[email protected] wrote on 03-04-2007 13:13:39: > David Gravereaux wrote: > > Ian bell wrote: > > > >>David Gravereaux wrote: > >> > >>>That's a push-pull arrangement. As the ripple is identical and > both tubes have > >>>matched gain, the ripple noise cancels in the transformer. > >>> > >> > >> > >>Are you sure about that? > > > > > > Yes. > > > > > >>Each valve conducts for only half a cycle. > > > > > > Not true. You forgot about the class-A range of overlap. 30% overlap isn't > > uncommon. It's only at low output level where PSU ripple is a concern. > > > > Debatable on all counts. Unless it is pure class A, there will be > parts of each cycle where only one valve is conducting. Unless the PSU > is fully regulated ripple will increase with increasing signal. Ripple > to signal ratio is likely to be similar at all signal levels. > > > >>It > >>is like summing two single ended amps fed with anti-phase inputs. Each
> >>is still affected by ripple. > > > > > > Yes, and they cancel in the transformer. > > > > I don't see how they can. I'll need to go simulate it to check it out. > > > Is this topic done yet? I'm done with it. > > > > I'm not. It's interesting. > > Ian as your not done!! My 2ps worth, if you use pentodes the high output impedance sort of isolates the hum as well because the current in the transformer doesn't change much with the bouncing voltage, doesn't work so well with triodes, who's anode impedance is lower.... dc
