10-27-15: Dracule1
Thanks Al. I got the distinct impression from George at TubesUSA to stay away from rectifier tubes that draws more than 2 amps. The list of rectifiers recommended by George for the Elrog 300Bs I think draw no more than around 2 amps (5R4, 5R4GY, GZ34 and GZ37). I'm afraid to try my EML 5U4 (3 amps) with my Elrogs.
The 5R4 tubes and the GZ34 are approximately 2 amps, although the GZ37 is about 2.8 amps. I’d question whether George really meant to indicate that you should stay away from tubes drawing significantly more than 2 amps, given that all of the tube choices Lampi supplies with the Big 7 are around 3 amps, and the choices provided with the GG are 2.4 and 3 amps. And if he did really mean that, I would want to know what his reasoning is.
The 2 amp draw of the Takatsuki you are using will result in some increase in the voltage supplied to the filament of that tube, relative to the voltage that would be supplied if the tube were drawing a higher current, due to the lighter loading of the 5 volt winding of the power transformer. Matthias of Elrog made exactly that point, although in connection with a 300B used in a power amp rather than a rectifier used in a DAC, in a post in this thread dated 6-19-15 when he said:
If an EML 300B XLS is the right choice we have 1.5A heater current at 5V. There is no voltage regulation. A transformer a rectifier bridge and a CLC only. May be the XLS is underheated (so I hope) ore all other (right) 300B are overheated, because they draw 1.2A (some russians and chinese 1.25A)only. This would result a heater voltage above 5V...not so good for any 300B and very bad for an Elrog 300B, because the use of thoriated tungsten for filaments. Higher heater voltage will result a (much) higher anode current and a much lower lifetime...
Now that increase in filament voltage in this case might be small enough to not matter, or it may not be. We have no way of knowing, and I’m not sure that George does either, with respect to the specific design of the Lampi. And given that a 300B and these rectifiers both have 5 volt filaments, perhaps the design even utilizes the same transformer winding to power both, which would result in the 300B seeing more than the 5 volts it is designed for. That suspicion is heightened by the indication you provided that the Lampi applies AC to the filament of the 300B.
So regarding your statement that "I'm afraid to try my EML 5U4 (3 amps) with my Elrogs," it seems very conceivable to me that using a rectifier having a 2 amp filament current draw might be more stressful to the Elrog than one having a 3 amp draw. Especially given the very high 600 volt plate rating of the Elrog tube, which together with Matthias' statement quoted above would seem to indicate that it is more vulnerable to a filament voltage increase than to a B+ voltage increase that might result from a rectifier change.
There are undoubtedly other significant differences in the electrical parameters of these various tubes, which I haven’t taken the time to study and which don’t seem to be available in all cases. But I would certainly expect that between those differences and differences in the value of the filament voltage resulting from differing current draws there would be significant differences in the DC voltages supplied by the rectifier circuit to the rest of the DAC, including the output tube. Which can certainly be expected to result in sonic differences, for better or worse, even apart from differences in the intrinsic sonic characters of the tubes. But not having knowledge of the specific designs we have no way of knowing what the long-term effects on reliability might be.
As I said earlier, changing a rectifier to one that draws significantly different filament current than the one the unit is supplied with amounts to incorporating a design change into the unit. As an electronic circuit designer (not for audio), it goes against my grain to implement a design change without having either a means of understanding its ramifications, or an ok from the designer or some other source having specific knowledge of the design.
Bill (Brownsfan), it looks like the CV387 is a small pentode, so you may have mis-typed that number. Otherwise, though, given that I see on Brent Jessee’s site that the VU-71 is equivalent to the 5U4GB, and given the rest of the facts stated in your post, including operation at a lowish line voltage, I don’t see that you have any issues. Enjoy!
Best regards,
--Al
