My two FM Antennas Compared

 

I have two outdoor FM antennas that were installed by electricians 8 years ago.*  One is normally connected to the Living Room FM tuner and the other is normally connected to my kitchen FM tuner.  For almost 20 years I've kept the Living Room tuner set to 88.3 KPAC for background music listening, and I use the kitchen tuner (conveniently next to my kitchen chair) for searching for and tuning in other stations, especially 90.1 KSYM and 91.7 KRTU, two college radio stations.

(*This was a very expensive project because of the grounding requirements.  Each antenna has its own grounding rod, and all grounding rods are connected to the main house grounding rod by a thick wire that runs 3 feet underground across the back yard.)

Five years ago I decided I wanted to measure and compare the two antennas on the same tuner.  So for this purpose I moved my classic Kenwood 600T FM tuner to the kitchen.  I was going to use the 600T's marvelous 10dB calibrated signal strength meters, multipath meter, and scope outputs with a scope to compare my two FM antennas.  Otherwise, the 600T is not my favorite sounding tuner.  I planned to do the test in a few days and then replace the 600T with one of my other tuners that sounds better.

But after setting it up in the kitchen, I discovered the 600T sounds fairly good if the MPX Filter position on the Mode control is selected.  (Otherwise it sounds bright and tinny.)   And because of that and other things, you know how it goes, I never got around to doing the tests until now (last week).

One antenna is the famous Magnum Dynalab ST-2.  Another is the less well known Godar FM DXR 1000.  They are both whip antennas not because that is the best but because that is the easiest to install, since they can simply be screwed to the side of the house and not atop a mast.  I've long dreamed of having an antenna mast but it would be a very complicated outdoor project, too complicated for me so far.  Whip antennas also have an advantage in that they are omnidirectional and never need to be rotated.  This can also be a disadvantage if you want to block some strong unwanted signal.

(Most of the differences here probably have little to do with the whip antennas themselves, which should have fairly comparable performance in the FM band, but in the differing heights of the antennas as installed, and the fact that one of them--the Godar--is now bent.  Strangely, however, the bent antenna actually worked better on some stations in this test.)

I have the ST-2 is mounted as near as it could get to the peak of the roof, as high as I can go just mounting the antenna to the side of the house.  I couldn't mount it right at the peak because it would interfere with my locked attic door.

The Godar is mounted about 4 feet away, and situated about two feet lower because of the roof slope.  The Godar is tuned to what I determined to be best by SWR in the low FM band, probably 88.3.  The ST-2 has similar fixed tuning, best at the low FM band (which is where all my stations of interest are anyway).

Sadly tree branches messed with the Godar last year, and is bent to an angle less than half way up.  So how well is the Godar still working anyway, I wanted to know.  Do I need to buy a new antenna?

The Godar antenna, because it has greater RF bandwidth* than the ST-2 is also used for my Uniden SDS200 scanner--whose importance I'm beginning to question now that police bands here are 100% encrypted, which Scanners can't legally break even if they could do it technically.  Supposedly I can still get fire and EMS communications though I need to update the programming in my scanner for the newest systems, which changed about 2 years after I bought my scanner.

*I long figured the Godar has greater specified RF bandwidth, up to 800 Mhz because of having a lower inductance choke.  However it might also be because of the thicker whip made out of aluminum rather than steel, and because you can make the antenna much shorter.  I have it extended to nearly the maximum length so maybe I am not getting the top RF bandwidth after all, but it did work OK with my scanner at frequencies around 850 Mhz when the scanner was still programmed correctly for local services.

So in between the Godar and the living room FM tuner there is a splitter, which reduces the signal sent to the FM tuner by 3dB.  I tried an FM/TV splitter, which could theoretically have zero reduction in the FM band, but decided that it curiously caused non-linearities in the FM band which were even worse than a splitter (or so it seemed).  So I went back to a plain vanilla nice quality 1000 Mhz bandwidth splitter.  The signal then passes through permanently wired RG-6 panels in kitchen and living room to reach my living room tuner, a Pioneer F-26, one of the best FM tuners ever (like many of the other FM tuners I have collected) and from the late 1970's.

