---   The Modulator Power Supply   ---

Glow Blue!   No, I'm not a U of M football fan. It was indeed "Glow Blue!" that was heard 'round the W8VR hamshack today when for the first time in nearly a half century, BigRig exercised a pair of its 866A mercury-vapor rectifiers. The blue glow was truly exciting to see, and missing only the speech-following blue pulsation of heavy AM modulation to make the image complete. That'll come soon though!

It was a bit of a relief to crank up a newly-built 2.5 Kv supply and not hear a single arc or spark. Nonetheless, several issues surrounding the modulator power supply come to mind as worth talking about. Here's a few:

    1.    As I mentioned in one of the "before" pictures, the modulator power supply seems to have been once re-wired by me, although I have absolutely no recollection of it. I was gabbing about that just today, and it occurred to me that the wiring techniques used were those that I learned after about 1953 when I began working as a technician at Armour Research Foundation in Chicago. After coming to Detroit, I transferred BigRig to friend W9RX (then W9EXE) in Chicago, along about 1961, so the rewiring occurred during that 8-year window. I find it strange that I remember nothing about it!

    2.    The metal work was the worst I've encountered so far, with respect to extra and unused holes, signs of parts having been replaced with different parts, parts not mounted straight, not mounted symmetrically, bent chassis, etc. The large rectanglar hole in the chassis, underneath the plate transformer, smacks of there once being a double-bell transformer mounted there. I don't remember hearing of a double-bell transformer with as large a VA rating as the present plate transformer, so perhaps the original transformer was once replaced to get more audio power. Speaking of the present plate transformer, I've learned enough about its electrical characteristics, but I've been quite unable to determine its origins. I've been unable to find it in any of my several old transformer catalogs, and I do have several. Where I've no catalog, I've emailed a few manufacturers and described the unit, but nobody will own up to having been the maker. Often, they too have no data remaining from their operations of over 50 years ago. I can faintly hear the refrain from within their engineering department walls during one of their cleaning sessions - "Throw out all those old drawings, we'll sure never need them again!" Well, the transformer seems in good working order, so I guess the issue is more a curiosity than a necessity. If, though, you come across any info on who, in the 1930s or 1940s made a plate transformer bearing model number PL-4, please do let me know. The unmarked primary is a single 110v or 115v winding, and the secondary is marked 1850-1525-0-1525-1850.

    3.    I discovered that the UTC filament transformer was by its markings, a 10 volt CT unit. Strange, because the 866 rectifiers have 2.5 volt filaments. I've seen 866 filaments in parallel, and a pair in series operating from 5 volts, but there's no reasonable way to run a pair of them from 10 volts. These 866s were clearly in parallel as easily seen from the "before" photo of the chassis underside. A little investigation revealed that the filament transformer had been re-wound for a 2.5 volt output! I was excited to discover that because it made me a little proud to recognize that my dad had been willing and able to undertake the rewinding of a transformer! I suppose that he may have acquired it otherwise, but it pleases me to think that he rewound it! Either way, the re-wound filament transformer supplies the correct voltage, adequate current, is properly center-tapped, and the insulation doesn't hiccup when the supply is putting out its unloaded 2450 volts.

    4.    Many have heard the famed Indianapolis Speedway being called the "Brickyard" because of the huge number of bricks that once paved the track. Also storied, is how when the track was re-surfaced, a few of the bricks were retained for purely ceremonial and nostalgia reasons. So it is with BigRig's modulator power supply. The filament wiring of the original power supply was formed from a couple of untinned square-copper-bus conductors. (No, not the kind of bus conductors that ask for your tickets.) The bus conductors here are easily seen in one of the "before" pictures. It occurred to me that just like the speedway, BigRig should retain a ceremonial bit of its original wiring. So, I saved the two peculiar pieces of square copper bus, and installed them into the rebuilt modulator power supply. They'll be the sole remnants of BigRig's original wiring in the restored BigRig, and they're easily seen in one of the "after" pictures.

    5.    Some may be curious about the brass tube visible in the underchassis "after" view. Let me explain. It was then, and still is, common practice to use the chassis as the common conductor for "ground." That practice often points to a lack of in-depth understanding of the meaning of "ground" and also often results in a myriad of bugs in the performance of radio gear. Yes, it may be convenient to do so, but the use of various points of the metal chassis for a bunch of power and signal returns, frequently causes difficulties through "common-impedance coupling." And, the higher the frequencies involved, the greater the frequency and magnitude of the headaches! Common impedance coupling can allow one voltage or current in a circuit to be varied by some other voltage or current in a different circuit. The most-often-described situation is strong 60 Hz. power currents wandering back and forth through a chassis because various points of the the chassis were used to connect circuit returns. If low-level signals like mike inputs, use the same chassis as a "ground," then the voltage drops in the chassis metal itself, can end up in series with the mike's low-level signals, and produce lots of hum. To avoid those pitfalls, one connects all "grounds" only to a single common point, and my brass bus, at low frequencies, is a reasonable approximation of such a commmon point. In the entire re-construction of BigRig, I have never allowed more than one point of a chassis to be used as a wiring point, so I expect that when I get the thing on the air some day, any hum will be only because I don't know the words to the lyrics of "There Ain't No Currents Through This Here Iron!" And, I still need somebody to compose the melody . . . .

    6.    I did make a couple of very minor changes to the modulator power supply. As I've probably said before, at least once (!), I've been struggling with my longing to change this, change that, improve that other thing, and on and on. I have to keep reminding myself that my task isn't to build the best transmitter that I can, but instead to re-build this one. Yet, there are some things that don't materially change BigRig, but are important for safety of transmitters and radio hams, so I did add a couple of fuses in the primary circuits, and to help in noise control, I added a few bypass caps on the power lines. Another change was to the power connector. One can see in the "before" picture, that the two 120-volt inputs were via a chassis-mounted female connector. Yes, but . . . that means the mating connector on the cable was male! Not too cool for operator safety, so I changed the gender of the connectors.

    7.    Lastly, I changed the HV wiring from the PVC-insulated wire that I found there, to the red 10 Kv wire that you see in the pictures. I don't know the characteristics of the original wiring, there were no markings on it. I remember no arcing or sparking, though, and there were no visible signs of breakdown. Yet, for HV circuits, it just didn't look like proper wire nohow, so away it went.

Well, that's the story of the modulator power supply. At least for now. This was the 5th deck of the 8 decks of BigRig, so I've only got 3 more to do. It's getting exciting to think that with any degree lof good luck, this thing might actually be on the air one day soon!

June, 2009. 73 de W8VR

Click   here   to return to the home page.

Click  here   to return to the Rag Chew links.