Elrog 300B Review - The Best 300B Ever?

@purr1n regarding the tone bursts... Is it possible the Elrog has a higher bandwidth than the other tubes?

This pattern is similar to what we would get out of a high quality speaker.

Unfortunately too many modern speakers with their hard coned drivers are dominated by D3 instead which also doesn't decrease as much with lower SPLs as D2 does.

Now the total killer would be if the phase of the distortion was opposite to what a speaker would do so it might cancel out a bit.

Someone on DIYAudio did an FFT and the SET amp had lower distortion at the output of his horn speaker than the SS amp did.


Anyway, very interesting stuff. Now I wonder about harmonics vs frequency for different output powers. Like maybe 0.01W, 0.1W and 1W. Higher harmonics could be vanishingly low at low power.

 

@purr1n holy crap, that is impressive for an OPT coupled amp! 100kHz flat. No signs of major ringing, too. Otherwise I also didn't expect much of a difference between the tubes as you said. Thanks for measuring it anyway.

I bet distortion vs frequency vs output level is going to be more interesting.

 

Well, the Line Magnetic stuff (albeit that's higher power) goes to 50kHz only.

 

The WE lines for 1v and 2v look to be farther apart, so distortion increases more towards maximum output than with the Elrog and EH. How about thd vs gain at fixed frequency? Does that reveal any qualitative differences?

 

I dont think ive been around long enough for most other investigatory threads on SBAF. Seeing it all roll out gives a better sense of the collaborate effort, almost like those shows where internet groups get together to solve a crime.

As someone who is heavy into and heavily favors vintage tubes, thia hunt for "vintage tube arcana" is very compelling. I bought a pair of date matched Amperex 7308s because they were made a few miles away from where i grew up (the place is now a uniform factory). I jave nothing to use them on though.

Some barely educated thoughts, please correct me if I am way off base or am just too general, but doesnt glass shape influence a decent chunk? Many saying the pinched waist 6922 variants that were some of the best even though it was a defect. The pinched in waists tended to keep all the great aspects but have a tighter overall control. With the many 300B shapes maybe there are some clues there.

 

What can we do to measure transient performance?
FFT, square waves, multi-tone? Where can I apply the bees knees to my oscilloscopes.

can these also be applied to transducers?

it seems people had criticized csd in the past but it seems like times are changing, and those may soon be under consideration for the mass *marketing*

I think for transducers the sauce is in the impedance v fr + phase.

 

Just some ideas I've been playing with for a while. Damn it I wanted to write about them since last winter, but they're still half baked in my google doc and my mind. The article of transient response on DAC digital filter is closed to finish though.

1. Maybe it's the flicker noise or thermal noise of those tubes are different. It's a pretty decent metric to chase for while building solid state amps, but it's pretty rare to see voltage noise density charts of tubes. There might be a difference in this case.

2. It might be helpful to plot the impulse response of these tubes. The plot might be more meaningful if the y axis is plotting the absolute value of the voltage levels in log scale, so you can see more of the little tails. It's been pretty helpful for me to visualize digital filters in such way. Additionally CSD plots might help visualize the microphonics.

3. For the distortion characteristics, you can curve trace the tubes around the operating points, and retrofit the I/V curves with equations like x^1.5, x^2, or e^x. I've tested some mosfets, some fits the e^x curve more closely (which will sound more like BJT), and some fites the x^1.5 curves more closely (which will sound more like tubes). I don't suspect there's a huge difference between those 300Bs though, since they are based on similar constructions.

 

I don't think square waves get enough loving. A super-fast rise from a 10 kHz wave, without ringing or overshoot, should represent the worst a transient has to offer.

 

For anyone that's heard all or some these: how do the Takatsuki's compare to ER (plain/moly/TM versions), WE, and LinLai Elite/Psvane Acme globes?

 

I got a pair of these (non-Mo/TM, just the original ER300B) in yesterday for Studio B second hand, sub-500ish hours, manufactured last year. It's early days, but they're the real deal. Marv isn't kidding when he said they're dead neutral and not "classic" 300B. Out of the five pairs of new production 300B's I've tried (three Chinese, Gold Lions, and these), they're far and away the most idiosyncratic, just totally their own thing. The thoriated tungsten filaments just don't sound like anything else, I think. This is true even with the Studio B, which definitely exhibited less change/delta between all the other 300B's I've tried, and is known for sounding good even with cheap 300B's. There's more extension on either end but with a nod toward the highs. I haven't really decided if I feel like there's "less" midrange, but you'd need to be careful with component/transducer matching, or you could definitely end up feeling that way - they're that less bloomy and tooby than any other 300B I've heard. If you're like Marv and feel ambivalent about the 300B sound, these definitely may fit the bill re: not sounding like a "classic" 300B. But man, the resolution and the dynamics (and resultant sense of air) they emphasize without that 300B syrup “in the way” can really get your attention -- with something like the HD800 and Studio B that already play things toward those qualities, some recordings can start sounding spooky, others can sound harsh, although still in a way that made me attribute those qualities to the recordings more than the tubes. Keep in mind with my impressions though that Studio B is also adding some feedback in there to shave off the 300B toobiness already, so with an amp sans that small amount of feedback, you might gain back some tonal meat on the bones/bloom.

Really would encourage anyone who has the opportunity to give them a shot, they're good enough that, like Marv indicated, if they float your boat you might feel inclined to dial things in around them instead of reaching for another tube. I know Marv mentioned the possibility of using SBAF funds to purchase a pair from Elrog for a loaner when he made the thread, would be cool to get more ears around here on them.

edit: Can confirm that these play better with a driver tube that is more tooby. I was using Bendix 2C51 with the Linlai WE clones, but it is too clear/detail-forward/neutral with the Elrogs, they do much better with a JAN GE 5670W. Getting a 396a in tomorrow and a couple other tubes that should also be more tooby than the Bendix (although Bendix is on the least tooby end of the 2C51 family, prob, so not hard to do).

It just goes to show you that Marv is right about being a good chef. Neutral plus neutral might be too neutral. I’m finding Yggdrasil GS to be fine in this config, but that might be a sonic priorities thing for me - I am usually much more tolerant of sharp transients and high frequency fuckery than a lot of folks. I can see where someone might want to run LIM with these or something else darker, even a NOS DAC.

 

Shitty iPhone picture to show the cool thoriated filament toob glow. Cameo by Amazon doorstop nervosa and Daruma doll/tiny succulent plant.