[TowerTalk] Fiber optics for amateur radio..

Jim Lux jimlux at earthlink.net
Tue Dec 4 00:45:17 EST 2007

-----Original Message-----
>From: "Richard (Rick) Karlquist" <richard at karlquist.com>
>Sent: Dec 4, 2007 12:11 AM
>To: Jim Lux <jimlux at earthlink.net>
>Cc: towertalk at contesting.com
>Subject: Re: [TowerTalk] Fiber optics for amateur radio..
>Jim Lux wrote:
>> I haven't browsed through the Mouser or Digikey catalogs for this, but it seems that there must be analog transmitter/receiver components that are fairly inexpensive for HF thru UHF use (I know there are microwave links for fiber from Ortel, etc., but they arent' cheap)
>> I wonder what it would take to build a HF preselector/amplifier/fiber transmitter and corresponding receiver.  You could put the widget up the tower, and completely eliminate the whole "feedline as a radiator" or "feedline carrying lightning" issue.  This is a receive kind of application.. It's probably a bit bold to suggest putting your kilowatt linear up the tower, but the idea is there.. 
>I used to design fiber optic communication test equipment at Agilent,
>up to 40 Gb/s or so.  We looked into sending analog signals optically,
>but all the optical hardware is now highly integrated assemblies that
>are designed to send digital signals only.  The closest you could get
>to analog transmission would be to frequency modulate an oscillator with 
>your spectrum and then FM demodulate at the other end.  You might
>as well just digitize the spectrum in that case.

We've been getting analog fiber links for microwave signals (we use them for fiber optic delay lines, for instance), but they're not cheap.  I think the modules have something like a laser diode in a feedback loop.

I think Ortel is now part of Emcore, who have products described as:
"EMCORE's linear fiber optic links are an excellent alternative to using coaxial cable systems to transmit 10 MHz to 40 GHz signals. They offer significant improvements in reliability in microwave communication networks by transmitting the RF signal in its original format. As a result of these properties, these microwave transmitters provide significant improvements in signal quality for a wide variety of applications, including antenna remoting, timing and reference signal distribution, telemetry, measurement, and delay lines."

>> Or, for relay/tuner controls... 
>> A complete link is still in the $200 range (2 ends+cabling) but that's not totally unreasonable.  For instance, if you had to string several hundred feet of multiconductor rotator or SteppIR controller cable.. copper prices are rising
>> For instance, 10 conductor AWG22 cable runs about $1/ft these days (Mouser catalog page 904), so a 300 ft run is going to be $300.  Compare that to a couple 75 dollar boxes on the ends and a cable at $20-30.. Browsing page 2075 in the catalog, it looks like the low end cables are about $0.03/ft....Single mode (higher data rate) runs about $21 for a 20m cable and $16 for a 10m cable ($5/10m or $0.15/ft for the cable)
>I already run CAT5 (6 cents a foot new at Fry's) to my tower and only 
>have the heavy duty stuff running up the tower.  K7NV sells a remote 
>rotor controller to do this.  Using Power Over Ethernet, you can use
>CAT5 for both signalling and moderate amounts of power.  

I just ran across someone selling fiber (duplex LC) to 100BT converters for $20 each (as a discontinued item, I grant you).. but still.. that's pretty darn cheap, and you get galvanic isolation for free.

To remote
>a linear, this is the strategy I have arrived at:  Step up
>the power voltage to 480 VAC using a 5KW isolation transformer
>wired as an autotransformer. 

Hmm. hoisting that 5kVA transformer to the top of a 150 foot tower could be an exciting proposition.  But hey, the power company does it.

 At the linear, step it back down
>to 240VAC. I have several of these transformers that people
>gave me for free, if I could haul them away, hi.

Those beasts are quite common.  Especially in places where they're decommissioning factories with heavy equipment.  The factory uses 480V delta to run all the motors and heavy loads, and they use small dry transformers to get 120V or 240V to run appliances, lights, etc. at point of use.

  This allows me to
>use fairly small wire.  But the other magic bullet is to get a
>solid state linear with a "Universal" switching power supply.  These
>supplies automagically run on any voltage between 85 and 260V.
>Thus it doesn't matter if the line voltage at the linear sags 10%
>or 20% on key down.  The output is always 48V.  Another strategy is
>to put batteries at the linear and only use the connection to the
>power company for charging them.

That's the direction I've been heading (on a smaller scale) for portable phased arrays.. Accept the loss in the wire, and put a good wide range regulator at the far end.

More information about the TowerTalk mailing list