One way would be to use a common cable tv splitter, and feed two dipoles, one vertical, one horizontal.
Situate the dipoles so that you have a cross. Or you can put the horizontal on the floor and the vertical above that at about 1/2 wavelength of F.
Should solve the multipath problem.
RFB
The dipoles sound like a very easy to do approach. I will include it in the testing plans.
But I am looking also for an opinion about my "L" proposal described above. I will be trying it for comparison with other methods.
Might work if the TX your using had more than 11 nanowatts coming out of it.
A simple L is not going to get you dual polarization. More likely that will act like a standard vertical whip even with the lower horizontal element.
Separate the vertical and horizontal of the L so they are fed independently, then you will get your c-pol.
RFB
The C.Crane type of FM transmitter is most commonly used indoors in the attempt to provide wireless stereo audio to various locations within the dwelling.
The telescoping antenna on the C.Crane is 11", nowhere near a reasonable fraction of a wavelength.
The part 15 rules on certified transmitters require using the installed antenna, any changes to the antenna make the certification invalid.
Strangely, the Wholehouse 2.0, a certified FM transmitter, has a stub of a wire at the RF output, a wire-connector, and a long wire, with a formula for calculating a full wavelength, which interestingly is about 3-meters. What's obviously strange is that the user can cut his own antenna without losing certification.
We elected to hang a 1/2 wave vertical antenna from an indoor "art project" bamboo tower, and made it a four-element cage-monopole type.
Despite the advantage of the lengthier antenna, the Wholehouse does not perform any better than the C.Crane, indoor reception is great for about four feet and doesn't go around corners or through walls.
None of that matters in our case, all we need is a solid link between the FM transmitter and two receivers which feed AM transmitters. Our vertical link located on the same floor works fine, but the downward link 1-floor below is riddled with multipath, because the bamboo tower puts a null toward the downward direction.
That explains why we are re-engineering the tower to provide a horizontal signal for downward thrust.
Questions about part 15 compliance are virtually moot in this scenario, because there is no outward penetration beyond the building walls.
We elected to hang a 1/2 wave vertical antenna from an indoor "art project" bamboo tower, and made it a four-element cage-monopole type. ... That explains why we are re-engineering the tower to provide a horizontal signal for downward thrust.
Acceptable results might be produced by inclining even a simple, linear, non-cage radiator so that the perpendicular bisector from that transmit antenna is directed toward the receive antenna on the lower floor.
....use a simple piece of wire tilted towards the lower deck where you need more signal. Arrange it so that one angle looks below and the other looks up.
Worth a try.
RFB
Inclining the perpendicular bisector of the receive antenna on the lower floor toward the transmit antenna also may be beneficial.
(edit) I see that RFB posted something similar a few minutes before I first posted the text above.
Earlier I described an "L" shape which seems to be the same thing being here proposed. Am I in tune?
And I agree, the receiving station down below needs also to be correctly "aimed".
Earlier I described an "L" shape which seems to be the same thing being here proposed. Am I in tune?
An L-shaped radiator doesn't have the same radiation pattern as the linear (straight) radiator I described.
But nevertheless it might be useful in your situation, especially if the pattern of the receive antennas you need to serve are compatible with the radiation pattern of that transmit antenna.
Going back to what rich suggested I believe, to put it in other words, a straight wire is being described, but instead of being vertical, it would tilt toward the downward target, making it equi-distant between vertical and horizontal.....
The receiving antennas could be angled accordingly.....
Am I gaining speed?
Having taken a crash coarse in analyzing the spectrum analyzer, I actually found my FM whisper-signals.
The FM band on the scope looks like an insane cartoonist's animated drawing of a city of high-rises bursting out of an empty field on either side.
By narrowing down the signals I first found 101.9 coming from the C.Crane, and it was harder to fish out 107.1 from the Wholehouse 2.0, which registered lower despite having a bigger antenna.
Once I had 107.1 singled out on the scope I cut all but one of the elements comprising the cage-monopole, and the power level remained the same on the scope.
Since then I've been trying to think of what metal I have to build a tilted straight antenna, and how to put that on the tower so the structure doesn't tip over.
The Yagi idea is much better than the folded dipole that's on the receiver. I think.
At first I "ran with the idea", to use an expression.
Then I walked around the idea, and finally sat on it.
The "idea" was to build a straight-line antenna mounted in a tilted direction which would put signals both downward and backward; that is, toward the back of the house.
Mechanical realization of the idea seemed to be impeded by integration questions; that is, fitting it into indoor space so it wouldn't poke somebody in the face or look like an avalanche in progress.
What was needed was a temporary quick-fix.
Down below at the LPB CC transmitter the Technics FM tuner has a signal-strength meter, which was showing a fluttering signal between 2 and 2.5 on a scale of 5. What I did was move the folded dipole until I captured a 4.5 signal on the meter.
We are now functional, probably compliant, and the tilting antenna can await a brain-storm as to how to make it stand there, tilting.
What the preceding entry illustrates, is something particular to indoor antenna design, that of making it part of the decor, without creating hazards or ugly "wife offending clutter".
Rich's idea of a linear antenna tilted so as to serve both downward and forward poles of transmission is simple to describe and imagine, but putting it in the room stumps the smartest interior design expert.
Contents of indoor rooms are mostly vertical or horizontal. Perhaps the back of chairs and sofas have some "lean", but not much else. Well, dogs and people may tilt occasionally.
Getting to the point, the very heart of indoor antenna science/art consists of 50% technology/50% interior architecture.
This project, the design of indoor antennas for part 15 radio, will not accept a schlock approach. A "schlock" approach is that often taken by a short-cut artist with no social skills. Part 15 is a no schlock operation.
The geometry of antennas is on my mind tonight. They have towers that stand straight up. There are self supporting towers wide at the bottom narrow on top, but their geometry is for support, not RF efficiency.
FM antennas are more swirly, with curves and twirls to get that signal into the vertical and horizontal planes.
The cone antenna is an interesting combination of playroom geometrics.
But in part 15 the antennas tend to be straight up, side-to side, 2-dimensional at the most.
I am looking for 3-dimensional antenna designs. The helical "slinky toy" example is a start.
What else do we know about 3-D antenna geometry?