[papagenohyun]

hi... hagtech.

i saw the circuit of cymbal poweramp, and amazed at how simple and brilliant ideas its topology has.

to my understanding, one 6h30 has dual function of voltage gain and phase splitting. and also it has ccs on its cathode .so it blocks mixing signal pathway and powersupply pathways in voltage gain and also helps to make symmetrical phase splitting role.
ususally i have seen many circuit or idea using input transformer which they insists on better performance. but i think your idea has very simple and cheap solution .

i am also interested in your powerstage design.
at first, i think it is usual and common pushpull design. not thing new.
but i have read it several times, and i doubt that it is somewhat different from usual pp design. because power tubes also has ccs on their common cathode.
i think it is partially cathode bias and is also helped by ccs.
maybe i am not tube amp designer or speciallist. only hobbist
so my understanding is not perfect or far from truth.
but my guessing is that
if the ccs can carry all the currents needed for output tubes, i can remove C10 (220uf) cap and cymbal amp will be born as fully differential poweramp.

maybe it is not easy job to make ccs which can carry those current ( heating problem, and stablility...) but i guess that you must have tried it.

anyway, i am very impressed with your idea .....

fully differential amplifier is really something special and very different from usual poweramp.

south korea

dong hyun

[hagtech]

Dong Hyun,

Thank you for your comments!  Your insight and analysis of the input stage is correct.  One dual tube does all that.  And provides high input impedance, gain, and low output impedance.  Something a transformer cannot.

The output stage is fully differential class A up to a point.  For very large signal swings (or if speaker impedance is way too low), the circuit falls into class B.  But only when rated output power is exceeded.  R24 and R32 form a self-bias reasonably constant current source.  A true current source could be used here (see Allen Wright's new stuff) but clips hard on the peaks, like a solid state amp.  That's why I stuck with the resistor.  Clipping on peaks is not so hard.

C10 isn't really in the circuit for the output, as it is on the other side of R29.  Ok, it affects current by 5%.  But it is needed to remove any residual ripple from the 13V tap and is used as supply bias for the input stage current source.

jh