[quote=Patrick Turner;957412]On May 25, 11:26*am, John L Stewart
wrote:
In a SEP amplifier an unbypassed R supplying G2 results in degeneration
of the signal. Too much R & the power handling ability will be
restricted.
If you think of it, the entire plate family is being modulated at the
signal rate. Just like a freq conversion mixer, so lots of Dist products
are possible.
In a PP amp an unbypassed G2 R supplies provides some balancing of the
signal, similar to an unbypassed cathode R as used in a diff amp, but
not as much. Works OK for Class A, but fails badly otherwise.
One of the primary reasons for using low value resistors on G1 & G2 is
to prevent parasitic oscillation of the amplfier. In this case the R
should be physically small & attached as close as possible to the
socket. The R lowers the RF Q of the parasitic circuit. Needs to be
physically small in order to avoid skin effect if possible. I've seen
parasitics to 150 MHz in audio amps I've built.
Regulated G2 is nice & gets the most out of a power amp. But then you
will need a stiff plate supply to take full advantage. Otherwise you
will simply modulate the B+ at twice the audio rate.
As Patrick has pointed out, get the soldering iron going & make lots of
observations. Apologies to others if this has already been covered. Too
busy here with manual labour & the bicycle!
During the last 12 months I rode 11,000 km. Speed average at 23kph.
Nobody over 55yo has overtaken me in last 3 years, except a few blokes
who ride in Canberra's Pedal Power group. There's plenty of blokes
fitter than me. But I continue with activities to keep me sane, and
out of hospital and out of jail, if not quite off the streets. But I
see a life as a stack of cards, and as one grows older, the stack
becomes taller, and ever so much more likely to tumble down. Life is
but a temporay therapy, with a final side effect. Doctors just delay
it.
Ya can't beat working out in a shed with adequate gear to teach
yourself, while doubting yourself, but then proving your doubts wrong.
When ya return from shed back to life after learning, one should be
better equipped to make better amps, otherwise, why go to the shed in
the first place? Why not paint a room, repair a floor, repair a toilet
cistern, or go for a bike ride instead? I like having reasons to
exist.
G2 is a control grid like G1, but with much lower gm. G2 input
resistance is low, and not linear.
I have models at my website for pentodes, in terms of current
generators for G2 and G1, with Ra as a shunt R, but what I have just
realised is that my models for the G2 current generator don't include
the input resistance to G2 in the models, however, in the models
explained, the outcome for a given pentode or beam tetrode can be
predicted better than otherwise methods all much more vague, more
approximate and unclear. Even the best of brains in 1955 didn't quite
create the best of equivalent models for tubes as can be devised, to
allow the examination and prediction of what happens when you have a
high value R in series wuth G2.
In such modelling, voltage gain and Effective Ra can be calculated
fairly well without worrying about THD because one always works with
small signal levels when THD is always going to be low because the
tubes is considered while working in class A, and the macro
transition between class A and AB, or AB2 etc is not considered. In
practice, PP OP tubes are usually always working in pure class A even
though they are biased at low Ia with high Ea to allow maybe 10 times
the possible pure class A PO in class AB, or AB2.
I have never ever seen any commercially successful tube amp design
which has high value R between B+ and G2, or between G2 and UL taps on
OPTs. The old Mullard notes about using EL84 and EL34 in UP PP amps
for class AB1 do include mention of using maybe 1k0 series R to G2
from UL taps and I recall they say THD is lower, but it ain't much
lower.
One would think everyone would have used a couple of cheap resistors
to feed G2 and leave it fed without any C bypassing if it led to
better amp performance, because resistors are cheaper to implement
than a tranny tap, and leaving out 1 bypass cap means cheaper and the
CEO can more likely afford the Cadillac.
But we see the vast majority of PP amps made using the UL method, and
with many having no R between G2 and OPT taps.
The only way to avoid the *possible* horrid effects of high R between
G2 and either some fixed Wg2 supply OR some Vac source is to feed G2
by a direct coupled cathode follower, which means its anode must be
say 300V above the B+ for OP tubes, and then the its grid can be fed
from a high value pot strapped across an OP primary to vary the UL
tapping, without using the tap. One could also use a mosfet. By the
time one finishes the experiments one might conclude plain old UL
works fine, and that the doubt one had about learning anything useful
before using a CF like this was well founded, and no use to persue the
alternative and more costly/complex way to skin the poor little cat.
Patrick Turner.
[color=blue][i]
Cheers, John
------------------------------------------------------------
I managed just 5200 km on the bike for the year ending Apr 30/12, not bad for an old fart in his 80th year. A lot of that during the Winter which is fairly real here. That & many km on the XCSkis.
Didn't do much with toobz over the past year but the manual labor here on more than 2 hectares here keeps me buzy. Lots of fire wood to cut & split.
Dug out the access to pump the septic tank a couple of weeks ago. Its under the brick patio. So got the bricks are back in now so its on to the next thing. One of the tractors needs some fixing. Living in the country for more than 40 years brings lots of physical labour.
Still beats sitting inside with the computor!
Cheers to all, John
|