The first order of business was to take a look at the transmit spectrum of the TS-450's as it seemed as though at least one of them was more "hissy" than the others - that is, when it transmitted, everyone else heard a hiss - even if not on the same band.
Throwing the rig on my service monitor (which includes a spectrum analyzer) I could see that both of the TS-450's were pretty much created equally. What surprised me, however, was the fact that from the frequency of the radio's lowpass filter (on 20 meters, this was somewhere around 19 MHz) on down to below 1.8 MHz (where the response of the final amplifier itself dropped off) there was a fairly level noise floor - approximately -34 dBm in a 60 kHz detection bandwidth. When translated to an SSB bandwidth, this turns out to be approximately -50dBm - plenty high enough to be audible under many conditions even when the "path loss" (that is, from the transmit antenna to the other station's receive antenna) was taken into account - and the two radios don't even have to be on the same band!
The next measurement was with the built-in antenna tuner switched into the transmit path. Interestingly, even when the tuner "matched" to the service monitor - that is, "matched" to a 50 ohm resistive load - I could see that the magnitude of the off-frequency noise decreased markedly, implying that there was still some "L" and "C" in the signal path and doing some "matching." Typically, the amount of noise reduction on other bands was anywhere from 8 to 20 dB, depending on the transmit frequency and band, of course.
The upshot of this is that it's worth "eating" the additional loss of the antenna tuner (a bit less than 1dB) just to gain an extra 8-20dB of reduction in the "hiss" emitted by the transmitter!
One problem that this does not solve is that of in-band desense - that is, running two stations in the same band at the same time. Practically speaking, there are only a few ways that one can minimize this problem:
- Separation of antennas. This is fine, if you have the room! It's worth noting that Field Day rules actually place a limit on how large the site may be (that is, all radios and antennas must fit within a 1000 foot diameter circle) and still be counted as one station.
- Sensible arrangement of antennas. In North America, it's generally better to place antennas in a north-south line as one will, generally speaking, be sending signals east and/or west, putting the other antennas off the sides where the response will generally be reduce and the isolation between antennas be enhanced. Obviously, the nature of this sort of strategy varies a bit depending on your location!
- Selection of rigs. Many years ago, we banned Icom IC-706's (and all variations) from our field day site as it seemed as though they did a great job in radiation tons of noise and absolutely wilted in the presence of another transmitter - even one on a different band! While it would be nice to have a really fancy, high-dynamic range and ultra-clean rig and/or one of the old standbys that could tolerate multi-station environments (such as the venerable Drake 4 twins) we have what we have. Compared to many others we have tried, the Kenwoods aren't really that bad...
- Filtering. Keeping one's emissions confined to the band on which operation is taking place is fairly easily managed with a number of off-the-shelf "transmit-through" bandpass filters, but these do nothing to keep two stations on the same band from bothering each other! Years ago, we tried using 1/4 wave stubs, but these turned out to be a nightmare to manage - especially as the collective memory from year to year was imperfect and expertise on what worked or didn't worked was lost. In the case of 20 meters, last year I devised a fairly simple notch-bandpass approach that worked nicely to keep the SSB and CW stations out of each other's hair - but I'll talk about that in a later entry.
Next on the list was the old TS-820. In turning it on, I realized that my vague memory had been correct and that it did have the stock, 500 Hz CW filter in it. After cleaning it up, spraying a few pots and switches with cleaner to get rid of the "scratchies" and tweaking the frequency display into calibration, I set about modifying it for low-voltage, positive keying.
Being an older radio it has grid-block keying which meant that -60 to -90 volts appeared across the key, making it incompatible with many modern CW keying circuits that expected a low, positive voltage across the key. Constructing a simple circuit using 3 transistors and a bit of RFI protection, I wired it into the original key jack, tested it, and then added a label indicating that it had been modified - I'll post details on this modification in later entry...
I also tested the TS-820 to see what its transmit spectra looked like and was pleased to see what I'd expected: The noise output was lower than that of the TS-450's (around -50dBm as compared to -34dBm for the TS-450's) and it was confined to no more than +/- 10% of the operating frequency - a testament not only to the tunable bandpass filters driver chain but also the Pi network on the tube finals - both of which effectively limit the useful bandpass of the output amplifier signal path to those frequencies near those for which the radio was tuned.
Now, for the rest of the planning of Field Day. Stay tuned!