John Byrns wrote:
In article , Patrick Turner
wrote:
I have built a couple of subwoofer projects which
included a 12" speaker in a large ported box driven by an SS
amp to augment a small pair of floor standing speakers.
The L and R signal from the line stage part of the integrated tube power amp,
( used for most of the signal ) was taken out with two cables to two 22 k
resistors, and at the join of these two R is a mono signal, and this is then
fed to a high input impedance emitter follower transistor stage,
Don't these 22k resistors create a noticeable amount of cross talk, given
the relatively high impedances involved in tube amps?
Not much cross talk because the junction of the R is shunted with a cap to ground,
and the bass signals are usually the same for both channels.
The take off point from the line stage is from a relatively low impedance point,
being the wiper of the 100k log volume controls and in normal use, its at a low
impedance setting; at 12 oclock, at -20dB, the output impedance
of the pot is 9k.
The stereo info is contained in F above 250 Hz, and the cap shunting the 22k x 2
prevents much cross talk above this F.
By 5 kHz, where you sure don't want any cross talk, my method is
quite OK.
Why not include a
couple of high input impedance emitter follower transistor stages in front
of the resistors?
This is the more purist way of doing it, but then its an extra stage.
Mr Gangyi has to learn about transistors and filters.
Before I ferret through a mountain of paper circuits I've used,
and scan them in and transmit them, I think he ought to just try fooling around
with a few 10c small signal transistors, and a few R and C elements.
It will be a revelation.
He should find out that in common emmitter mode, with a bypassed emitter,
the transistors producing 3 vrms are hopelessly distorted, over
several %, but at 4 mA, the gain will be huge.
Then when the emitter R isn't bypassed, the gain will plummet,
and the thd will then seem similar to a tube, but still not as low.
Then in emitter follower, the transistor is at its best, ie,
has the lowest thd, and competes well with a cathode follower.
The way I did it means there is very low thd, and very little circuit complexity.
This sort of project is just perfect to train the mind to basic analog electronics,
some understanding of filters, and cascading them, settinjg bias currents,
choosing load values, understanding load lines, and it'll put a rumble in a lounge.
There would be many readers wondering where to start;
I suggest they buy a few small NPN signal transistors, and some pre drilled
board for experimental circuits, some 0.6 mm dia solid hook up wire,
and first build a little +/- 12 volt supply, and a volt meter,
and a book about basic transistor circuits.
Opamps could easily be used instead, but in this case, there is no need.
Since my idea needs emitter followers, cathode followers made from
12AU7, 12AT7 would be quite excellent.
But there is a lot more trouble to be gone to for just a bit of low bass.
There is no need for any gain in the active devices.
The following SS power amp input sensitivity can be
adjusted for say 0.5 v for full power by adjusting the FB network
if need be. Some presettable gain pot for the sub amp is usually necessary.
Patrick Turner.
Regards,
John Byrns
Surf my web pages at, http://users.rcn.com/jbyrns/