[TriEmbed] OT: RF questions

[Pete Soper]

I modeled 1/2" diameter elements with EZNEC and the length of each 1/4wl 
element comes out to 29.1 inches when cut for 98MHz with the antenna 10 
feet up.

The graph of SWR vs frequency from 88 to 108MHz is below.
<http://triembed.org/images/FM-DIPOLE.png>

The closer to resonance the lower the SWR and greater proportion of RF 
out vs turning the power into heat. So if you end up needing to transmit 
at, say, 89MHz you'd simply make the elements 98/89 of the above 
dimension and it should keep the SWR as low as it can get. If you were 
transmitting at 107MHz you'd multiply by 98/107.

But the other problem with going unbalanced into a dipole is that it 
doesn't necessarily radiate like a dipole (i.e. two lobes perpendicular 
to the elements). As Dan mentioned, the feedline ends up radiating and 
it tends to be at wonky angles relative to the axis of the antenna 
elements. The coax coil (or purpose-made balun below) solves this by 
isolating the feedline from the antenna. It's a pain to have the antenna 
only be effective for a piece of the azimuth range you need. From the 
description you probably want an omni pattern. The gain off the ends of 
a horizontal dipole is terrible and the gain falls off pretty severely 
more than around 40 degrees right or left of the broadside direction.  
So in addition to a balun you might consider making the dipole vertical. 
A vertical dipole is omnidirectional outward with the nulls up and 
down.  But the balun is key to getting a predictable pattern.

You can get baluns from Digikey, by the way. The MABA-011040 
<https://www.digikey.com/product-detail/en/macom-technology-solutions/MABA-011040/1465-1702-1-ND/5131112> 
is rated for 1-300MHz and doesn't look like it would be hard to solder. 
This could go at the antenna and then you'd use a simple run of coax to it.

-Pete AD4L

On 9/28/20 12:29 PM, Brian via TriEmbed wrote:
> Hi Folks,
>
> I know there are several radio-smart people on this list, so I hope 
> the rest will forgive the noise as I pose a couple questions here.
>
> The questions first; I'll provide background afterward.
>
> I have settled on a digital FM transmitter chip (Silicon Labs' 
> SI4721), and a simple dipole antenna.  Two questions arise from this 
> decision:
>
> 1. The transmitter has a single-ended RF output, but will be fully 
> isolated from earth ground in operation (running on batteries or an 
> isolated AC/DC supply).  A dipole is a balanced load, but since 
> "ground" of the radio circuit is isolated, can I just treat it as 
> "balanced" and connect the circuit ground to the other half of the 
> dipole?  Or do I really need to use a balun for a proper balanced 
> output?  Coverage area actually needs to be very small (< 100'), so 
> I'm not majorly concerned with impedance mismatch losses, etc.
>
> 2.  I'll be using 1/2" copper pipe as the elements, held inside a 
> larger PVC enclosure.  What's the best way to bond wires to the pipe?  
> Should I just solder them on?  Tap a hole and use a screw to clamp 
> them?  Some kind of shark-bite approach?  Does it even matter at all?
>
> Here's the background:
>
> My church does a Christmas program called the Drive-Thru Christmas, 
> which is made up of five live-actor scenes distributed around our 
> parking lot.  Guests are typically given a narration on CD which they 
> play inside their vehicles as they move from scene to scene.  In order 
> to improve our social isolation this year, I'm doing some R&D on the 
> "talking sign" idea, using five separate short-range FM transmitters 
> to broadcast the scene's narration to the guest's FM radio in their 
> car. Each transmitter would broadcast on a different frequency, and 
> some system would step each transmitter through the list of 
> frequencies in time with the car's movement through the scene, so we 
> can maintain our 5-car pipeline but not require the guest to re-tune 
> their radios.  We tried using an internet streaming option last year 
> (for folks with smartphones linked to their car stereos) but that, I 
> hear, was an abject failure with many people unable to access the stream.
>
> Well anyway, thanks in advance for any advice!
>
> Cheers,
> -Brian
>
> _______________________________________________
> Triangle, NC Embedded Computing mailing list
>
> To post message: TriEmbed at triembed.org
> List info: 
> http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
> TriEmbed web site: http://TriEmbed.org
> To unsubscribe, click link and send a blank message: 
> mailto:unsubscribe-TriEmbed at bitser.net?subject=unsubscribe
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.triembed.org/pipermail/triembed_triembed.org/attachments/20200928/3e910b14/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: opiadpmmadeneekh.png
Type: image/png
Size: 98588 bytes
Desc: not available
URL: <http://mail.triembed.org/pipermail/triembed_triembed.org/attachments/20200928/3e910b14/attachment.png>