Tropospheric Scatter / Ducting on 900 MHz

train_radio_guy · Post by **train_radio_guy** » Tue Nov 14, 2006 6:16 pm

Has anyone had any experience with Tropo Ducting on 900 MHz? Looking into reusing some old cold-war parabolic dishes, in a point to point link that spans 85 files. Originally the system ran around 500 to 1000 watts. Any suggestions?

- trg,

AEC · Post by **AEC** » Tue Nov 14, 2006 7:41 pm

You would probably be far better off using high element count yagis on 900, those dishes are not small, add lots of wind loading and are a pain to align at times, especially if you are unlucky enough to have to do that during a windy day.

Cost will be more as surplus is usually far cheaper, but in the long run, you'll save rigging time and end up with a rock solid link nevertheless.

At 900, the signal will scatter due to the dish size, focal length and feed, so I really doubt you'll see any improvement usign a dish link.

Also, if you are one of those that are unlucky enough to live in the snow belt, then you have ice buildup to consider.....then those chunks falling on someone as well.

train_radio_guy · Post by **train_radio_guy** » Wed Nov 15, 2006 3:23 am

Well the dishes have been in place since 1956. One is located in Central Illinois, while the other is located in Northern Illinois. They are elevated about 30-40 feet on a solid lattice structure. The span between the two dishes is approximately 85 miles. Alignment seems to be good based on aerial photography, and a map overlay.

It seems there are four points that need to be addressed:

1) Method of transmission (NFM, SSB, etc..) . Which would be most effective?
2) Type of receiver, since this is not a LOS link. (Sensitivity Requirements?)
3) Feed line interface ('N' to waveguide transition)
4) Minimum power requirements.

The link originally utilized extremely sensitive receivers at both ends, and power amplifiers that ran around 1 kw, plus the Central Illinois site had a second antenna for diversity.

I've got leads on 1kw amps, but I'm concerned about the side effects to people & equipment in the path at either end. 1kw feeding a high gain parabolic antenna, that's low to the ground is rather disconcerting.

Remember how the high power microwave bombardment of the U.S. embassy in Russia affected the personnel in the building? It was the proverbial mouse running around in the microwave!
Lots of things to consider.

- trg,

AEC · Post by **AEC** » Thu Nov 16, 2006 7:30 pm

Since you mentioned the actual locations and the power previously used...FORGET IT!

900 is already chock full of paging, trunking and toy RF garbage, not to mention baby monitors and 'wireless' devices.

A.C power requirements, RF signage posted all around, then security for the site.

Not to mention upgrading the cabling and ensuring the transitions are capable of operating in the intended bandwidth.

What was the original frequency range of these dishes to begin with?

Since they have coax to waveguid transitions, I'd think they were probably for 5-14 Ghz. which would mean a complete realignment of the feed structure and focal points on both dishes for use at 33 CM.

That is a big undertaking in its own right.

Next is power generation for 900...a couple of 100 watt bricks heatsinked to the max would be about right, then additional high velocity fans, thermal cut-offs, RF bypass switching in case of PA failure.

You get the picture...C O S T L Y !

80+ miles is a huge volume of space to cover on 900, and 10 watts will not cut it, even with high gain yagis on both ends of this link.

Wireline would appear to be the lowest cost solution here, and it's 'on' 24/7 as well.

train_radio_guy · Post by **train_radio_guy** » Fri Nov 17, 2006 5:09 am

Keep in mind this an experiment for amateur radio. It's not intended to be a commercial communications link. A telephony based 56k or T1 circuit could be ordered, if that was the intention of the project.

Here's the 'low-down' on the sites:

Location A: The former DEW (Distant Early Warning) Line Training Facility, which is located southwest of Streater, Illinois. (http://www.lswilson.ca/streator.htm).

Location B: The relay station for the former DEW Line Training Facility, which is located just south of U.S. Route 20, west of Rockford, Illinois.
Both properties have since been sold by the government, and are privately owned. The dishes are intact, but the condition of the existing waveguide is unknown at this point.

Both dishes were previously used on 900 MHz, so resonance shouldn't be an issue. (http://en.wikipedia.org/wiki/User:Dual_Freq/WACS - See Operation section).

Both sites already have commercial power & are fenced-in by a security fence.

The waveguide to coax transition is still of some concern to me, but I've got a few friends in the commercial microwave business, so they may have a realistic solution.

As far as amplifiers go, if the paging amps don't work out, I'm looking at the TPL HMS series of amplifiers. Their power output is rather low in comparison, but they make a reliable product, and stand behind it as well.

I'm also still concerned about the mode of transmission. Narrow FM may not be the best method. Some hams in the Chicago suburbs regularly experiment with SSB microwave (10 GHz), and have had good results, although these are strictly LOS comms. I don't know enough about tropo ducting / scatter, to competently say which mode works best.

