Quote:
Originally Posted by Beefy /img/forum/go_quote.gif
So over the last week, instead of writing grant applications
, I've read the entire thread here on Head-Fi, as well as about 100 pages in total over at diyaudio.org. I understand most things happening with the tweaks, but still have a couple of lingering issues.
Firstly, output resistors. I understand some people don't like them because they hurt dynamics...... but isn't that a direct result of most people using carbon film resistors? I know that the likes of Stackpole, KOA etc. are the tube-enthusiasts-darling, but it was my understanding that carbon film is inferior to metal film across pretty much all measured characteristics.
|
Decent guess, but no. Metal film just sounds bad in the output positions, period - and yes, I've tried both on identical amps at the same time. A high value metal film will sound noisy and lose dynamics - a very bad sounding combination.
It's true that it hurts dynamics. This is immediately apparent on high impedance phones such as Sennheisers. With low impedance - something like KSC75's or Grado's, the dynamics are not noticeably hurt at low resistor values, but some of the "tizziness" goes away and there's a soft sheen on the highs (with some sort of carbon resistor, there).
Quote:
| Secondly, power transistors. How are people honestly comparing the different options? I find it difficult to believe that you can get such big distinctions as people claim between these parts, when so much time and effort passes between testing one versus another in the same amp. Unlike testing between caps, where plenty of people have identical builds except for the caps, nobody seems to have rigs where the only difference is the transistors. Or if they do, they are Black Gate models which would probably sound subtly different even with the same transistors, depending on burn in status and individual tube characteristics. |
The differences are not subtle, and yes - I have 6 MAXes right now, each with different output transistors - also had four regular Milletts before that, each with different transistors in the DB's. With the Millett, you could take the same amp and plug/unplug as many different DB's as you had on hand. That's probably how Steinchen made his original reviews.
Moreover, I've heard nothing to dispute Steinchen's original work on such reviews (as referenced in Regal's link), except that I disagree slightly with his description of "rough highs" on the 2SC3421/2SA1358. (I think with the right output cap and higher bias currents on the MAX, the highs are fairly smooth - they're definitely emphasized in those transistors, though - so is the bass, but in extension, not in slam.) If you read the older Millett threads over on DIYForums.org, Steinchen at one point declared the 2SC34221/2SA1358 pair his favorite. So, things change as you try other transistors and note the differences.
There's also a point I made not too far back in this very same thread. AMB took issue with it because it read as if I was saying that the tubes did not make the most difference with sound. That's true to an extent - what happens in my experience with these things is that differences in tubes of the same type are more or less overwhelmed by the buffer. This was obvious in the older Millett, since you could plug and unplug the DB's versus a monolithic buffer. With the monolithics, even small differences in the tubes between brands seemed to make a difference. That goes away, IMHO, when a strong discrete buffer is applied.
Now, that said, changing from a 12AE6 to a 12FK6 will affect a huge difference in sound. Personally, I still think it's debatable in some instances whether that difference is greater than the difference between changing from a 2SC3422/2SA1359 pair to a 2SC3421/2SA1358 pair. Those two are very different, for instance. Whereas, differences between the 2SC2238/2SA968 and the 2SC2344/2SA1011 are much more subtle.
As stated in some of those posts, though, the gain and sound difference are large enough between the 12AE6 and 12FK6 that many people settle on one or the other. That makes any changes from then on are effected almost entirely by the transistors.
Keep in mind that at this point, there are very few amps that have as much history and refinement in the DIY headphone community as the Millett Hybrid. The basic circuit has been built for years now, and some of us are on our way to building a dozen or more.
Quote:
Yep.
