View Single Post
Old 14th May 2022, 7:03 pm   #32
nemo_07
Triode
 
Join Date: May 2022
Location: Cologne, Germany.
Posts: 28
Default Re: 6-gang FM stereo tuner heads

Quote:
Originally Posted by RichardGM View Post
Thanks for the answers to my question, but I'm still a bit puzzled. Yes you need to filter out the image frequency and any nearby strong signals, but most domestic FM tuners do OK without multiple stages of filtering before the mixer. Excluding the BBC's very special case of a receiver "based at the site of a working VHF/FM transmitter", when/why are those extra stages of filtering needed? Is it to receive very weak distant broadcast transmissions?
Those are not only required to enable reception of weak stations but also to suppress spurious signals, created at any gain stage, and capable of blocking weak transmissions. Typical gain stage is told to be linear, but this holds only for really small input signals, say up to the limit of few mV at best (for a BJT; FETs are more immune in this respect). There doesn't have to be a BBC transmitter site to make a trouble. In any crowded urban area you will find signals from local FM stations and/or FM relay nodes, various RFI sources, etc., with levels of several mV, perhaps even 100mV RMS a simple dipole can deliver. Such strong interfering signals, with frequencies lying several channels away from desired will
a) mix themselves in the RF gain stage into a "ghost" RF or IF signal,
b) another ones will reach the mixer (amplified by the gain stage, not much weakened by poor selection),
c) 2-nd harmonic of unwanted one (Fi1) will mix with another one (Fi2, spaced by IF/2) into IF (IP3 issue),
d) yet another ones' second harmonics produced by the gain stage will mix with the second harmonic of LO.

Whenever any one of such mix product fits into the IF bandwidth, you will possibly hear a station, where there is none (a "ghost"). Such a ghost, when stronger than desired weak transmission will quiet the last one in the process of IF limiting.

To any desired frequency Fd you can find multiple of "unwanted" frequencies that fit into one of the mentioned schemes.
In case b) for example, we have at least twenty candidate pairs of following unwanted frequencies: Fu1 = Fd + IF/2 +/- n*(channel spacing) and Fu2 = Fd - IF/2 +/- n*(channel spacing), n = 1 to 10. Easy to see that, in a typical "economy" class front end they will encounter regular amplification (~26dB), and some 12...26dB attenuation before reaching the mixer, and will all mix into IF bandwidth.
A rough calculation can show, that immunity limits to the listed phenomena are for typical 3-gang front end ... let's say, "not outstanding at all".
nemo_07 is offline