---   The Modulator   ---

The modulator is an essential part of an old AM transmitter. By old, I mean an AM transmitter that emits a carrier and two, yes two, sidebands. These new-fangled SSB transmitters (they'll probably never catch on) emit darn little carrier and only a little more than half their sidebands. I don't see how anyone can enjoy making a contact without any carrier and only half their sidebands, so to me it all seems just like a waste of good resistors and condensers. Besides, why won't anybody admit that this SSB stuff is really AM anyway - just a different kind of AM!

BigRig is a plate-modulated transmitter. Plate modulation was generally favored over several other methods of generating AM despite its unfavorable economics, because plate-modulated rigs unarguably produced the best signals. Plate modulation wasn't all that hard. You just had to amplify the microphone's audio output up to a level needed for the transmitter. BigRig needs about 130 Db. of voltage gain and about 146 Db. of power gain, between mike and final amplifier. The plate-modulated Class C amplifier, for adequate modulation (i.e., 100% modulation) required a modulating signal placed in series with the high voltage for the final amplifier - a modulating signal of a power equal to one half of the DC power input to the final amplifier. In this case, about 400 watts That's 400 real r.m.s. watts, sir, not those phony "two-channel peak music power" pseudo-watts touted by over-reaching hi-fi advertisers.

BigRig, like practically all plate-modulated rigs of its day, used a Class B amplifier for its modulator. Class B amps were favored because of their electrical efficiency and circuit simplicity. The transmitter was first designed to use a pair of Taylor TZ-40 triode tubes in the Class B stage. They were popular because they were specifically designed for Class B modulators, and moreover, designed to operate at zero grid-bias (hence, the "Z" in their model designation). Zero-bias operation was welcomed by hams because grid-bias was generally obtained from a battery, which punished its user by its high cost and relatively short life. In use, if the operator were not alert, the gradually dwindling voltage of tired bias batteries caused lots of low-level audio crossover distortion in the Class B stage, and resultant on-the-air criticism. And too, the modulator's gradually-climbing idling plate current often climbed right into that bright-orange-anode area that immediately preceded the need for a new set of tubes.

Eventually, Taylor went out of business and TZ-40s were not to be found. A suitable, and perhaps even better tube, was the then-new 811A. At plate voltages of 1250 volts or less, a pair of Class B 811s ran happily at zero bias. However, BigRig needed every last watt of the 375 watts a pair of 811's can officially produce, and perhaps even a few more. To get all them there watts, one needed lots of plate volts. BigRig's modulator power supply produced about 1600 volts, and at that voltage, the 811s demand a little bias. A 4.5 volt "C" battery (no, not 3 1.5 volt C cells) did the job, but such things don't exist today.

I've tried to resist the many temptations to "improve" BigRig by incorporating some modern circuitry techniques and parts. I often recall with a little giggle, the now-defunct Telrex company, once the makers of the world's finest beam antennas. Telrex had a notation on many of its blueprints that read "Specifications are guaranteed if you make no new inventions." Although, it's a bit of a struggle, I'm trying not to make many too many new inventions in BigRig. The engineer in me sees scores of ways to improve it, and there's many an opportunity to beneficially incorporate parts that didn't even exist in 1945 when the rig was built. And some of those 1930s and 1940s parts are just darn hard to find! And, the son in me winces at discovering less-than-wonderful implementations that were made by Dad, excusing them with the rationale that economics, not lack of skill, forced some junk-box design solutions and some not-quite-the-right-tool construction methodologies.

Here, though, was a bit of a quandry. No real C batteries for the bias supply. I mulled over several battery options for the needed 4.5 volts, including a bulky 6V lantern battery with a two-diode series drop, but instead, hoping nobody will notice, I've planned on supplying the modulator bias from a little supply comprising a wall transformer and a Zener diode regulator. Please don't tell. It's interesting in a way, though. The 1930's and 1940's literature on Class B amplifiers strenuously reinforce that the bias had to derive from a "low impedance" power supply. The bias had to remain stable even though often-substantial grid current flowed backwards through the bias supply throughout nearly all 360 degrees of the audio input cycle, and tended to upset the bias voltage. This "low impedance" requirement needed some thought with the techniques and parts of 1945, but it's trivial with today's gizmos like 3-terminal regulators and big Zener diodes. A bias supply of a half dozen cheap parts is so economically reasonable, so much an easily-implemented technical no-brainer, and has such an extremely low internal impedance, that an old-time Class B modulator designer would likely cut off his right arm for it. Now, the big question is, should I use Fahnestock clips to attach the bias leads? After all, Burgess C batteries did!

Well, today is April 4. The modulator metalwork has been liberated from the powder-coating man, and has been re-assembled. A mix of old parts and new. Mostly old techniques. A couple of new pictures show it's status after re-assembly but before the wiring has commenced. It'll probably be finished in a week or two.

One more sidelight. The modulator is not the heaviest of BigRig's eight decks, but it does weigh in at a very healthy 55+ lbs. It's a bit much to wrestle around on the workbench, but for some reason I don't seem to remember that being a problem when I last worked on the rig 45 years ago.

Click   here   to return to the home page.

Click  here   to return to the Rag Chew links.