Elrog 300B vs Takatsuki 300B tubes

I have measured this in my amp. Since my EML 300b XLS filaments draw 1.5 amps I have to adjust the dropping resistors so I still get 5 volts because voltage gets pulled below 5 volts. Other 300b tubes that draw 1.2 amps makes the voltage rise above 5 volts making me re adjust the dropping resistors. It's ohms law in practice. I heat my filaments with a/c. Jet

Thanks very much, Jet. Could you give us a rough idea of the magnitude of these voltage changes that you see in your particular amp?

I did a few minutes of research trying to find specs on the resistances of filament windings of transformers that would be used in these kinds of applications, and came up pretty much empty. It is surprising that such specs aren't commonly provided, which makes it hard to pin down the issues being discussed in any kind of quantitative manner. Which opens the door to speculation and differences of opinion.

Not to mention that it seems highly unlikely that data would be available for the particular tubes being discussed, or most other tubes for that matter, which would quantitatively indicate how their MTBF (mean time between failure) varies as a function of filament current, among other application dependent variables.

In fact the only conceivably relevant data I was able to find was for some military grade 60 Hz filament transformers specified in MIL-PRF-27 and its associated documents. For transformers intended for applications that are roughly similar in terms of voltage and current those documents generally indicated filament resistances in the area of 0.1 to 0.25 ohms. Which seems reasonably small **IF** the transformer and its turns ratio are chosen by the designer based on a loading assumption that is at least roughly in the general ballpark of the actual load.

Best regards,
-- Al

Al-
when I got home this evening I measured my Brooklyn line voltage with my fluke meter and got 124v which is more or less normal. My EML300b XLS filaments measured 4.97v and my Sophia mesh measured 5.06v. That's about a 2% increase. I am using 5v filament transformers that are over spec. They are rated at 3amps.

Thanks very much for providing that info, Jet. I've done some fairly lengthy calculations based on it, the results of which are listed below. These results assume (as I would expect) that the dropping resistor you mentioned is on the primary side of the transformer, that its setting wasn't changed during these measurements, and that the tubes were reasonably warmed up for the measurements. The results of my calculations are as follows:

Filament winding impedance at 60 Hz (which I would expect to be not greatly different than its DC resistance): 0.33 ohms.

Open circuit output voltage of the filament winding (i.e., the voltage it would put out if zero current were drawn from it): 5.46 volts.

The 0.33 ohms is just a bit higher than the 0.25 ohm spec I had mentioned for one of the MIL spec transformers for which I found specs, that transformer being reasonably comparable in terms of its voltage and current ratings.

I note that in contrast to most manufacturers EML actually indicates a tolerance on the 5 volt filament specification of their 300B-XLS, that being +/- 5%. Kudos to EML for providing that. Although I still wouldn't count on prolonged use of the tube at or near the upper limit of that tolerance to not result in some degradation of MTBF (long-term reliability).

5.06/5.00 is an increase of about 1.2%, which seems very comfortable, at least in terms of reliability if not sonics. But I suspect that you have done a very wise thing in using a dropping resistor to correct for your relatively high (although certainly not uncommonly high) line voltage. Even if the transformer is designed for an input of 120 volts, as opposed to say 117 volts, 124/120 is a 3.3% increase all by itself. And the MIL specs I looked at generally indicated transformer tolerances of +/-2% or more. The sum of those two deviations, if not corrected for as you have done, would already exceed EML's 5% tolerance recommendation if the deviation in the transformer happens to be in the upper half of its +/-2% range.

Turning to rectifiers, for a 3 amp tube obviously a winding having a somewhat heftier rating than the 3 amp rating of the one you are using for the 300B would be used, so chances are its resistance would be a bit less than the 0.33 ohms. Let's say 0.25 ohms, the rating of one of the MIL spec transformers I looked at. Substituting a 2 amp tube would result in a voltage increase, relative to use of a 3 amp tube, of 0.25 ohms x 1 amp = 0.25 volts = 5%, which does not take into account line voltage variations, transformer turns ratio tolerances, and other factors that could further increase the voltage applied to the tube. Not an immediate disaster by any means, but certainly undesirable and a cause for significant concern over the long-term.

Thanks again. And I hope that some of those reading this thread will find this analysis to be helpful. Best regards,
-- Al

P.S: A sharp-eyed reader may notice that at one point in my post just above I kind of juxtaposed EML's +/-5% tolerance with the 5.06 volt measurement Jet provided for the Sophia tube. I'm assuming that had Sophia provided a +/- tolerance on the 5 volt supply, the number would have been similar to EML's.

Best regards,
-- Al