Selectivity vs. sensitivity

You want these stations in your car ? Well, a buddy has a Ford Taurus with a radio with extreme sensitivity and doesn't exhibit that noise like older tuner due to the digital signals etc. on the carrier.

Some tuners vary the IF bandwidth with signal strength or even multipath re ception. (Shotz ?) this gives you the low distortion of the wide bandwidth on better signals and limits distortion on weaker or much reflected signals . A quite good design I say.

When you are dealing with a weak signal, narrow bandwidth is almost always the choice as noise goes up with bandwidth, and it of course affects select ivity. It affects THD especially in stereo because the sidebands are blocke d. Sure, you can get by with a +/- 75 KHz bandwidth but it sounds like shit . How picky are you ?

So you have questions that sales is not going to be able to answer here, I don't know what to tell you. Look at audiophile specs on these tuners if yo u can get them. the quieting curve, THD vs. signal level. No way can they q uantify how it handles multipath, but that is usually less of an issue with a signal originating farther away. (unless you are in NYC itself)

So in the end, both parameters you ask are important. I tighter bandwidth w ill shield out local stations that are stronger. If you have them 200 KHz a way you definitely need the utmost in selectivity. Sensitivity is related b ut shorter bandwidth will result in better quieting. Thus a tuner with shor ter bandwidth will measure better in selectivity with all other things equa l. I used to custom align IF strips on old tuners for that.

Another factor is that the frequencies you mentioned are at the low end of the band. that means the varactors are operating at higher capacitance and inaccuracy and drift are worse. this is almost impossible to tweak, I would not attempt it and lack the equipment to do it anyway.

As such, an in person audition ight be necessary. Go to the store whether y ou buy there or not and search the lower end of the FM band on various unit s. if you find one with really superior performance buy it, and insist on g etting THAT ONE, not another one in the box. that may be the only way to do it.

Anyway, that is my take on it, take it or leave it.

Both, and both have complications. For selectivity, if your FM radio can hear HD Radio (IBOC), it needs to have an IF bandwidth of at least 400 KHz. Here's what it looks like on a spectrum analyzer: It won't work if the IF bandwidth (-3dB down) were exactly 400 KHz as it would be clipping the corners of the IF bandpass and probably have horrible group delay. It has to be wider, typically about 500 to 600 KHz IF bandwidth. However, even if it was 400 KHz wide, you would still have an adjacent channel problem. In the US, FM channels are on

200 KHz intervals. That means the digital part of the spectra overlaps the conventional analog FM part of the spectrum in the adjacent channel. If you were listening to a conventional FM station, and there were an HD Radio digital station on the adjacent 200 KHz slot, its digital signal would slop into your IF bandpass and all you would hear is digital garbage.

The closest approximation to a solution are the digital FM receiver and demodulator chips by SiLabs. The IF bandwidth is programmable and very much a brick wall. The designers can narrow up the IF bandwidth to exactly 200 KHz which will remove most, but not all, of the HD Radio garbage in the adjacent channel. You'll find these chips in radios by Tecsun and Meloson. Not sure about car radios.

So the importance of selectivity depends on what you're listening to (conventional FM, or HD Radio) and what's on the adjacent channel.

It's almost midnight and I'm beat. I'll dive into the sensitivity part later, when I'm more awake. Incidentally, FM receiver sensitivity is rather oddly measured in dBf (dB above 1 femtowatt).

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

selectivity is the opposite to sensivity.

h a radio with extreme sensitivity and doesn't exhibit that noise like olde r tuner due to the digital signals etc. on the carrier.

I had more than a couple of people complain that their factory Ford radio h ad better AM DXing range than their aftermarket radio.

I don't know about the new Fords, but Fords through the early 2000s had ver y sensitive and reliable factory radios.

The last Ford I had was a 2000 Explorer, and that radio could pick up WFAN and WCBS (New York AM stations) almost like they were locals. I could list en to ballgame from New York when others couldn't. Too bad Ford didn't kno w as much about automatic transmissions as they do about building radios.

Where were you when you listened to WFAN and WCBS?

I had a 1946 Studebaker which could pick up Los Angeles radio stations when I was in Twin Falls, Idaho.

If you are really IN Baltimore, there's no way you can get DC stations on a car radio. I lived on the Eastern Shore of VA for a while, and I did listen to my favorite DC stations, but I built a Yagi antenna and pointed it up into the sky. That was a fixed location and home stereo equipment, not a car radio. That worked pretty well, but you got a lot of flutter any time an airplane was crossing in the middle.

