Created 6 June 2002
OK, so it's fairly well documented now how to build an AO-40 station. Here's how you can do it from scratch. Is your old AO-10/AO-13 Mode-B station gathering dust? Then all you need is an old TV satellite dish, a homebrew feed, and a $95 downconverter. A couple of hours' work.
But... now you've built it, how do you use it?
One of the things which afflicts us all when trying out something new is the fear of the unknown. The AO-40 satellite has its lowest currently operating downlink in the 2.4GHz band (or 13cm band, or S band). That in itself has led to many operators abandoning all hope of even trying to operate AO-40. The purpose of this page is to try to help out.
However easy some may say it is, there is a big difference between operating on the lower bands compared to the microwave bands. That difference is that unlike the lower bands, there's almost nothing to listen to in a very wide band to know that your equipment is at least working.
You can't just swish the tuning knob and expect to hear something unless you are also someone who wins the lottery jackpot on a regular basis. Add to that the fact that even if there was someone to listen to, antennas in the microwaves are normally relatively very directional, so you also need to know where on earth (or in space) to point the antenna.
And, of course, without any prior experience of operating, what have you got to compare with to know if your system's working well or not?
Yes, there are terrestrial beacons on the 13cm band available for the lucky ones. But even if you are in range of a terrestrial beacon, most downconverters for AO-40 don't cover down to the terrestrial sub-band where those beacons are.
Here's a typical AO-40 receive system...
I've separated the different areas of the system. Hopefully we know that our 144MHz Rx is working OK. The rest is what I call the 'unknowns'. With some tips below we'll be able to check out some of the components individually to reduce the uncertainty.
I recommend that you concentrate on resolving the component parts in the order provided, ensuring each part passes its test before continuing the next.
To begin with, prepare all your equipment to be together in a semi-portable arrangement. If you've already put your dish and feed up on the tower life will be tougher when you're trying to diagnose faults.
Even disconnected from the antenna, but with the power on, as you plug the downconverter into your SSB receiver you should hear quite a big increase in noise level. That's due to the huge gain these units have. More than enough for our purposes, so some folks use attenuators between the downconverter and the receiver.
Base S-meter readings are typically S5 to S7 or so. If you don't hear a large increase in noise as you plug in the downconverter, something's wrong.
If you're feeding power up the coax, check that there really is power at the downconverter end.
Check the downconverter's drawing a reasonable amount of current (depending on model, 100-200mA).
Model Current SSB Electronics UEK-3000 164mA DB6NT MKU24-OSCAR 150mA (cold), 107mA (warm) AIDC 3731 (modified by K5GNA) 137mA AIDC 3733 (modified by G6LVB as per K5GNA instructions) 147mA Norsat 2500 173mA
If still nothing, or it's not a big noise increase, check your cables.
Above all, don't transmit into your d/c!
OK, if you only ever purchase one piece of test equipment for 13cm, get yourself a signal source. They're available for about $20. Make life easy on yourself.
AMSAT-UK provides a complete kit of parts for David G0MRF's signal source.
Bob K5GNA (see also pricelist) does an ex-MMDS Gardiner unit which can be used without modification for a weak source at 2397.840MHz (receive at 141.840MHz). With an hour to spare you can modify it to 2400MHz with a new crystal and also modify it to make it much much stronger. In the interests of knowing if your system will pick up weaker signals, it might be wiser to just change the crystal! Fully modified, I can hear this signal source half a mile away with omni antennas.
Also check out Jerry K5OE's page on ideas for signal sources.
As an antenna, stick a short piece of wire (3 or 4cm) into the centre of the antenna socket of your downconverter. Don't plug the signal source into your downconverter or you may blow the front end.
Place the signal source a foot or so away. Now find the signal. This should be VERY strong. Indeed, you may find spurs before you find the real signal.
With the modified Gardiner unit, there's nowhere in my apartment I can hide it (even in an Aluminium case) without it being loud at the downconverter.
If you don't have a signal source, you will have to test the downconverter and feed together.
The frequency of the downconverter is about the only 'unknown' which can't be answered precisely easily, but at least I can tell you what to expect.
If you know something about the innards of the downconverter, and you already have an receiver which can tune to the downconverter's crsytal oscillator (or a harmonic) and is accurate enough, you may be able to confirm and/or tune the local oscillator to a certain accuracy. But, remember that not only is the crystal oscillator multiplied up potentially hundreds of times to the LO frequency, so are the inaccuracies.
