The Microwave Station

One way to achieve success in a VHF contest is to locate your station on a mountaintop and work stations that no one else can hear, or even knows is there. But, to win a VHF contest today, that is not enough. Not only do you have to be on a mountaintop, but you must have a VERY decent station on every microwave band, and then have the stations to work on those bands. Back in the early 60's, that was somewhat the case as well. W1MHL/1, a perennial winner in New Hampshire, would send out portable stations to various spots, all with the intention of working those stations for contest points. I remember reading that the 2.3 GHz station was a pulse transmitter and a wideband receiver. The sections were the individual states, so the portable stations went to Maine (Mt. Agamenticus), Massachusetts, Vermont (Mt Ascutney I think) and Connecticut too! Such operations were a good way to test out the gear and make real contacts over real paths. It also increased the score of W1MHL, and generated interest that begat more activity. Later groups that came after W1MHL included W1FC, and then W2SZ among others, who all utilized the microwaves to advantage. Their pursuit of ever higher bands to operate, and then making rigs to send out to remote sections and grids, raised the bar for what was required for a high score. Today, the use of microwaves is absolutely required for a high score in all VHF contests, and especially in the September affair.

Our contesting plan was to build the absolute best microwave station that we could muster. The location here is good, but it is not a mountaintop, so we are at a disadvantage to begin with when compared to the likes of K8GP, W2SZ, or other "mountain toppers"! The effort to improve the microwaves has been spread out over a number of years, and recently has become a big factor in our scoring. I figured that a very high performance microwave array would be a big equalizer in the September contest. If you look at the K1WHS September, 2001 contest score, with its' best tropo in 20 years conditions, and compare the microwave results we get today under dead band conditions, you will find that our scores and grid totals today are better on all bands above 432 than they were in that historic 2001 tropo opening. Why is that? There are two reasons. The first reason is that each band is now so much better, equipment wise than before. Reason #2 is that there are a few more grids with activity over what was available in 2001. I place 85% of the success at the increased horsepower and receive capability that has been built into the station. Another portion of our plan was to activate the empty grids to our Northeast. I am not sure why, but there is no UHF activity from Eastern Maine and VE1 land, with only a few exceptions. That area of the planet is sparsely populated, for sure, and employment opportunities there are slim for anyone with a technical nature. Only chainsaw repairmen and marine engine mechanics come to mind. There is no electronic industry. VHF hams there are as rare as hen's teeth, so we have to send someone out there to work all those empty grids.

Below, you will find a description of each band, interspersed with comments concerning efficiency or circuit protection, that goes hand in hand with setting up a full bore contest station. Pictures are being added as time permits.

The Microwave operating area. The righthand K2 runs 903 thru 3456. The left K2 runs 5, 10 and 24 GHz, while a third one stuck in the far rack powers the auxilliary microwave setup on 903 thru 3456 with a separate antenna system. The big rotor boxes control Prosistel PST 71 and PST 51 type rotors. The aux microwave rotor is a one of a kind Japanese rotor with a nice digital readout. A few solid state amplifiers are visible in the far rack as well. The large one is a 300 watt 903 ss amp.

33CM

The 903-3456 MHz station consists of an Elecraft K2 IF powering individual transverters for 903, 1296, 2304, and 3456 MHz. These antennas are all co-located on the same mast so that makes sense. There is an interface box with a 144 MHz DEM transverter and automatic band switching for each of the four bands. Bandswitching is initiated at the keyboard of the logging computer. (Writelog) The bandswitch signal changes bands on the K2, which in turn sends the required information to the Microwave Interface Box (MIB), which selects the proper transverter and routes all IF and keying lines to the correct place. The K2 has a neat accessory called the KRC-2. This little box will control external station accessories such as amplifiers and transverters, bandpass filters etc. all with the K2 bandswitch data. For more information check out: www.elecraft.com/KRC2/krc2.htm We use the KRC2 to change bands on our homebrew MIBs. In actual operation, the microwave operator types in the frequency for the sked on his keyboard, and then starts sending with the keyer (or keyboard). He can also grab the microphone and start calling the station on SSB. How easy is that? In the past, many operators were hesitant to operate the higher bands here, as the transmit switches and bandswitching knobs were not automatic. In fact, they changed at almost every contest as things were added to the station. Things got pretty intimidating at the microwave station as a result. With an automatic band changing station, the operation is seamless, and anyone can operate 10 or 24 GHz as well as they can figure out how to operate on 6 or 2 meters! Another good choice for ease of operation, is the Software Defined Radio, or SDR. It makes for simple QSY on all the microwave bands. Similar automated circuitry there can really streamline operation too! I have used the SDR-1000 and really like it on the microwaves for finding stations calling while off frequency. (A common event as we all know!)

