In Search of the Perfect 300B Tube

In the last two years, the available choices in 300B output tubes for one's low-wattage single-ended power amplifier have become an embarrassment of riches. If we include the souped-up versions—which, in fact, deviate from the original 1930s Western Electric (WE) specifications—the number of 300Bs to choose from now includes about 15 different brands and types.

This calls for an evaluation. A few of the 300Bs examined here have been covered in subjective reviews in other magazines (Vacuum Tube Valley, Fall/Winter 1997, issue 8; and Glass Audio, March 1998, Vol.10 No.2), but the variety of 300Bs tested below is, I believe, the widest yet, and Stereophile's first. I hope my findings help Stereophile readers in their choice of a 300B suitable to their amps and tastes.

I decided to limit the test field to tubes that are assumed to more or less adhere to the original specifications, as only these can be tried in any amplifier designed to take standard 300Bs without having to worry about compatibility. "Heavier" versions, designed for higher anode dissipation, invariably draw far higher filament currents than the 1.2 amps the old WE needed, and usually draw more anode current at a given anode voltage and grid voltage. KR Enterprise's VV52B, for instance, draws more than 2A of filament current and about 60% more anode current with a normal grid-bias voltage. As it cannot be used in most amplifiers designed for a 300B without the amp having to be modified, it wasn't included in this overview. The new Valve Art (VA) 5300 is a borderline case, with a 1.6A filament current, but as the widely used VA C60 also draws 1.6A, I decided to let it pass.

In most cases, the higher anode-current draw of some 300B variants is not a real problem. The bias arrangements of 300B amps come in two forms: "automatic bias" (mostly passive, using a high-power resistor in the cathode circuit), in which case the anode current is only mildly dependent on the tube's actual properties; or "fixed bias." In the latter case, the amplifier's user can adjust the negative grid voltage so that the anode current specified by the amp's designer is met—unless, of course, the tube under consideration appears to be out of the amp's range of bias adjustment. If you cannot sufficiently reduce the anode current—especially if it is more than 20% higher than it should be—do not try a 300B in your amp for more than a few minutes. You might blow the amp's output transformer or power supply.

A higher filament current cannot be readjusted and therefore can be a problem. Filament current supplies come in three forms: AC, regulated DC, and passive DC (ie, silicon-diode rectification into an electrolytic capacitor, often followed by a low-value power resistor or a choke and a second electrolytic). In all cases, the filament winding on the mains transformer must be able to deal with the current drawn. Fortunately, most designers build a bit of extra margin into their mains transformers. Still, you should monitor the transformer's temperature, especially in the first few hours after replacing the tube. If the amplifier starts to smell funny—or, worse, if there is even the slightest trace of smoke—switch the amp off and never again use that tube in that amp.

In the case of a passive DC filament supply, safe operation also depends on the current capacity of the silicon diodes (and the series resistor, if present). If the current is too high, the diodes die without warning—maybe after a minute, maybe on second turn-on, maybe after an hour, maybe after a day. Fortunately, diodes can be easily replaced by someone who knows what he's doing.

In the third case, the amp is probably equipped with a 5V voltage regulator from the 78 three-terminal chip series. These come in two varieties: the standard 7805 (designed for 1.5A maximum output current) and the 78S05 (good to 2.0A). Though these regulators have internal current limiting and thermal overload protection and should therefore be indestructible, in practice not all samples are. It is therefore safest to operate these devices somewhat away from their maximum current and dissipation. They can easily be replaced in case of failure—again, if you know what you're doing. Always use the 2A S-version (78S05) as a replacement.