For example, for a 144MHz IF, the LO will be 2400-144 = 2256MHz. For a modified AIDC 3733 with 144MHz, this 2256MHz is derived from an 8.8125MHz crystal source (multiplied by 256). Tune your SSB HF receiver with a short receive antenna close to the downconverter and you'll be able to pick up the crystal oscillator.
In this example, if the crystal oscillator is only 100Hz off frequency, this will be over 25kHz off at 2256MHz!
So don't be surprised about a +/-20kHz or even 30kHz inaccuracy especially if you have installed a new crystal as a modification.
Bear in mind also that drift will occur when switching the downconverter on, often 10kHz or so over the first ten or fifteen minutes of operation as the crystal oscillator warms up.
Feed (+ Downconverter)
To test the feed, you're going to need to hook it up to your downconverter.
Again, the signal source will be especially useful here
Generally you're looking for some improvement and directionality over the piece of wire you used for a test antenna when checking out the downconverter. If your signal source is very strong, it will be advantageous to check the comparative signal strengths and directionality outdoors.
There's also a test you can do without the signal source. This will test the sensitivity of your feed and downconverter combination. Set up the equipment outside. If your SSB receiver allows you to switch off the AGC, do so. Pointing the feed to the sky and then down to the earth, you should be able to hear a very small increase in noise if your feed and downconverter are pretty sensitive. If you can hear this difference, you are doing well so far.
Dish (+ Feed + Downconverter)
There's not much to go wrong with the dish itself, other than being very very badly bent. Things are surprisingly forgiving at 13cm. The larger the dish the more difficult it will be to point and find the bird, as the beamwidth is smaller. But, with a smaller dish, the quieter the beacon will be. 60cm diameter is about the comfortable minimum.
You should have checked that your feed and downconverter are functioning before attempting this stage.
Make sure your feed is in about the right place. +/-1" is probably of little consequence in the big scheme of things in practice. You can optimise it later.
The correct place to put the feed depends on the feed, the type of dish and its dimensions. Essentially the phase center of the feed (where it emits the maximum amount of radiation) should be placed at the dish's focal point. For a helix at 13cm this is about an inch or so in front of the feed's reflector.
If you have a signal source, this will be invaluable in testing out the system.
Take the equipment outdoors and check the directionality and gain of the system using the signal source. Again, as signal sources can be very strong, you might want to place it a hundred yards away or so from your dish.
If you have an offset fed dish, they are generally about 21 degrees off true vertical. So, if the feed arm is at the bottom, and the signal source is at the same level as the dish, the dish will point 21 degrees down towards the earth.
You don't need a signal source for the next two tests. A test of sensitivity is to check for ground noise by pointing the dish up to the sky and down to the earth. Again, switch off your SSB receiver's AGC if you can. There should be quite a marked increase in noise.
A further test is to point the dish at the sun. A small increase in noise should be detectable. If this is the case, your receiving system is in very good shape so far.
Prediction, in fact, is only an unknown if either you're not too familiar with operating some of the esoteric parts of your software, or are unfamiliar with some of the Phase-3 satellite specific terms.
If you've operated the LEO's, many of these concepts will be second nature, but some may not be.
Some computer prediction software will automatically update the PC's clock online, or on a modem dial up. There's also a number of utilities to do this online. Alternatively you'll have to do it manually. As AO-40 is slower moving, the time is not as critically accurate as for LEO's.
Depending on the software you're using, this can be fiddly to get right, especially when you switch time zone when travelling, or as you switch in and out of daylight saving time. Some newer Windows software is pretty good at compensating for this automatically.
Check that the observer position in your prediction software is correct. You may need to enter this more than once in some prediction software depending on the mode of operation (eg, real-time or listing modes).
Although AO-40's orbit isn't going to be changing much, it's good practice to update your Keps fairly regularly, say every couple of weeks or every month. Keps (or Keplarian Elements) describe the orbit of the spacecraft.
Even though the bird may be visible from your location, check with the AMSAT-DL AO-40 news page that the Middle Beacon (or MB) on 2401.323MHz will be switched on for the given MA (Mean Anomaly, or simply the position in the orbit).
Be aware that the MA on the WiSP software wasn't working correctly at the time of writing.
The concept of MA isn't often used in amateur terms for LEO's.
Squint is simply a measure of how well the antennas are pointing your location. Aim for trying to listen to the bird at squints under 15 degrees to begin with. A squint angle of zero degrees means that the antennas are pointing directly at you.
