So, I don't think that the Impedance won't create any trouble.
Matching is always possible but this variation is quite high..