In article ,
Ian Thompson-Bell wrote:
Eeyore wrote:
John Byrns wrote:
I concentrated on the design of the mixer section
Yes, that'll require a triode with AC coupled nfb from anode to grid to form
the virtual earth, the remainder being routine mix Rs, level controls etc.
That's hardly a DESIGN. It's a stock circuit.
Graham
Hmmm, that assumes virtual earth mixing is the way to go for tubes. A
typical triode will have a stage gain of 20 maybe 30dB so the virtual
earth will not be that good (Rfb/30 maybe) and there's not really enough
gain to make up for mix losses. Also Rfb needs to be large enough not to
significantly load the anode, so we are talking 300K or so here which
means the mix resistors will be 100K for for 10dB gain make up.
I have at least for the moment dropped the idea of virtual earth mixing
from my design, however my original virtual earth design used a feedback
resistor around the tube providing the virtual earth function, equal in
value to the resistors used for each channels connection to the virtual
earth mix bus. Driving this relatively low value feedback resistor was
done by using the same buffer circuit, perhaps a CF, used in the channel
modules which have the same problem driving the mixing resistors. With
this scheme of equal value resistors the mixer circuit proper has a gain
of unity.
If you want to stick with VE mixing then a tubed long tailed pair and
another triode will make a simple op amp like circuit where Rfb can be
much smaller and the open loop gain is much higher making a better VE.
Or simpler still just stick a CF on the end of the triode and feed its
cathode back to the grid via a lowish Rfb.
That was essentially my original scheme, noting that the problem of
driving the feedback resistor is no worse than that of the individual
channels driving their mixing resistors.
Regards,
John Byrns
--
Surf my web pages at,
http://fmamradios.com/