-----Original Message-----
>From: David Robbins K1TTT <k1ttt@arrl.net>
>Sent: Dec 4, 2007 3:32 AM
>To: towertalk@contesting.com
>Subject: Re: [TowerTalk] Fiber optics for amateur radio..
>
>Fiber is nice for transmitting control and data information. It of course
>is a bit more complicated than just running copper with the same
>information, but has advantages of being less susceptible to rfi. There is
>one big thing that everyone forgets though, you can't transmit POWER via
>fiber. You want to turn a rotor via fiber? You have to run copper to the
>rotor for power, need relays or some other electronic converter to take the
>control from the fiber and convert to signals to turn on the relays. So you
>have taken what was a simple copper wire from a simple supply in the shack
>and added at least 2 more complex (relatively) opto-electronic converters.
>Want to get position indication back? You need another fiber (or you could
>re-use the same one at the expense of more complex converters), some kind of
>digitizer or analog to fiber converter, calibration, and decoding to some
>kind of display in the shack... much more complex and prone to failure than
>a simple copper connection from a pot to the meter in the box.
I would venture to say that as technology improves, the model of unregulated
power out and a fiber each direction, with some intelligence at each end, will
be more reliable and less expensive than the "run separate wires for each
function".
Indeed, you have to get power to the loads, but nothing says that power has to
come from your shack, from a special box (rotator controller or relay
controller)..
As far as reliability and calibration goes.. I think that with decent design,
the remote sensor and digitizing scheme is better. No worries about changes in
the properties of the wire or connections. Once you've bought into having any
smarts at the far end, then it can do multiple things.
Say you want a computer controlled rotor like the Green Heron box. The optical
fiber approach isn't much different, except that some of the functionality now
lives up on the tower, rather than back in the shack. Either way you have to
convert a pot position to a digital number. Either way you have to turn power
on and off to a motor. Either way you have to talk RS232 or some other
interface.
Likewise the relay boxes..
I'd note that for the last few decades, the trend is away from the "separate
wire for each function" in almost all other industries. CAN bus and similar in
automobiles, MIL-STD-1553 for avionics systems... and all driven by some of the
same basic reasons: reduce the number of wires that need to go from point a to
point b. Each wire has a cost, not only in the raw material cost and weight
(important in cars and planes), but in testing and manufacture (each connector
pin costs something to install and test) and in increased reliability.
Doubling the number of pins doubles the probability that any one pin will
fail..Pretty quick (around 4 pins), it becomes more reliable to multiplex the
data on a redundant pair of wires.
Want to put
>the rx/tx up the tower and feed by fiber... great, no feedline loss, but you
>still have to run power up the tower to the box with the fiber so you still
>have to worry about lightning protection, but now its in the direct line of
>fire on the tower. And it has to be all remotely controlled via fiber...
>final transistors fail in the middle of a contest??? Can't just pull a
>spare rig off the shelf and plug it in, you have to drag it up the tower.
Heh, heh.. just put the spare rig on the tower.. think of the desk space it
will free up..
I don't want to claim that this is the be-all, end-all today.. just that it's
got a lot of potential, and the cost is no longer prohibitive. 20 years ago,
the cost alone was a significant disincentive. Maybe, folks at the edges of
the hobby will come up with ways to address the concerns.
Jim, W6RMK
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|