I was motivated to do the test now both because of the Godar antenna being bent and because I really really want to start playing my other tuners, including the Pioneer ST-9500 MkII that I finally got down from the top shelf in my climate controlled storage building a few months ago.  A friend believes that to be one of the best tuners ever, possibly even better than the F-26.

It just happens that I haven't been listening to KRTU much on the kitchen tuner recently, but I continue listening to 2 hours of weekly space music on KSYM.

Now finally testing the Godar on the 600T playing KSYM, I was shocked to find that the signal level was 13 dB lower, and not inconsequentially (26dB vs 39dB).  But over time I became aware that it was actually sounding better.  This became more and more clear as listening went on for a few hours.

At first I was thinking the noisier sound with the ST-2 was because of some kind of distortion, possibly caused by ground loop isolator.  Over the years, I'd used a number of ground loop isolators with the kitchen tuner because otherwise I'd get a serious hum on the kitchen audio system.  Some of those isolators seemed to cause notable distortion.  But it appears that I've removed all the isolators, and now I simply run the ST-2 signal through the RF surge suppressor built into a Monster brand AV power strip.  This also ensures it is firmly grounded to the kitchen audio system ground (or vice versa).  For a moment I was thinking of bypassing the Monster surge protection.  But it would be a big hassle because of where that power strip is located.

So what else could be causing this problem that makes the ST-2 sound worse on KSYM, I thought to myself, despite having much higher signal strength.  The Multipath meter hardly ever even moves, so it didn't look like it would be "multipath."

It was necessary to hook up the scope to find out.  And it turns out the scope is far more revealing of "multipath" distortion than the multipath meter.  In fact, the 600T multipath meter is basically useless.  You need a scope to see what is actually going on.  

Mitch Cotter, who had done a study on FM tuners for Consumers Union, convinced Saul Marantz of this, and the result was the legendary Marantz 10B tuner, inroduced in 1962 with a built-in scope, and many successors with built-in scopes made by Marantz and other companies.  After the Marantz 10B was introduced, most top tuners from other companies, even if they didn't include an actual scope, have generally included scope output jacks that could be connected to scopes.  McIntosh top tuners had the scope outputs and McIntosh themselves made a purpose built scope for consumers, whose last incarnation was the Maximum Performance Indicator MPI-4 (leave it to McIntosh to give it such a fancy name).  Kenwood made various scopes primarily for amateur radio (I have one, it's too big and clunky to mess with much, and doesn't show a good picture except for Kenwood FM tuners) but also a few for consumer audio purposes.

I've normally used an aging and increasingly dysfunctional Tektronix CRT scope for this purpose, but this time, for the first time ever for me, I used my relatively easily moved and situated Rigol DS1102E digital oscilloscope.  I had never figured out how to use this as an X-Y vectorscope which is what is needed for seeing FM multipath.  It turns out that this is an X-Y option in the Horizontal menu for this.  

KSYM using Godar DXR 1000

Hooked up to the Godar antenna, the scope shows a notable V pattern.  The thickness of the V is no more than 2 screen divisions.   That "thickness" is what indicates the degree of "multipath" distortion (which technically means unwanted in-band signals which can be caused by many things including multipath, adjacent stations, and other in-band stations).

KSYM using Magnum Dynalab ST-2 antenna

Hooked up to the ST-2 antenna, the pattern might best be described as "blob" but more importantly the middle doesn't condense down to just one screen division.  You see a smattering of dots going nearly up to the top of the screen.  It's not much of the total pattern, but it's enough to make the FM demodulation much noisier.  The multipath meter wasn't showing the total spread, just some kind of average which loses these details.

I believe what the V pattern is indicating is the presence of an on-band but weaker signal, which is being rejected by the FM tuners "capture" characteristic, which gives the upper hand to a slightly stronger signal at all modulation levels.  This is specified in an FM tuners Capture Ratio which is very good for the 600T tuner, 0.8dB in wide mode, which is almost a miracle except that a few tuners, like the legendary Sansui TU-X1 do even better.

(And tellingly, the Wide FM band selection always sounds better on KSYM, even though there are alternate strong channels on both sides.  This alone suggests the interference is in-band.  And there are several in-band candidate stations possibly causing this in-band interference, including one at 100 miles away at a higher elevation--KTXI)

What seems to be happening for the ST-2 is that since it rises higher, it's picking up more of the offending in-band interference signals, which are coming from farther away.  So it's not so much the channel I want that's getting stronger, as shown by the signal strength meter, it's the other channel I don't want that's getting stronger even faster, and just enough to start overwhelming FM capture at the fringes, enough to add some very annoying noise.