Obviously more research is required, and yes this project could be potentially costly.

- trg,

SlimBob · Post by **SlimBob** » Fri Nov 24, 2006 5:07 pm

I've thought about what it would take to get a KW on 902, and for the most part my brain keeps coming back to the kilowatt linears that used to be in the back of the ARRL Handbooks for 1296MHz. Of course, you might need 300W to drive those monsters to full power, so you still need your nukie.

AEC · Post by **AEC** » Fri Nov 24, 2006 9:09 pm

What is the main mode of transmission used through this link?

If SSB, then you'll require converters and filters to remove IMD products, as well as pre amps, and probably at least 2 circulators per station location....again, for removing IMD products from TX and RX signals.

85 miles is no cake walk, but should be possible using SSB, but noisy with FM I'll bet, then comes Fresnel zone refraction/reflections, path loss, actual link budget(S/N), feedline losses, and if remote access/control is added into this, the appropriate controlling system for both ends as you would need a method of killing transmit if problems occur.

Damn, 33CM tuned dishes....and with a built-in path to boot!

Y'all got yourselves a right fine system foundation, I wish you luck making it all work the way you hope it will.

train_radio_guy · Post by **train_radio_guy** » Sat Nov 25, 2006 12:13 am

Those amps in the ARRL Handbook may be our only option, since finding a solid-state 1kw amp for 900 MHz appears to be difficult, and the costs would probably be out-of-site. TPL's biggest amp for 900 MHz runs around 250 watts continuous duty.

We're still throwing around ideas on the best mode of transmission. I'm trying to determine what was originally used on the link, back when it was constructed in the late 1950's. Just guessing, I would suspect it was either similar to the format used on older analog microwave systems (base-band / channelized format) or possibly PCM (Pulse Code Modulation). In either case, we won't be able to recreate those formats, so we're left with either NFM or SSB. This is where the hams, who have experienced with LOS (Line of Site) 10 GHz SSB hops, will hopefully be able to lend a hand, and offer some direction.

Since the path not LOS, I suspect preamps will be a must. I understand they had some pretty 'hot' receivers on the original link; so recreating that in it self could be a challenge.

Initially, we'll probably just have someone at each end with the 900 MHz gear, and coordinate our activities by cellphone. If we find that contact can be made reliably, then we'll start considering all the aspects about controlling the equipment at both ends by remote control.

This whole idea was something the guys had been throwing around for a while as an interesting, if not unique project; not to mention reusing cold war radio sites for ham radio seem pretty darn cool. How realistic this project is, only time will tell.

- trg,

MSS-Dave · Post by **MSS-Dave** » Tue Nov 28, 2006 6:32 pm

Just for grins, I ran a few quick calculations on this...

Free space loss at 902 mhz for a 85 mile path is 134.4 dB or so. The Fresnel zone of the signal is around 300 feet at the middle of the path. Since your antennas are only 40 feet max above the ground, the distance to the horizon is less than 20 miles. I plugged some numbers into this calculator here..http://www.terabeam.com/support/calcula ... m.php#calc and found in theory that it ***MIGHT*** be done. I used 25 watts as transmitter power and -100 dBm as a receive figure which is pretty conservative for a good FM receiver. Absolute best is RX level of -60 dBm with this info. If the antennas are a large as you say, 20 dBi gain may be too low of a figure to use. It's all going to come down to what's in the path between the 2 antennas (other than the ground at 20 miles or so....

) and your RX sensitivity. I think most would agree, the best thing to do is to try it and see what happens!! You have a great test bed to experiment with there. I think you can find equipment to test fairly cheaply. Waveguide adaptors to N type are available new but real high priced, try used at Fair Radio or Surplus Sales of Nebraska.

If you can, give it a shot and post back with results. I'm interested if high gain can overcome some of the path issues. Of course, it's winter in Illinois, might work now but when the trees grow the leaves back...

Dave

train_radio_guy · Post by **train_radio_guy** » Tue Nov 28, 2006 8:03 pm

Right now, we've only got radios capable of producing 20 watts, so I'm looking at some amps from either Down East Microwave, or TPL. Both are going to be rather costly.

Securing access to both sites is also another obstacle. The site southwest of Streator is owned by the adjacent landowner (a local farmer), so that shouldn't be too difficult to work out.

The site west of Rockford is managed by a tower management company, and the adjacent 300+ ft radio is home to a cellular provider, several LMR systems, & the FBI, so that might be a bit trickier.

I need to get a good look at the vertex - feed point, on the rear of the dishes, to get a better idea of what I'm up against. If the opportunity presents itself, I try to get some pics.

