PSN-L Email List Message

Subject: VHF telemetry
From: sean@...........
Date: Thu, 9 Nov 2000 15:32:57 -0600 (CST)


Has anyone responded regarding your questions about radio telemetry?

Over the years I have been able to operate several RF telemetry networks
in difficult situations with quite long distances and many repeaters.
The "standard" station is a 1 to 6 channel seismic pre-amp/VCO unit
with FM audio carriers feeding a narrowband FM (5khz) VHF transmitter
with 100 milliwatts (0.1 watt) power operating on 50 milliamps at
13 volts from a 1100 Ampere-Hour Zinc-Carbon-Air battery pack for up
to two years.  The frequencies are from 165 to 175 Mhz, with 5-element 
Yagi antennas with 6 db gain at each end of the link. 

The largest network has been the 36 station network in the New Madrid 
region. We could manage distances up to 70 km with the station
antenna about 10 meters up in a sturdy tree and the receive antennas
30 meters up a cable TV tower. The terrain there is very flat, but
we did use the highest hills as remote station sites. Because we had
only 12 RF frequencies to use, the layout was critical so we wouldn't
interfere with our own signals.

In the Aleutians, the mountainous islands with no trees made transmission 
easy from 3 meter antenna masts, again with some 50 km links. To see its
extent, get your map of Alaska and look for the islands Adak (the central 
station), G. Sitkin. Umak, Kagaglaska, Tanaga, and Kanaga, with repeaters 
on SE Tanaga and the Yakak peninsula at the SW corner of Adak.

In the Aegean network, I put a station on western Crete transmitting 
to Milos and repeating to the SE Polyponeseus and then repeating to Athens.

You might have trouble finding maps for the network from Nurek to
Dushanbe in Tadjikistan and at Totugul in Kirgisia, but again we used
high hills in the relatively barren Tien Shan and Hundu Kush, with 2 meter 
pipe masts for the antennas to transmit as much as 100 km. Our highest 
station was at 4100 meters. The problems there were that the Russian 
"air-cell" batteries were very unreliable, and the nomadic horsemen would 
lasso the antenna and pull it down.

Regarding the performance of a low power VHF radio link for narrowband 
seismic telemetry:

You can estimate the effectiveness of a VHF link over any given terrain
knowing the transmitter power, the antenna gains, and the receiver threshold
for a given "fade margin", the percentage of time that the signal received
will be too weak to use. And then you can throw a mountain in the way
and estimate the added loss.

Also be mindful that VHF refracts downward in the lower atmosphere, 
increasing the propagation over the visual horizon as if increasing the
earths' radius by 1/3. This increase in distance is often estimated 
when plotting the profile of the propagation path by using 4/3 radius
profile paper, or a single curve representing the "flattened" earth.

Assume that your typical VHF receiver unsquelches at a signal of 1 microvolt,
at 50 ohms, which is -137 db. (the db power ratio is 10*log(power[in watts]
referred to 1 watt (0 db)).

A 0.1 watt transmitter puts out -10 db to the antenna. The yagi antennas 
at each end improve this by +12 db. The propagation path loss is estimated
by A = 37 + 20*log(f) + 20*log(d), where f is Mhz and d is miles. So if f is
170mhz (+44.6) and d is 50 miles (80km) (+33.97), the path loss is 115.58 db.

So we get -114 db to the receiver, for a fade margin of (-10 +12 -116) -(-137)
or about 23 db, which is a relatively strong signal.  At 20 db, the signal
reliability is 99%, but at only 10 db, it is 90%, and subject to atmospheric
variables, like inversions that refract it right out to space. For seismic
telemetry minor fades and bursts of static can usually be handled by the
data discriminator if the audio carrier filter is properly tuned and the
active squelch is adjusted properly. Very rarely do these trash a seismic
event since they only last several seconds.

Bear in mind here that the "line of sight" constraint for VHF propagation 
is for the most part wrong: the antenna is not putting our a "beam" until the
frequencies go above 1 GHZ, but rather is launching an electromagnetic wave 
in free space. An object in the way is a scatterer, not a brick wall, as long
as it resides near the center 1/3 of the propagation path. At Adak, we 
transmitted directly toward a mountain range in the center of the path from 
the the station and repeater on Yakak peninsula.

If we throw a mountain in the way, its addition to the path loss can
be estimated using nomographs. A "knife edge" peak near the center of a
50 km propagation path will cause an additional loss of -10 db if it rises
50 meters above the 4/3 curve. If the mountain is 100 m high, the loss
is -14, and if 200 m high, the loss is -18. These losses are highly dependent
on the geometry of the mountain, and may be twice these values. But I have
found that a profile grazing within about 30 m of a ridge top will add a 
loss of about -10db, and a conical peak in the path is of little concern.

So what if the fade margin is less than 10db?: (and some things I have done)
1: The received signal may be full of static, which is often largely 
 eliminated by the narrow-band carrier filter of the seismic discriminator.
2: Some receivers specify a threshold of 0.5 microvolt, or -143 db, or
 about a 6 db improvement. Note that this is an optimal performance value. 
3: The optimal consistent performance of both transmitters and receivers
 assumes that they stay bone dry (ie. in a monitored, desiccated sealed 
 container), since humidity will change the geometry of the tuning coils.
4: Many transmitters can be "peaked up" for an increase in power to 0.2
 watt, for an increase of +6 db, at the expense of battery longevity. (In the
 Aleutians, we needed 2 years of power from the battery pack, limiting our
 total station current to 65ma (at 12.6 volts)).
5: Receivers can be fine tuned for a given link to lower the noise by 10db
 or more. I have packed a portable oscilloscope 300 feet up cable TV towers
 to make an otherwise marginal signal strong and reliable. This, or course,
 has to be repeated if either the transmitter or receiver are changed.
 Prior to deployment, pre-tuning each radio link in the lab using a crystal 
 frequency standard (10^-6 accuracy) can make a big difference.
6: Arrays of antennas (at the receive site) can improve the signal by
 10 to 20 db, but add cost and complexity.
7: It is best to carefully plan station sites beforehand, but sometimes
 relocating a station to a better/higher location can do wonders. Changes
 in the seismic array geometry by several kilometers have little impact
 on the seismic event location accuracy, but a large impact on site noise 
 and the reliability of the radio link.

I hope these comments are helpful.

Public Seismic Network Mailing List (PSN-L)

[ Top ] [ Back ] [ Home Page ]

Larry Cochrane <cochrane@..............>