Cheap Microwave Gear?

In my last blog post, I mentioned my conversation with Andy W5ACM at a Houston QRP Society meeting where he challenged me to make contacts through OSCAR-51 before criticizing it. That conversation has borne more fruit by sparking another idea, and I thought I'd write it down before I went on to
work on other things.

That conversation, now nearly three months past, was a general condemnation of what I saw as the direction of AMSAT in the late 1980's and early 1990's. Back then, the thrust seemed, to me, to be toward satellites with high (40,000km+) apogees and S-Band (2400 MHz) equipment. Back then, I saw that as suicidal because S-Band transceivers were not available to the radio amateur at any price and one would pay several times the cost of a VHF transceiver for an S-band transverter.

Since that time, some radios have had support for L-Band (1300 MHz) as an option, but I still know of no options for S-Band except upconverters and downconverters which allow the use of a VHF or UHF radio on the S-Band.

Anyway, since I was trying to make contacts through satellites, I thought it would be a good idea to subscribe to AMSAT's "amsat-bb" mailing list. The amount of noise in that list's signal is not bad for a general discussion mailing list, and it had an announcement a while back that I found interesting.

You see, it turns out that a guy named Paul Wade, W1GHZ has designed some transverters, and announcement was made on amsat-bb for a group putting together a group buy of parts to kit out transverters based on his design. I briefly considered the ordering some of the kits which, sadly, I decided I couldn't afford right now, but the important thing to understand about these kits is that you can build an L-Band or S-Band transverter based upon them for around $100 each, plus whatever you consider a couple of hours of your time to be worth, and for a hobby activity I consider my time to be worth a negative amount.

Another reason that I decided to pass on one of these kits is because of how they work and what, precisely, they do. You see, they are cheap because they use inexpensive non-microwave components and printed circuit boards and deal with the losses that those choices imply by inserting cheap and readily-available MMIC (Monolithic Microwave Integrated Circuit) amplifiers. Local-oscillators suitable for use in microwave transverters have proven to be a problem in the past, but if you use a frequency that is a multiple of a frequency that is commonly used in desktop computerers, that's not a problem any more.

What is a problem is the fact that the frequencies of the oscillators and such were chosen to translate the output of a 2-meter (144-148 MHz) transmitter into the small signal part of the L-Band and S-Band and, well, I'm interested in signals in the satellite part, not the small signal part. To use the L-Band converter to work AO-51, I'd need to drive it with an FM transmitter on 116.7 MHz. Of course, I don't have a transmitter that emits anything on 116.7 MHz, let alone FM. (It's in the aircraft band, and airplane radios are AM.)

So, what to do? I could modify my IC-706MkIIG such that it isn't restricted to just transmitting within the ham bands, but I'm not really that enthusiastic about opening up my main radio and taking a soldering iron to its innards. I could also find a different oscillator frequency that would just happen to be cheap, common, and which had a harmonic that feel into the range of about 1120 to 1125 MHz, but I gave that up after about an hour. Besides, it occurred to me, the venerable Icom wasn't necessarily the appropriate equipment to drive a transverter. Transverters don't need a lot of drive and typically can't deal with very much driver, and so I'd have to do something to ensure that I didn't accidentally blow up whatever transverter I got.

I then observed that this wasn't a problem on receive because lots of radios have general-coverage receivers so it doesn't matter if a downconverter puts the signals I'm interested in at some frequency I wasn't allowed to transmit on. What I needed, I thought, was a special purpose exciter that could generate FM signals (or, really, any mode) on any frequency from 100 MHz to 150 MHz but which wouldn't generate too much power for the upconverter to handle. I thought that would be an appropriate companion device for those really cool transverters.

With that in mind, I set out to see if anyone has done this already, and I couldn't find where anyone had. However, what I did find was that all the hard parts are done. A search for "exciter transverter" led me to a 1993 QST article by Rick Campbell KK7B "A Multimode Phasing Exciter for 1 to 500 MHz". To be honest, I don't think that it's really an exciter. I think that it's really just a modulator that needs a carrier signal (two of them, 90 degrees out of phase, to be specific) to modulate.

So, where to get the carrier signal? It would be best if the frequency were synthesized and what I know about PLL-based frequency synthesizers suggests that they really aren't all that easy to work with. Then, I remembered hearing about a technology called DDS, for "Direct Digital Synthesis", which is a radically different technology for synthesizing a signal than using a PLL, but had the limitation of requiring that the signal be lower in frequency than a reference, and so was limited to relatively low frequencies.

When I searched for "direct digital synthesis arrl" (because "direct digital synthesis" isn't specific enough and the other article was found at arrl.org) the first link was to an article in the current (or next--it isn't clear) issue of QEX magazine about a device based upon a thing called the "NimbleSig III" DDS synthesizer, designed by Thomas Alldread VA7TA.

So, I searched for "NimbleSig III" and found an amazing device that can generate two signals each with a specified independent frequency and phase shift at anywhere from 100 kHz to 200 MHz. Um, wow. This is just what the doctor ordered and why the preacher danced. Combined with KK7B's device, I can use it to generate AM, FM, SSB, CW and anything else that can be derived from those over the entire desired 100 MHz to 150 MHz frequency range. Of course, these things aren't available assembled, or as kits, but as bare printed circuit boards, and they use all surface-mount parts. I've never done surface mount, but this is a powerful inducement for me to try.

The thing is, although I haven't tried to price the parts just yet, I can probably build each one of these devices for around $100. So, I could build an S-Band or L-Band transmitter for less than $300 and get the downconverter I'd need, so I can receive, too, for free. I wonder if anybody's done this?