- trg,

ke3ht · Post by **ke3ht** » Wed Dec 06, 2006 7:27 am

Hello all, I like this forum and found your info request interesting. I currently use 902mhz "SSB" at distances over 356miles on a regular basis. So 80 miles is nothing for me even with hills in the way. I have 65 watts and a 6foot dish on my Van. [/url]http://www.ke3ht.org/rover[url]for pic's. 150miles with the catskill mountins in the way is no problem if the other end also has a 6 foot dish. Your desire I would assume be for a high speed data link. I think you could better estimate your link abilities using some ham free software called "radio mobile". It is a very accurate link prediction system using radar maping and RF predections. Plug in all the factors and it will tell you if it will work. I also might suggest if you need MORE POWER seek ebay. I just purchased a 150watt amp for $35 and the same guy sells 300watt amps for $85. The trick is that these things need some minor changes mostly to get a decent RF connector we hams can make use of (like N). His ebay site has a link to the instructions to do this. I do not think 80 miles is very far at all for a slow speed data link. Anything above 19.2k will require special TNC's anyhow. If your talking auidio then NBFM will add some capture effect in the even you get some snow or high winds.[/url]

train_radio_guy · Post by **train_radio_guy** » Wed Dec 06, 2006 9:35 pm

So you think our 900 MHz link is feasible? Well that gives me some confidence that our endeavor isn't in vain.

Initially NBFM is going to be our first method of choice, especially since commercial gear for 900 MHz is fairly easy to come-by these days, thanks to the new technologies, and the 're-banding' folks at the FCC.

A typical 10-15 watt Motorola GTX900 fetches around $ 150 with all the trimmings. This should provide adequate drive for a bigger RF amplifier. The price range you mentioned sounds a great deal better, than the price quotes we've been receiving from TPL & Downeast Microwave.

If we can get a solid 100-150 watts into these dishes at both ends, I'm confident we should be able to copy one another. The quality of the signal is still up in the air, but access to the sites & feeding the dishes still needs to be resolved as well.

With the storms we've had so far this winter, we may have to wait until spring to perform any site testing. In the mean time, securing access to both sites, and acquiring the necessary equipment will be my two main priorities over the winter.

Any suggestions are welcome.

- trg,

Astro Spectra · Post by **Astro Spectra** » Mon Dec 18, 2006 6:17 pm

The NZ amateur radio record at 900 MHz is 385.5 miles done using Motorola Spectra radios!

The antennas were a tripod mounted 4 foot grid dish at one end and a 6 foot trailer mounted grid dish at the other end. In both cases the sites were no more than 200 feet above sea level.

There was a lot of fading on the over water path but with detectable signals at all times and occasional lifts to good quieting. Best results were with the 'Hear Clear' option enabled.

The equipment used consisted of Spectra radios driving 100W nominal amplifiers based on Motorola MRF847s.

So it can be done![/img]

train_radio_guy · Post by **train_radio_guy** » Tue Dec 19, 2006 7:39 pm

If the path is reliable enough, we were 'tooling-around' with various ideas, of what we could use it for, beyond the initial point-to-point contact. Here's some of the ideas: Amateur Television, long-range inter-connected repeaters on either end for local voice, phone patch, & weather storm reporting just to name a few.

With the right equipment, several of the applications could be 'muxed' together over the path. This all is very interesting, but before the cart is put in front of the horse, we will need to get access to both sites, and make some test transmissions. For the time being, the project is going to be put on hold, until the holidays are over, and everyone is back from vacation. Weather permitting, we'll try to get the ball rolling early next year.

SlimBob · Post by **SlimBob** » Tue Dec 19, 2006 11:53 pm

train_radio_guy wrote:With the right equipment, several of the applications could be 'muxed' together over the path. This all is very interesting, but before the cart is put in front of the horse, we will need to get access to both sites, and make some test transmissions. For the time being, the project is going to be put on hold, until the holidays are over, and everyone is back from vacation. Weather permitting, we'll try to get the ball rolling early next year.

This is something I mill about in my head every time I think about linking sites at 900MHz and up -- why do one voice channel at 5KHz when we can multiplex several, and some data? Why do 9600 BPS packet when you can do 2Mbit/s+? Next thing you know, you're looking at trying to pass wide signals through repeaters with linear amplifiers in really inaccessable sites.

It's a common denominator problem. You have to have an accessible common denominator. Which is the venerable 900MHz radio, otherwise the barrier to entry is too high to get casual hams involved. Granted, your toys are more specialized than the network in my head. I'd start with FM first -- with the equipment we have on 900MHz already, we can do full-duplex FM, and multiplex a few carriers as well.

train_radio_guy · Post by **train_radio_guy** » Wed Dec 20, 2006 3:21 pm

The possibilities are essentially limited to the constraints of the imagination, the availability of the equipment for the desired technology, and the budgetary/fiscal backing of our (hopefully generous) benefactors.

