Wow! Some pretty robust debate! A few points to address it if I may:
1. Yep, it's a Magnavox. The effort was hardly to try and make a Picasso out of it, or imply that it became one. Rather, with so many (if not virtually) all of these units now being used outside of their original habitat, the effort was primarily aimed at making it be the best it could be for what it is, within its newly re-purposed use application and operating environment. Since so much of that changed when it morphed into its new service duty, the effort then was one of helping the basic design adapt to its new surroundings and responsibilities. For me personally, I was simply trying to fulfill a request from fellow AKer Kidmoe regarding the wide spread love for the 9300 series of amplifiers, honor my end of the deal we struck, and produce an end goal wherein we both helped the AK community in the process. In the end however, its still a Magnavox, with the same tubes, chassis, transformer set, and circuit topology. And there's a reason for this: I have always felt it is important that when a device is modified for improved performance, that as much of the original essence of that device is maintained. That was a guiding principle with this project, as it has been with all the modification projects I've offered to the AK family, Magnavox, or otherwise.
2. Operating with EFB, the modified amplifier draws only 90 mA under quiescent conditions (less than the original design), which improves even further on one of Magnavox's original design goals for the unit: extended tube life. At full power output (15.5 watts/ch) into 4 ohms with both channels being driven, this current level (basically) doubles to 185 mA . I have found that even the JJ GZ34 -- which has a dubious reputation at best -- does just fine at the voltage and current levels that most 6BQ5 stereo amplifiers operate at. Their failure record multiplies exponentially in higher voltage and current level applications.
The primary reasoning for changing to the 5AR4 tube was to eliminate the B+ surge at turn on as produced by the 5U4. With today's higher line voltages, this surge reaches about 425 volts. With the 5AR4, the surge is completely eliminated, while the tube's lower internal drop also acts to maintain the same B+ voltage produced after bucking, as is produced by the 5U4 without bucking (based on the current draw of the original design). For the majority of folks who operate these amplifiers into an 8 ohm load this this is important, as it allows the modified amplifier to produce at least as much power (actually slightly more) as the unmodified stock design does into that compromised loading arrangement, while maintaining sane heater voltage levels at the same time.
3. Unless there were different power transformer manufacturers, I would be surprised at those finding a reasonable heater voltage with 120 vac (or more) applied directly to their amplifier. In both this and the previous (8600) Magnavox project, heater voltages were found to be uniformly high, causing need to address it. Of course, the installation of in-rush current limiters will reduce the need for the bucking connection.
4. While need for the bucking connection centers on the heater voltage produced, the ultimate goal actually has little if anything to do with the life of the heater itself. After all, the heaters of modern vacuum tubes rarely fail, except in those tubes used in a series string that is powered directly from the AC line (i.e., the All-American Five radios). No, the real goal of the bucking connection is to produce the optimum cathode temperature within the tube: If the heater voltage is too high, it will act to accelerate the demise of the cathode, and also cause some of its coating to boil off and deposit itself onto the next nearest element: the control grid. This in turn promotes reverse grid current issues which can then cause a tube to runaway with itself when operated near maximum dissipation levels. In this project, the tubes are operated nowhere near maximum dissipation under normal use conditions, so reverse grid current issues are not a concern in the modified amplifier. However, even if those design issues aren't present, the damage (cathode material boil off) within the tube is still occurring in most cases if a correction is not made. Why subject your tubes to that environment if you don't halve to? The bucking connection then is a real tube life saver -- but not for the element it directly addresses.
The combination of the bucking connection and 5AR4 rectifier tube allows then for all elements of the tubes to operate at optimum temperatures, without any loss of performance in the process, while adding EFB to this mix then (among other things) extends tube life to well beyond that offered by the original design -- even when operating from its intended design center AC line voltage level.
Dave