I can just as well listen to KSYM on the living room system with the F-26/Godar for space music as background music from now on, and take advantage of the superior sound of KSYM on the Godar.  In the kitchen, where I'm usually sitting in front of computer screens, the sound can be too in-your-face for background music purposes, especially during the announcements.

And it turns out that the Godar is also working better on KPAC as well, the station I usually use it for too.

On KPAC the Godar measures about 10dB weaker (only 3dB probably caused by the splitter, 7dB caused by the lower height antenna) but the multipath trace is about half the thickness of the ST-2.  Despite the lower signal strength, the sound is just a bit cleaner with the Godar antenna, as suggested by the mulitpath traces on scope.

KPAC using Godar DXR 1000

KPAC using Magnum Dynalab ST-2

Both antennas have the V outline similar to the Godar antenna on KSYM.  I think this is another capture situation with the higher antenna picking up more of an interfering station.  However, not as bad, it doesn't have random dots reaching to the top of the screen.  The main strong signal (at about 60dB on the better sounding Godar as compared with 72dB on the ST-2) is sufficient to suppress the interfering stations much better, even though it's 10dB weaker.  So the overall situation is similar to KSYM on the Godar antenna, just somewhat less so.  The meter says it's worse but the scope and actual results are that it is better.  But the sound and visible multipath on the scope are very good with either antenna, it's hard to tell the difference.

I see it may actually help for KSYM and KPAC that the Godar antenna is now bent, since this might add some directivity, suppressing other directions.  Fortuitously, it's bent in the direction of the transmitters of KSYM and KPAC.  The bend would tend to block stations coming from the sides FWIW.  Since it's working so well I don't need or want to touch it.  Perhaps a brand new antenna would be higher and therefore worse just like the ST-2.  If the Godar finally breaks off I'll probably have to replace it then.

So it appears that both the signal strength and the "multipath" meters are useless in situations like this, and probably many others.  You simply need to have a scope to get a feel of what's going on.

The best FM antenna is a big multielement antenna on a rotor on a mast, which has long been my dream, but I've hardly been able to contemplate setting it up, and basically nobody does that kind of thing for hire (at any price a mortal can afford, anyway).  I called "antenna installers" and all they wanted to do was tack up a small VHF antenna for digital TV to the side of the house, just as I have done with the whip antennas.  (I have two such brand new unused antennas in my junkpile, a legendary APS-10, and a classic Radio Shack FM-6.  How to put em up has always been the problem, as well as the grounding stuff I got sorted out for my whips.)

I also dug up a quick RF disconnect for one of my antenna cables and if I install another one on the other cable, I can swap antenna cables as needed (if I want to listen to KSYM on the kitchen tuner, for example).

KRTU still measures and sounds better on the ST-2, and by no small margin.  It has 10dB more signal strength AND less "mulitpath" visible on the scope.  Even less than on KPAC with the Godar.  It's obviously best to listen to KRTU on the kitchen tuner with the ST-2, which is what I always do.

KRTU on Godar DXR 1000

Despite having less apparent in-band interference, the multipath display for KRTU on the ST-2 still shows the V shape--and perhaps even finer than the stronger KPAC.  So I tuned around the dial a bit, and it appears all the stronger stations show that "V" characteristic, even with signals as high as 75dB.  So it appears that the V shape is just what the 600T tends to show on its scope outputs, given the way it's designed to produced the scope outputs.  That actually varies a bit from one tuner to another.  I would have expected the ideal display to be a flat line.  But "idealness" for the 600T appears to be how thick the trace of the V is, or if you can't even see the V at all because it's so blurry.  It's also great for tuning as the tiniest off-tuning results in a detectable tilt one way or the other at the bottom, which is flat right at the center.  The meters are more difficult to read or interpret.  For precise fine tuning, the center tune meter is fairly useless as it's correctly tuned at the far right of the inner bubble (a through tuner alignment would fix this).  The signal strength meter is not biased like that, but around the center tuning position there are only very small changes.