Welcome! Need support, you got it. Or share you ideas and experiences.
24.21.130.133
In Reply to: RE: eros: one man's review (amongst a dozen or so) posted by jivetrain on June 27, 2009 at 23:26:57
I agree with a lot of that! I still have more background noise than I want, but while the music plays it sounds great, so I may just get used to it. I tried a couple different tubes and hear very little difference in the noise or musicality of the preamp.
As for improving the amp, longer screws might lessen the cursing involved in the socket installation. The shunt regulation can be improved - the lm431 is an 11 cent part causing some issues - it at least should be bypassed. And the printed circuit boards shouldn't have jumpers - they should be customized for each module.
All that said, it still sounds really really good when playing music, and that's the point.
I'd like to address some of the issues here, for those interested in the design questions and decisions.
First, the EF86 is a low-transconductance pentode so it naturally has more white noise than a high transconductance triode like the 6DJ8/6922 types. This is somewhat, though not entirely, mitigated because it is an audio design, not a VHF radio design, so there is some reasonable hope that the low-frequency "pink" noise (a.k.a. 1/f noise) will be more well controlled. And the greater gain keeps the second stage from contributing as much high-audio-frequency noise. In most cases its noise is still well below the vinyl surface noise. But the main point is, this is a great sounding tube - worth (IMHO) suffering a few problems to get that certain bit of aural magic.
I'm not clear on what issues you see due to the 431 chip - ping me, offline if you like, if you want to discuss specific things. I'm always interested in potential improvements! The shunt regs are bypassed in a way (at high audio frequencies) by the stability network. However, in general it's hard to design a shunt regulator that does not rely on capacitors in the signal path, and caps are always a bit problematic for the sound. For what it's worth, we did extensive listening tests with 6 different shunt regulation circuits before settling on this one as the most musical sounding. In the case of the Eros, we went to some trouble to isolate the gain stages from the regulated power supply with current sources, further minimizing the PSU contribution to the sonics. But as I have said before, I have never met a poser supply improvement that was not also a sonic improvement.
I certainly agree about the advantages of specific PC boards for specific purposes. They would be a lot easier to design, too - I spent many many hours making these to serve in all the Bottlehead products! However, the economies of scale are pretty compelling for PC boards. Dedicated boards would add a bit over $100 to the price of this kit.
Post a Followup: