Philmore FM Separator

Last night I found an old FM separator, made by Viewsonics, for Cox Cable.

I immediately set up the whole thing (which is a bit complicated) and it worked.

FM was about midway between as good as with the Yamaha direct on the higher antenna, and what it was before.

But things sounded wrong, grainy.  When I off tuned it on the downside it quickly became very distorted.

Today the new Philmore FM Separator arrived.  It is not like the one in the catalog photo (which, IIRC, looks like my Viewsonics).  It is much larger, made in China in a very solid box, and has specs (<= 2.5dB within specified ranges).

[Updated at bottom.]

It works far better.  There is no "off" sound and the noise is about as good as the Yamaha on the other antenna.

But I also wanted to start observing these things, so for the first time ever, I hooked up a scope to the F-26.

The Kenwood vector scope I bought a few years ago and had resting atop the F-26 for years had insufficient gain and was worthless.  I had to use my Tektronics.

It's clear the alignment is off, and the auto-lock pulls the station off tune, which is aggravated  by the, yes, narrow band that is always selected.  I tried some local rock stations, and they too seemed to be pulled in only in Narrow, even on the outdoor antenna.

I have a mind to surgically remove the auto-lock and auto-narrow, leaving wide the only choice.

And perhaps improve the alignment.

All looks pretty good, though, if I just tune right above where the tune indicator says.  Then the auto lock doesn't engage, and that turns out to be the best spot too.  And the IF modulation seems very flat, until it hits the stops on really wide modulations, because it's stuck at narrow.

No obvious issue with the new antenna system though.  I remember that Cox had their own FM channels and didn't care if you got KPBS-FM.

The optimal separator for a scanner would not be TV_centric, and just divide the spectrum between 108Mhz and 120 Mhz (so no division below 88 Mhz).  I'll keep looking for that.  But this one seems OK.

The new arrangement which made all this possible now has a shelf reserved for the living room tuner, so I could put another tuner there while I'm working on the F-26 more easily.

Something cool running with no drift.  The 9500 might be a good choice.  And now the stronger antenna signal means there won't be so much of a noise issue with tuners having more primitive MPX.

[Later tests demonstrated the Philmore 9500 is adding a kind of hashy distortion to 88.3 Mhz.  So compared to indoor antenna, the noise is reduced, but an irritating edgy distortion takes it's place.  I first noticed this on the already edgy sounding Bedroom System I have been trying to make sound better.  But then I found, indeed, I can hear the same artifacts on my other systems, they are just far less noticeable.  So then I tried a UHF/VHF seperator, which simply divides the spectrum at 300 Mhz.  Well it seemed to do the same thing!!!  (Compared running the antenna feed straight through, using a double female F adapter.)  Finally, I tried a mere splitter, which causes 3dB reduction in signal to both outputs at all frequencies.  This one claims to be good to 1500 Hz and is fairly chunky.  That does not add any hashy distortion to the signal.  I can barely tell the difference between it and the mere adapter, and to the degree I do, I might prefer the splitter.  It would seem the FM separator could have a problem at 88.3 Mhz, because TV channel 6 reaches 88 Mhz...any reasonable filter has to be in steep decline by 88.3 if not 88.2.  But why the issue with the mere VHF/UHF separator then?  Possibly there are ripples created by these things way beyond the separation point.  For low fidelity communications, these are adequate, but for hifi FM stereo, phase ripples cause distortions and delays in the demodulated signal.  And/or possibly also there could be nonlinear effects.  I noticed in one separator I opened, it had ferrite filters in it.  So I don't know whether the problems are caused by linear or non-linear effects.  Either way, a better design, using better parts, might be able to do better.  But after many hours of scanning, I've found nothing better looking than the Philmore.  Designing something like this is beyond my demonstrated capabilities, though I've seen discussions of making similar DIY devices, so I think I could figure it out after some time.]