To get a squint measurement, you'll have to get the current AO-40 ALON & ALAT figures published and updated regularly here. ALON & ALAT don't get updated when you update your Keps. Check that you enter these parameters correctly: some software annoyingly asks for the data in reverse to the way it's published.
Also, some software will not accept ALON figures above 360. If this is the case, simply subtract 360 from the value.
As the ALON/ALAT nomenclature is somewhat esoteric, many packages are not very intuitive about where or how you enter the ALON/ALAT data.
Another gotcha is that some older software doesn't cope with the fact that AO-40 has two sides, one for the high gain antennas and the other side for the omni directional antennas. As the antennas in use on AO-40 are on the high gain 'reverse' side, squint indications may be 180 degrees out with software which doesn't understand the two sides. Check for later versions of your software if you can't find out how to set this.
The ALON/ALAT and squint concepts are difficult to grasp at first. ALON/ALAT is also quite fiddly to enter into some prediction software.
Above all, Read The Manual for your software on how to enter the ALON/ALAT data.
A number of computer prediction packages incorporate, either directly of indirectly, an indication of the Doppler shift proportional to the relative speed of the satellite to yourself and the frequency under scrutiny. This will help greatly in narrowing down the frequency range you sweep around when searching for the beacon. At 2.4GHz on AO-40 the observed Doppler range can be as much as +/- 35kHz or so, so narrowing this down will help a good deal in finding the beacon.
AO-40's Middle Beacon (or MB) is the normal beacon and is nominally at 2401.323MHz.
As an alternative to using the computer, you can use my low tech approach to Doppler which I find quite handy, particularly outdoors.
Using your prediction software and the AO-40 schedule, choose a time of a pass when...
- The beacon is on
- The squint is under 15 degrees
- There will be no local obstructions to the satellite - including trees
One thing the computer revolution has taught me is that the concept of a paperless world is as far away today as it's ever been. On the big day before you're ready to try, take the opportunity to print out a listing of the predicted Az/El, MA, squint and, if your software will do it, the Doppler corrected beacon frequency.
I usually run the listing with predictions at about five or ten minute intervals.
The listing will help you plan where to put the equipment so there's no obstructions, as well as providing easy immediate access to all the data you need without having to fight with a computer program you might not be too familiar with.
One way of checking out your prediction software is to try it on a satellite which is very easy to hear such as UO-14. But this won't test your MA & squint angle prediction, or your Keps for AO-40.
The Heavens Above website can be used to check your AO-40 predictions are OK, but be careful. At the time of writing, predictions from the site began and ended when the bird is 10 degrees over the horizon. Heavens above won't confirm your MA or squint calculations.
Other than checking and double checking the settings of the ALON/ALAT for squint angle calculations, the only way you can confirm your squint predictions are close is to check with another local station who's already QRV on AO-40.
The Big Day
OK, so you know you have a tip-top receiving system and you're ready to try it out for real.
In addition to your pass listing, get your magnetic compass, protractor and plumbline out. Depending on your location, sometimes your compass may be way out (check out your Magnetic Declination). Most of us are lucky enough to have a magnetic declination which is negligible compared to our dish's beamwidth.
If you can get hold of an inclinometer too, even better.
Remember offset fed dishes need to be pointed about 21 degrees down from true.
With your prediction listing, point your dish as accurately as you can. As a guide, the 3dB beamwidth of a 60cm dish is about 16 degrees.
You'll need to be prepared to tune your SSB receiver around the Doppler corrected beacon frequency perhaps +/- 30kHz or so until you hear it because, as yet, you don't know the accuracy of your downconverter.
So, let's say the Doppler predicted frequency is 2401.320, you should be tuning your 144MHz SSB receiver from 145.290 to 145.350MHz.
By the way, have you noticed that the only 'unknown' at this point (I hope) is the exact downconverter frequency to within 60kHz? How many 'unknowns' did you start with?
Although this waving about method is not to be recommended for beginners, I've provided this video (2.5Mb) to give you an idea of what to expect with regards to accuracy of pointing and what it sounds like. Far better to be methodical about it. The video is of a G3RUH dish + patch feed, AIDC 3731 downconverter and an FT-817 for the 144MHz receiver. It was taken at 30 degrees of squint at a distance of 38,000km in my basement garden, so hardly optimum conditions.
It took me a several attempts to even hear a thing. But then I didn't have the benefit of hindsight. I hope all of that hindsight is written in this document...
Mail Howard, G6LVB