A simple NFM test is what we'll initially start with, and take things as they develop from there. Realistically, there is any number of things that could effectively bring this project to a screeching halt. That's why I'm taking small steps, and working on one issue at a time.

In all honesty, I'd be happy to make contact between the two facilities using a pair of Motorola high power GTX900 mobiles, each interfaced directly to the parabolic antenna. That said, my instinct tells me that we'll need a significantly larger RF source; somewhere in the neighborhood of several hundred watts. The only thing working in our favor is the lack of foliage, due to the season.

trg,

SlimBob · Post by **SlimBob** » Thu Dec 21, 2006 4:55 am

train_radio_guy wrote: In all honesty, I'd be happy to make contact between the two facilities using a pair of Motorola high power GTX900 mobiles, each interfaced directly to the parabolic antenna. That said, my instinct tells me that we'll need a significantly larger RF source; somewhere in the neighborhood of several hundred watts. The only thing working in our favor is the lack of foliage, due to the season.

Nah -- you've got something like 20+dB in gain from those dishes. So feeding a 25W radio into the dish gives you an ERP of 2500W. Remember, the pathloss formula is logarithmic.

http://gbppr.dyndns.org/path.main.cgi

Important to keep coax losses down -- for experimentation, you may want to see about putting a short piece of coax in and the rig close to the feedpoint.

N7LXI · Post by **N7LXI** » Thu Dec 21, 2006 1:51 pm

I cross posted the original post to another list I belong to that deals with AT&T Microwave stuff... http://groups.yahoo.com/group/coldwarcomms/

There were a few interesting replies.

This might be a good list for the anyone interested in this stuff to join. LOTS of info on these (and other) old cold war sites, and there are plenty of guys that worked with the gear when it was new.

train_radio_guy · Post by **train_radio_guy** » Thu Dec 21, 2006 7:34 pm

Thanks for the heads-up. I put in a request to become a member of the cold war comm. Yahoo group. With a project of this magnitude, any help offered would definitely be appreciated.

-trg,

illini52 · Post by **illini52** » Sun Dec 24, 2006 1:54 pm

http://www.radomes.org/museum/scripts/a ... ngCenterIL

I am not too far away from the Streator site.

train_radio_guy · Post by **train_radio_guy** » Wed Dec 27, 2006 9:26 pm

Thanks for the info. In my searches of the web, I hadn't stumbled across that web site. Much better pictures of the parabolic dish used for tropo scatter. I've got an old AT&T microwave guy that's offered to lend a hand on the tech support side. After the first of the year, I'm going to begin securing access to both facilities.

I work near the Streator Training Facility, and live near the Seward Relay Facility, which was briefly mentioned in an article of Popular Commincations in the early 1990's.

The Streator & Seward dishes are intact, but a closer inspection will hopefully determine how feasable this little project may be. Once I know more, I'll update the thread.

Thanks,

trg,

n9wys · Post by **n9wys** » Sat Apr 21, 2007 7:42 am

Sorry that I'm coming in to this discussion a bit late... Have you thought of using a Motorola paging PA (available on Flea-Bay) for your amp?

I picked a 150W unit up there for about $60, IIRC... I've also seen 300W ones, and one or two 600W ones. They pop up from time to time. (Probably a LOT less costly than buying new.) The conversion details for 900 amateur are here:
http://www.vhfsouth.com/tutorials/902.htm

I haven't played around with my amp - in fact, I just "donated" it to another ham who is building a 900 repeater close to me... and I have an old Glenayre paging transmitter I'm planning on using the PA out of for my machine...

Keep posting - being in N/E IL I'm interested in finding out if you are successful!!

willbartlett · Post by **willbartlett** » Mon Apr 30, 2007 4:43 pm

Man, I wish I'd been paying attention to this thread last fall. I had a half dozen of the Glenayre 1/4 kw 900mhz amps in my basement. I think I gave away the las one a few weeks ago. I can see about reclaiming them, might not be too troublesome. If I can come up with 2, I'll offer them up for the price of shipping, which might not be too bad. We have a bunch of 900 machines here near boston runnung 200+ watts with these amps.

Alternately, I have a car full of 2.1 or 2.6 ghz read (read, 2304 conversion) tropo gear. A company around here had an s band link system with 10-50 watts on either end with 4 or 6 foot dishes and had better than 90km at 512kb with some fancy dsp on both ends.

The other thing to remember about the dew line stuff on 900 is that the "hot" reciever may have been no better than a 5-6db noise figure. pretty respectable 50 years ago, but you can improve on that by about 5db for short money nowadays, even from mini-circuits. I can try to dig up some specs. My job was involved with the development of some of that stuff. I love the coldwarcomms group. Lots of amazing stuff there.

Tropospheric Scatter / Ducting on 900 MHz

Tropospheric Scatter / Ducting on 900 MHz

Ham links