On 33 cm, we use a Downeast Microwave transverter with a 145 MHz IF driving a converted solid state amplifier for about 400 watts output. Prior to 2007, we ran about 200 watts output. Both amplifiers utilized MRF899 type FETs operating at 28 vdc. The latest PA has 5 transistors in the final stage. DC power is a 75 amp switching 28 vdc supply. Feedline is 1 5/8" heliax up to the antennas, which are four times 47 element loop yagis in an H frame at 120 feet above ground. The receive preamp, coax relay and filtering is located in a weatherproof box mounted near the antennas. A separate receive feedline (7/8" heliax) runs down the tower to the receive port in the shack. Having separate RX and TX feedlines simplifies the tower mounted switching somewhat. The antenna system is located at the bottom of a 3" diameter mast that turns courtesy of a Prosistel PST-71 rotator. A 3" mast was selected to try to hold the microwave antennas on a proper heading even in high winds. With high gain on all bands, the beamwidth is rather tight, so dimensional stability is a requirement. This is also the reason for the main microwave tower to have star guying. With star guying, the tendency for any twisting during high winds is reduced to a minimum.

23CM

For 23 cm. We have been using a pair of air cooled 2C39 type triodes in a cavity amplifier. This arrangement delivers about 140 watts with almost no thermal drift in normal usage. Water cooling was not employed as the unheated nature of the shack made running water cooling quite a hassle during the cooler months. The maintenance time required to avoid freezing seemed to be prohibitive. Water cooling those 2C39s would have allowed up to 250 watts or so. Lately I decided that water cooling was needed for an effective 1296 station, and work was started on a 300 watt GS15B tetrode cavity amplifier. It is going to be used in 2007. The cavity was machined by KD5FZX. I added a G3SEK tetrode board kit with stiff screen regulation, and a negative screen current opto isolator protection circuit as outlined by K5JL. The GS15B can provide about 300 watts and no objectionable thermal drift with a good tube. The net result is a 3dB improvement on 1296 transmitting.

The 1296 antenna system of four X 45 element loop yagis, is located almost at the top of the looper stack on the 3" mast at about 133 feet asl. Feedline is 1 5/8" heliax with flexible sections of 1/2" Superflex above the rotor. The preamp housing is located near the 4 bay 45 element loop yagi array, and the receive downlead is 7/8" heliax into the shack. All the microwave bands are sequenced with DEMI sequencers. In addition, with four bands all stacked on top of each other, it is imperative that testing be done to insure that there is adequate isolation between bands so that RF energy from one band does not bully its' way into another! The antennas are arranged with 903 on the bottom, with 2304 just above that, followed by 1296, and then 3456 at the top. This was done to maximize frequency separation between adjacent arrays. The isolation was measured and in every case is below -60 dB, the limit of the equipment I used for the tests. Each band was swept so that any anomolies could be viewed quickly. 60 dB should be enough to prevent preamp damage between bands, and we have seen no problems in several years of use with this arrangement. To perform this test, we would sweep between 900 and 3500 MHz, and connect each antenna array together on the input and output ports and look at the amount of isolation between bands.

I have seen many puzzled comments on VHF reflectors about why preamp "XYZ" keeps blowing out. This seems to be a rather common problem with single operator home stations and rover stations. I doubt that many people realize how much power from another band can be forced down the feedline of the band you are trying to use. This problem is enhanced on the harmonically related bands as 144 and 432, or 432 and 1296. The third harmonic may be at such levels as to cause preamp burnout. The problem is exascerbated when spacings between antennas is small, or antennas are located on separate towers or masts at the same height. Before you go and install your super low noise preamp, it would be a good idea to check power levels from the other bands. Best way is to use a spectrum analyzer. The next best would be a microwave power meter and suitable attenuators. On the VHF bands, it is not uncommon to see a few hundred milliwatts back feeding a transverter or preamp. Unless the circuit has extensive filtering, damage will result.