If you are expecting to get the info from a few specs listed on the outside of the box, I wouldn't bother. Those specs are not likely to be very accurate or useful by comparison.

Jon

By no means. They're almost orthogonal in principle, although in practice they may tend to go in opposite directions.

Sensitivity is a measure of how well a tuner picks up the desired (on-frequency) signal... that is, how weak a "wanted" signal it can detect.

Selectivity is a measure of how well a tuner _rejects_ non-desired signals (on adjacent channels, alternate channels, and further away)... that is, how strong an "unwanted" signal it can filter out.

Cheap tuners may be both insensitive, and not very selective.

Really good FM tuners can be highly selective, _and_ highly sensitive. This generally requires having at least one stage of selectivity (i.e. one or more tuned stages) before the first RF amplifier - these stages keep strong, off-frequency signals from overloading the amp.

Tuners that are beloved of the "FM DX" crowd will tend to have:

- Quite a few stages of selectivity in the front-end. You'll see tuners described as having 3, 4, 5, or more "sections" in the front end - each section is a tuned circuit. Commonly, one section tunes the local oscillator, and the rest provide RF selectivity before the mixer.

- Several different bandwidths available in the IF section. A narrow IF (more stages of IF filtering, and/or more-narrowly-tuned filters in each stage) provides better selectivity. This can be at the cost of sensitivity (each IF filter stage adds some amount of loss, even to the desired signal) so a good narrow-IF design will include additional stages of IF amplification to make up the losses and restore the sensitivity.

- Post-detection filters, after the FM detector and before the stereo-multiplex decoder. These help filter out "birdies" and IBOC digital sub-carrier noise, from FM stations on adjacent or alternate channels.

No, it is not.

I want one with continuously variable IF bandwidth on the front.

With a DSP filter and demodulator, you can have exactly that. Spin the shaft encoder knob, and the filter coeficients will change to whatever you find useful. Not exactly what you want, but close:

ATR4255P For improved FM performance, an integrated IF filter with adjustable bandwidth, a soft mute feature, and an automatic multipath noise cancellation (MNC) circuit are fully integrated. A powerful set of sensors is provided for receiving condition analysis and stop signal generation.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

** Such a feature is useful with AM reception but not with broadcast FM.

The FM signal is inherently wide band, with +/-75 kHz deviation at peak audio level - if the IF bandwidth is less than 150kHz, distorted sound is the result.

I have a radio scanner ( AR 1000xlt ) with wide and narrow FM modes, 30kHz and 200kHz respectively. Listening to broadcast FM while in narrow mode is *intolerable*, in wide mode it sounds just fine.

.... Phil

I've had it in a couple of tuners and found it useful when DXing, and once in a great while other times but still DXing somewhat. Listening to strong local stations just leave it in wide though, then it practically worthless.

" I want one with continuously variable IF bandwidth on the front. "

The FM signal is inherently wide band, with +/-75 kHz deviation at peak audio level - if the IF bandwidth is less than 150kHz, distorted sound is the result.

I have a radio scanner ( AR 1000xlt ) with wide and narrow FM modes, 30kHz and 200kHz respectively. Listening to broadcast FM while in narrow mode is *intolerable*, in wide mode it sounds just fine.

** Continuously variable ????

Or just a couple of settings.

FFS stop oversnipping !!!!

.... Phil

Sorry, switchable, narrow and wide. I heard of a tuner with three positions but never had one and don't remember the name or model.

Doing variable control in the analog domain would be an incredible bitch bu t that's how I want it. Actually it only has to be a couple of stages towar ds the front, maybe even just one, but still. Maybe Studer Revox would do s omething like this (maybe they have) but I doubt anyone else. Them engineer s there I bet are showoffs. I'm not bitching, just saying. I saw the print of one of their tuners and said "WTF, are you sending the first mission to Mars or what here ?". I'm sure I still have it lurking around a drive here but it would probably be easier to find it online again than to look throug h all my backup.

Yes I am a nasty SOB who wants someone else to do what I can't. But they ge t paid for that.

I had an 82 Toyota Celica. Radio was horrible. It looked like it was designed for an external antenna amplifier and when one was added, the number of stations received increased.

It was very "hissy" when the speakers were upgraded,

I have a 2000 Solara and the Radio is pretty nice. 5 CD changer, cassette and AM/FM.

==

House wise, I have a Technics Professional ST-9030. See

I have been able to get 3 stations on the same frequency just by rotating the antenna.

The FM bandwidth is selectable from auto and wide. They probably should have made it auto and narrow. It may continuously change bandwidths in auto mode.

