I designed and built (after MUCH research) a tube poweramp that attempts to bring out the best of tubes for Hi-Fi, while being very clean. It's 2 6SN7's driving a push-pull EL34 output stage. Since feedback is limited to 0 - 12 dB, the plate load distortion in the output stage will give it a slight amount of real-time compression, and since the feedback path excludes the first stage, the distortion spectrum of the first stage is left un-tainted by the use of feedback. Since I replaced the usual plate load resistor in the front end with a very good current sink, linearity is improved and power supply rejection is very high. The middle stage is a differential circuit sitting on a current source. The output stage can be run in ultra-linear or triode mode. The distortion spectrum viewed on a spectrum analyzer is very natural, and close in to the fundamental, until clipping, and that's what it's all about; getting a natural distortion spectrum shape that doesn't extend way out from the fundamental, like virtually all high feedback designs do. It sounds very pure and natural.
Due to the inevitable mis-match of the phase splitter triode sections, and output tubes, there will be some 2nd harmonic generated there. The optional feedback can partially correct for that. A symmetrical circuit with high feedback will exclude even harmonics from the distortion spectrum, creating an un-natural distortion sound. To be fair, it's so far down with test signals on the bench that it's hard to say if we sense it at all. With real music, it may be an issue.
The downside is the output impedance, which lets the speaker resonance have it's way to some degree. My system is tri-amp'd, so this amp is just doing from 1.4kHZ on up, so the output Z is a non-issue. It drives the Seas Millenium 1 inch dome tweeters (same as in the Linkwitz Orion). I use hafler DH-220s (transistors) for the mids and woofers. The active crossover is 4th order LR, with active EQ for the open baffle mids and the closed box woofers. The system as a whole measures very flat from 20HZ to 20kHZ at my listening chair.
Due to the inevitable mis-match of the phase splitter triode sections, and output tubes, there will be some 2nd harmonic generated there. The optional feedback can partially correct for that. A symmetrical circuit with high feedback will exclude even harmonics from the distortion spectrum, creating an un-natural distortion sound. To be fair, it's so far down with test signals on the bench that it's hard to say if we sense it at all. With real music, it may be an issue.
The downside is the output impedance, which lets the speaker resonance have it's way to some degree. My system is tri-amp'd, so this amp is just doing from 1.4kHZ on up, so the output Z is a non-issue. It drives the Seas Millenium 1 inch dome tweeters (same as in the Linkwitz Orion). I use hafler DH-220s (transistors) for the mids and woofers. The active crossover is 4th order LR, with active EQ for the open baffle mids and the closed box woofers. The system as a whole measures very flat from 20HZ to 20kHZ at my listening chair.
Attachments
okay guys, thread closed for now, thread has run its course....- Status
- Not open for further replies.