13CM

The 2304 system is very similar in makeup, except for the power amplifier. We are currently using a 110 watt output solid state power amplifier located in the shack. We figure about 70 watts getting to the antenna. One of the projects this Summer is to revamp the tower mounted relay box and eliminate the SMA relay in anticipation of QRO on this band. We have another solid state amplifier waiting in the wings, that is capable of supplying enough power to melt most coax cable! Power for this amplifier will be supplied by a pair of 110 amp 28 vdc switching power supplies with 208 vac 3 phase input connections. We will most likely run the 2304 PA at the 350 to 400 watt level. It is capable of considerably more output! The 2304 antenna system is an array of four 76 element Blowtorch" loop yagis located just above the 903 array at about 127 ft up on the mast. The PHEMT preamp is located at the antenna to cut down on losses on the receive side. We figure we will have about 280 watts minimum at the antenna. In the past, we had about 70 watts at the antenna, so the 2.3 GHz signal should be much better at almost 6 dB over the old system.

9 CM

3456 MHz is a bit different in that the difficulty in routing RF power up the tower becomes a big problem. Large coax does not work at 3456. Use of 7/8" heliax results in too much loss. In the past, I had mounted the entire transverter and preamp up on the tower in a wx proof box. The box held a DEMI transverter along with a Pyro-Joe amplifier running 42 watts. This all worked quite well, but if anything ever happened to the transverter, I would be off the air for a long time. Due to the location of that wx proof box, at 135 ft above ground on a 3" mast, 15 feet above the tower top, any repair would be very difficult.

K1WHS wrestling with 3456 array at 135 ft on a 3" mast. N2CEI is assisting at the 120 ft level.

In 2006 I split the equipment up in the wx proof box. I left the 42 watt PA in the box, but removed the transverter to the shack. We decided to run 3456 TX rf up the tower on the old microwave transverter IF line. This cable used here was 3/8" LDF Heliax, which worked well as a 144 MHz IF line, but at 3456, the losses were quite high. A quick calculation produced about a 13 dB loss in the 3/8" Heliax plus the 1/2" Superflex cable above the rotor that ran to the box! We adjusted the power levels to accomodate the extra loss, and added a second stage preamp up on top so that the loss would not hurt our receive S/N ratio. Testing the entire system was done to ensure that we really were getting all of the power output that we thought we were. It was also important to verify that the receive noise figure was still good with the additional feedline loss as well. The latest 3456 setup has the transverter located in the shack in a 3" rackpanel, with a relative power output meter, and TX LEDs to indicate when the system is actually making power on top of the tower!

So all of the bottom four microwave bands, 903 thru 3456, are all co located on the same 3" mast and all antennas are pointing at the exact same heading. This makes for a very easy way to run stations up the bands. The 903 MHz antenna has a 7 degree beamwidth, and when a station is peaked on that band, then it will be peaked perfectly on the higher bands. This is very important for success when high gain antennas are being employed. 1296 has an 8 degree beamwidth. 2304 Mhz is 5.5 degrees. 3456 is the sharpest array with a 4 degree beamwidth. The rotator is a Prosistel PST-71 rotor. This has a slow turning rate and 1 degree readout accuracy. Such sharp antenna beamwidths would be impossible to use adequately with a lesser rotor. The PST-71 makes aiming a snap! We always try to get the six digit gridsquare from the other station, so that we can aim properly. There are some strange aiming effects that crop up on occasion. Some stations seem to wander around. In some cases this could be a simple aiming error, but other variations seen have been so great, that we are convinced that the variations are real. One location in New York State is consistently off by about 12 degrees. The effect seems to be more pronounced on 10 GHz, where signals from the same station have been known to vary by almost 3 degrees at times. I have seen no information about skewed tropo propagation paths, but I am convinced that it does happen. Just as arrival angles can change on the microwaves, I believe that multiple paths due to reflections and refraction, can cause the azimuth to vary slightly, as well!

The full stack in 2005.

5 GHz 10 GHz and 24 GHz

Not completed More to come, with pictures

\