|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
24.34.17.88
In Reply to: Jim Hagerman's Octal Cornet--Can't Lose posted by slowburn on March 3, 2007 at 18:11:55:
Good looking project. I've also got some old iron in the junk box that I would be looking to use as well.It seems that everyone that has built the Cornets has had very good results with them. This design also seem to respond well to component rolling. I have also read over at DIY Audio that the RIAA eq accuracy is very sensitive to tube rolling? Would you care to share your thoughts on this? Is this really an issue?
Follow Ups:
Indeed it is sensitive to tube rolling. Most other designers (ok, so far all of them) choose to minimize the impact of tube aging or rolling against any shift in EQ. I do just the opposite. I say bring it on.What I discovered was that the tricks used to de-sensitize tube variations also end up screwing with the sonics. The result is that the sonic flavoring of various tubes is retarded. Good tubes don't sound a whole lot better than lousy tubes.
The Cornet is designed to allow a tube's full sonic potential. You will hear distinctive changes when rolling tubes. The price to pay, however, is that you end up with +/-1dB variations in RIAA EQ. Older tubes will actually warm up the sound a little (rather than getting thin) with a slightly boosted bass and sloped down treble.
So take your pick. Super-accurate EQ, or fabulous tube rolling.
Hi Jim,Thanks for the explanation. It makes perfect sense. It does however bring up another question of how to maintain channel to channel match. I believe that a channel imbalance is going to be more audible and an absolute variation from the RIAA response. Resistors and capacitors can be matched between channels to quite tight tolerances. Am I correct in assuming that tubes would then also have to be selected for very tight section to section match? How close a match and what parameters should be matched? Will the individual sections age the same?
I know, a lot of questions, but I am just trying to understand the implications of using the tubes on hand vs. buying tightly matched NO$ vs. tightly matched current production.
The channel match is very important. And that is one of the big reasons I use dual triodes. This is yet another design tradeoff I make. I purposely sacrifice a little bit of crosstalk in return for matching. You see, both sections of that dual triode will age pretty much the same. So what happens in one channel, will happen in the other.Of course, the sections have to be pretty well matched to begin with. I use a VacuTrace for production testing. NOS usually aren't too bad, you just have to tell with your ears (not everyone can afford the fancy test equipment). If it doesn't sound right, try another tube. The important parameters to match are transconductance and plate resistance at the particular bias point.
The other reason to split a dual tube between channels is that it prevents a situation of instability. If you run two successive gain stages in one glass envelope, there is the likelyhood of positive feedback via parasitics. High gain and 360 degree phase shift is just asking for trouble.
One look at my construction ought to make it abundantly clear that I'm no circuit designer.I do, however, use fully functional ears and they're connected--somehow--to both my brain and heart.
Screw "accuracy," whatever that means. Plus or minus 1% in ANYTHING is unlikely no offend your sensory perception.
Mr. Hagerman's phono stage brings genuine flesh and blood to the party. Moreover, I've tried 15 or so tubes in various configurations in this little beast. They all sing but I like some voices more than others. And I didn't have to strain to wonder if I was imagining a different sound. In Mr. Hagerman's Octal Cornet, sonic flavors are easy to alter: just swap some decent old tubes, sit back, listen.
For now, I've settled on a WW2 Sylvania JAN 6SL7 in the first position. (black cylindrical plates, bottom getters, little US Navy anchor logo etched in the black base)
In the second stage, a 1960s RCA 5691. Fetishists don't drive up the price so badly on these newer, "modern logo" red base tubes, but they still sound fantastic.
Intuitively, ya'd think that the ultra quiet 5691 would be the best candidate for the first stage. Believe me, I wanted it to be so. But that old Sylvania has a top-to-bottom dynamic spark that's irresistable. Also, it is QUIETER than the the 5691. I kid you not.
Cathode follower's an early '50s brown base JAN Tung Sol 6SN7.
I never imagined that a homebrew phono stage built on my dining room table could ever run with a well-regarded commercial model. All I'd hoped for was a little DIY satisfaction.
I was dead wrong.
That ugly cake pan, stuffed with Mr. Hagerman's circuit and some sweet old tubes and iron, runs circles around most phono stages I've heard. That includes big-ticket items I've listened to at boutiques and shows.
And like I said, it utterly outclasses the very solid ARC SP6B's phono stage, even after my renovations and refinements.
In any case, I'd like to second Mr. Hagerman's assertion. It's a simple and elegant circuit. As such, it theoretically should reveal component flavors more than a more complex, feedback-ridden circuit. In practice, it does. Build it and you will hear that tubes sound different. So do signal path caps and resistors. That the Cornet sounds terrific with generic components (which I tried first, before subbing Riken, Kiwame, Cardas, etc.) is a testament to the designers skills, his ears, his heart.
I'd pit my cakepan against any phono stage. Any. I suspect that very few would sound "better." I'd bet that it might even get you 90% of the sonics delivered by JH's own, all-out, Trumpet. He may disagree, but then again, he ain't heard MY Cornet.
; )
Thanks again, Mr. H! I love my Cornet. I'm a fan for life. Wish I could afford your DAC.
This post is made possible by the generous support of people like you and our sponsors: