There is one matter regarding cathode feedback that must be kept in mind, and that is that e.g. the Quad II uses the cathode winding such that the output tubes work in UL mode. Thus, the characteristics are much closer to triode than tetrode operation.
Brandersnatch touched on this briefly. Simply comsidering a cathode winding as a pentode with (external) local nfb, would have been true had the screen grids been decoupled to the cathodes. Redrawing e.g. the Quad or similar circuits for ac, shows the screen to lie at the familiar UL tap position. Thus, should those cathode windings also be say 30% of the total primary, one really has almost triode operation before local feedback, with all the other advantages of UL.
Somewhat OT, just for interest: Such a figure for the 'taps' demands a quite high grid driving voltage, almost impractical - thus the use of McIntosh of an interstage transformer. In a case of my own I use 4 x 6L6GC in p.p.-parallel for a 100W output, where a 25% cathode winding demands some 380Vpp of signal to the power stage. I found this akward; had to drive hard with an E182CC from 520V h.t. (I have since lowered the ratio to 20%). I do however get 0,12% thd at full output with 22dB of global nfb and very low high order products as a result of the mainly triode operation; not to be sniffed at.
Brandersnatch touched on this briefly. Simply comsidering a cathode winding as a pentode with (external) local nfb, would have been true had the screen grids been decoupled to the cathodes. Redrawing e.g. the Quad or similar circuits for ac, shows the screen to lie at the familiar UL tap position. Thus, should those cathode windings also be say 30% of the total primary, one really has almost triode operation before local feedback, with all the other advantages of UL.
Somewhat OT, just for interest: Such a figure for the 'taps' demands a quite high grid driving voltage, almost impractical - thus the use of McIntosh of an interstage transformer. In a case of my own I use 4 x 6L6GC in p.p.-parallel for a 100W output, where a 25% cathode winding demands some 380Vpp of signal to the power stage. I found this akward; had to drive hard with an E182CC from 520V h.t. (I have since lowered the ratio to 20%). I do however get 0,12% thd at full output with 22dB of global nfb and very low high order products as a result of the mainly triode operation; not to be sniffed at.
With some of the more "modern" tubes and circuitry it should be easier to get rail to rail swing than previously. Forgetting about stinky 12AU7 and 12AX7's and going for ECC99 or 7119 and 5965's or dual D3a's for a "one off" for personal use along with some Sowter or Lundahl intgerstage trafos should do the trick nicely. I appreciate the comment about comparing a stock MCxx with a refreshed or even a tricked out MCxx would be something to die for. I have used MA230 output transformers on a beefed up Eico HF-89 and it took the HF to another level. Bigger filter cans, better coupling caps, a better front end etc... The difference was dramatic as the marketing people say. And the MA230 transformers are a level below the MCxx Unity Coupled beasts. Plitron may not spec their transformers at a reasonable power level but there is no tricking my ears that the PAT 3025's I have are anything less than stellar, and are good for more like 20 watts than the 13 they advertise. Let's give credit to them for not overstating their specs. All specs are based on a number of assumptions and they are no different.
When comparing output stage configurations on a numerical basis, the unity coupled come out on top for a reason, and the fact that they need a large swing should not deter us from trying. No one said it would be easy to make the best. If one does not need huge power as many do not with better, more efficient speakers, one could try PP triodes with cathode fb for close to the top numbers, to simplify the trafo and circuit design needed.
When comparing output stage configurations on a numerical basis, the unity coupled come out on top for a reason, and the fact that they need a large swing should not deter us from trying. No one said it would be easy to make the best. If one does not need huge power as many do not with better, more efficient speakers, one could try PP triodes with cathode fb for close to the top numbers, to simplify the trafo and circuit design needed.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.