To make things more interesting, I added a Carver TX1-11 signal process which makes multipath noise "go away".

More interesting yet, I have a 4bx dynamic range expander with impact restoration.

The tuner had two outputs. One filtered at 15 kHz so it would not interfere with tape decks and one with a higher response.

I also owned a automotive Blaupunkt Tucson which was very good coupled with an active AM antenna.

igned for an external antenna amplifier and when one was added, the number of stations received increased.

Ford and GM seemed to apply importance to their car stereos in the early 19

80s. Amps with DPL (anti-clippping gain control), speakers from Bose, tuner s from Hughes Aircraft which were also used in David Hafler tuners, and Bla upunkt for the cassette decks. Bose also made some amps for mainly Caddilac s, they were class D and ran into very low impedance. But that was no stran ger to Bose, the speakers in the 901s are 0.9 ohms each. They get near 8 oh ms being all in series.

And a buddy of mine has a 2005 Ford Taurus and that thing gets stations I c an't even dream of getting. Two of them I really want, 94.9 and 97.5 out of Akron, which is not all that close. And he gets them clean. I might have t o fell him over this. (jk)

able to get 3 stations on the same frequency just by rotating the antenna.

have made it auto and narrow. It may continuously change bandwidths in aut o mode.

I ould like auto, narrow and wide if I can't have a variable control. In fa ct I wouldn't mind auto because signal conditions change.

hich makes multipath noise "go away".

I bet all that is is a stereo blend with logic. Bob did alot of "magic" tha t way. Those asymmetrical charged coupled device tuners of his were not the cat's ass. All it did was to detect what it thought was distortion along m aybe with multipath which is easy to detect and start blending left and rig ht and adding a digitally delayed signal to the L-R.

toration.

I am going to build something like that one day if I live long enough, and all in discrete components.

ere with tape decks and one with a higher response.

Regular FM stereo is capable of 19 KHz but it takes some doing. It takes a very accurate pilot cancel circuit with a PLL generated null signal to assu re a perfect sine wave. Interference to the wave would be reproduced, but a lso prevented from interfering with the phase lock of the MPX decoder oscil lator. I don't want the job, maybe Revox.

th an active AM antenna.

Seems like you're into AM. Well, it is more people's media because it is mu ch cheaper to have an AM station than FM, and forget TV. Maybe, for the com mon good we should support AM radio.

I'll have to give that some thought.

A bit of hair-splitting here. The FM channel allocation in the USA is at 200 KHz intervals. However, if one adds the digital (HD Radio, IBOC, iBiquity, etc) modulation, the bandwidth is now 400 KHz wide: This is what it looks like on a spectrum analyzer:

In most HD Radio receivers, the IF bandwidth is set by a digital filter. For conventional FM, it's 200 KHz wide. For digital FM, it's

400 KHz wide. I suppose it could be front panel set by the user, but methinks it makes more sense to have the IF bandwidth automagically set by the mode and sub-channel.

Measuring Your IBOC Spectrum

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
Santa Cruz CA 95060 http://802.11junk.com 
Skype: JeffLiebermann     AE6KS    831-336-2558

Continuously variable for FM doesn't make sense. But there have been some FM tuners that could be switched between "wide" and "narrow", in relative terms. So for strong signals, wider bandwidth is fine. But for weaker signals, narrower bandwidth avoids interference from adjacent signals that are stronger. It wasn't uncommon for FM DXers to swap the ceramic filters in their FM receivers from the often 280KHz bandwidth to down about

180KHz, at one time one could go to a catalog and order Murata ceramic filters in a range of bandwidths. If you don't need FM, you can get by with narrower, though of course nt in the tens of KHz wide.

The scanner wants "narrow" for two way communication which is narrow deviation, 10KHz or smaller in recent years. The wide is for broadcast FM and maybe some other things, since yes, the "narrow" in this case is way too narrow for FM broadcast. Of course, the wider bandwidth can be useful for things like receiving weather satellites, which may have a wider deviation of something like 40KHz, but also because of doppler shift, an even wider bandwidth makes things easier. I know I've seen modifications for scanners to use with weather satellites, and they bypass the narrow filter at 455KHz, which leaves an FM broadcast band type ceramic filter at the first IF of 10.7MHz.

There was a time twenty years ago when I was bringing home lots of Delco car radios from garage sales. I'm not sure what the FM filter is in there, but they certainly seemed to have better skirt selectivity than other FM radios I'm familiar with. The AM filters seeemd sharper too.

Michael

for FM DX, you want a radio that you can switch to MONO mode.

mark