Berning EA-2101 power amplifier
Murray Zeligman introduced me to Dave Berning's creations. I remember driving out to Dave's house in Potomac one evening to look over an EA2-150 amp (which I later purchased). There was a cavernous basement filled with gear and a workbench against the wall. When Dave turned around to greet me, my jaw almost hit the floor: the guy looked liked he'd just finished high school. The "kid," however, was no technical juvenile. His circuitry knowledge was astounding, and the quality of workmanship was fully professional—far beyond the garage tinkering level. The EA2-150 was a copy of nothing; it represented an original and sophisticated design statement. Features such as variable feedback control and digital automatic bias adjustment were well ahead of their time.
Dave symbolizes for me the quintessential high-end designer: someone with a glint in his eye who pushes the art forward because of the desire to uncover the truth. Of course, the guy wants to make money, but this is secondary to the creative process. The art is not driven by business. You wouldn't go up to Van Gogh and say: paint me a $299 or a $499 painting. That's low-fi. Instead, allow the artist to express himself freely with a complete palette on a large canvas. If the result costs five kilobucks, so be it. No one has ever sold a ton of $5k amps, but that wasn't the point in the first place.
After experimenting with hybrid circuits, Berning settled down to using nothing but tubes; the EA-2101 represents his latest thinking on amplifier design (footnote 1).
One of the chores of the review process is the physical labor of moving the equipment, setting up and tearing down. Grabbing the EA-2101 by its cute little handles and hoisting it about didn't prove much of a problem. Weighing 40 lbs, it's not heavy for a tube amp. Even my Air Tight ATM-2, nominally a 50Wpc stereo amp, weighs more. Generally, you can gauge the quality of a tube amp fairly well by its weight:power ratio—the higher, the better. All else being equal, the more massive the power transformer is, the better the power-supply regulation. Similarly, the core size and weight of the output transformer equate with deeper and cleaner bass response. So on the basis of the heft test, the Berning did not look promising. I wondered out loud how on earth Dave had managed to pack two 100W channels of transformer-coupled muscle into such a lightweight package. The answer's a bit long, but the key involves the output stage used.
"Six times a week, and twice on Sundays, for the last 30 years, someone, somewhere, has introduced a new type of valve." Thus quipped a pundit in 1951 in the pages of Wireless World. Of course, he was referring primarily to pentodes. Since the pentode was invented in 1930, almost everyone lost interest in power triodes. Power pentodes and beam power tubes ruled the day for audio amplification. The last domestic power triode, the 6AS7, was trotted out by RCA around 1948. But with the introduction of the EL34 and the KT66, no "sensible" designer bothered any longer with power triodes.
The last big push in power-tube research and development was motivated by vacuum-tube television. Rugged tubes like the 6550 and 6LF6 were churned out by GE and Sylvania by the millions. Although the pentode carried the day, triode lovers who craved the purity and musicality of triode sound did not give up. They seized the opportunity, first in the '40s, of effectively using the pentode as a triode. It should be realized that, to function as a triode, the screen grid voltage must not be held constant in relation to the cathode. Thus, by simply tying the screen grid to the plate of a power pentode, its screening action vanishes as if in smoke. The result is triode behavior with increased plate dissipation which typically surpasses that of most true filamentary triodes. Through the expedient of "castrating" the pentode, triodephiles hoped to make the pentode sing sweetly. A compromise strategy is used in the well-known "Ultra-linear" connection, where, by tying the screen grid to a point on the transformer primary intermediate between B+ and the plate, tube behavior can be made to resemble that of a triode.
Berning, committed to triode sound, followed this path of using castrated pentodes—but with a novel twist. There are two pairs of 6JN6 power pentodes (you guessed it: TV tubes) in his output stage, operating in a push-pull mode, with each pair of 6JN6s tied in parallel. The suppressor, control grid, and cathode of each pair of pentodes are tied together and grounded through a 15 ohm resistor. The drive signal is applied to the screen grids.
Say what!? Berning castrated the pentodes alright, but by cutting off the control grid. The result is a triode with the "screen grid" as the controlling element. Interesting things happen when such screen drive is used, the most intriguing aspect of which is the possibility of linear triode behavior at extremely low idle currents. The quiescent current of each 6JN6 in this circuit is an astonishingly low 3.5mA! Compare that to the bias current of an EL34, typically 30 to 50mA. That's an order-of-magnitude reduction in quiescent current. The most obvious benefits are reduced plate dissipation (ie, lower operating temperature) and increased tube life. Berning rates the lifetime of these tubes at around 10,000 hours, which translates to an operating life of 10 to 20 years for most people. And when it's finally time to retube the amp, expect a modest cost of around $250 for a full tube complement.
Another benefit is the reduction of core saturation in the output transformer. This is due to unbalanced operation of the push-pull output stage and, I suspect, is a bigger problem than most people realize. Perfect push-pull operation, supposed to eliminate even-order harmonic distortion, is about as rare as an honest politician. When was the last time you saw a push-pull tube amp with no second-harmonic distortion? I certainly have not. And when the output stage is out of balance, DC current flows through the primary of the transformer, which pushes its core closer to saturation. The less the DC current, the smaller the problem. Thus, with screen drive and its attendant minimal idle currents, a smaller output transformer may be used without fear of core saturation. The EA-2101 features one of the lightest and smallest output transformers you're likely to see in a 100W amp—the main reason for the Berning's lack of pounds.
Footnote 1: David outlined some of his amplifier design philosophy in his interview with Ken Kessler, Vol.11 No.7, July 1988.—John Atkinson