On Jul 10, 3:48*am, John Byrns wrote:
In article ,
*Patrick Turner wrote:
On Jul 9, 9:19*am, John Byrns wrote:
[Much off topic drivel snipped]
You may well delete "drivel" like so many others here who have awful
personalities who cannot cope with being human or nice in any way. Ah
you going senile? Just can't cope with well rounded discussions? Its
OK, you don't have to hide such characteristics.
Your drivel is a fine topic for discussion, the problem is that it is off topic
in this group and you should take it elsewhere.
I see no reason not to include some background information which
indirectly relates to the subject.
Otherwise the group's discussion can be very boring, especially since
all this has probably been discussed before.
I've read the tonnesoftware.com site before and there's nothing there
that works better than my circuits using tubes, and I got less wave
form distortion that they show on their oscillograms.
The detectors used in the modulation monitors, the General Radio 1931, Gates
M5693, and Belar AMM-2/3 don't appear to have any visible waveform distortion,
although your eye is probably more sensitive to this type of aberration than
mine. *In my defense, back in the days before I became senile, when I worked in
the radio design factory, where we were designing "white goods" radios and
measuring power output at the 5% distortion level, I was consistently able to
set the signal level where 5% distortion occurred without reading the meter on
the distortion analyzer, looking only at the waveform displayed on the CRO.
Don't worry, when, or if I become senile, i'll not recognise 5% ona
CRO when its there. But most waves at that site showed slew
distortions and cut off distortions and so forth, all of which can
easily be avoided or minimised to be below 1% with my circuit over a
wide AF range and if the output voltage 1 Vrms. But even where one
listens to short wave where AF detector output 0.1Vrms, the sound is
good, although rather mauled by having travelled so far and being so
riddles with noise and fading up and down. I have never ever seen any
old radio or any old schematic of what was used for detecting AF for
use in re-broadcasting. AFAIK, not one single commercial example
exists of a radio with "infinite impedance detector" even though the
Selsted & Smith example is given in RDH4, page 1,495, Figs 27.56,
27.57
RDH4 does not have anything that works as well as what I invented for
myself 14 years ago. Maybe someone else invented it elsewhere, but
almost anything is better than the normal bean counter driven way of
AM detection in old radios.
I have two main problems with the tonnesoft web page. *The first is that he uses
an ideal diode model for some of his simulations, and doesn't say what model he
used in others, although I haven't looked at the web page in two years, and if I
read it again might find that he does specify the diode models he used.
I don't care about his modelling or what diodes he used; I just build
my REAL stuff and test it. If it measures well, and better than all
the websites I've seen online, they have wasted their time.
The second problem, which is very relevant to our disagreement over the use of a
cathode follower to buffer the diode from the IFT, is that he assumes that the
demodulators are driven by a zero impedance, effectively all his circuits
include your cathode follower, which confuses the issue.
There is no need to load the secondary of the IFT and one may indeed
connect it to an "infinite impedance" and it will still act like a
tuned transformer as intended. But one may load the IFT. No laws
against that. But loading can't be low, and so I have found its better
to convert the IF signal to low impedance with the CF and then the
diode is not so critical. Point contact GE types with high reverse
voltage ratings are probably best to use, forward voltage drop of only
0.2V.
I suggest you try my circuit rather than waste time trying to find
excuses not to use your soldering iron.
I probably shouldn't even be thinking about this cathode follower issue, let
alone thinking about soldering together your detector circuit. *I should be
saving my solder for building an improved 25L6 amplifier, however I am curious
about the effect of your apparently pointless cathode follower, and it is
possible that I will be overcome by a desire to show it to be worthless. *You
seem to possess a reverse "bean counter" element to your personality which
drives you to include circuit elements that serve no real function, thereby
needlessly driving costs up and reliability down.
Most bean counters justify their employment by being able to reduce
the parts and labour needed to make something, and therefore
increasing shareholder profits and most often reducing the sound
quality and reliability in electronics produced by the company.
I do indeed have an anti bean counter attitude, and most of them
should be frog-marched in public through the town square to where they
should be given a pick and shovel to dig their own grave. Where is Pol
Pot when you need the man? But I am not so extreme in practice, and
the radios and amps I build are amoung the best sound and most
reliable on the planet.
You can set out on a wanton voyage of defiance of good sense if you
like, but meanwhile I have clients I like to please and who pay me to
put in the stuff that the companies left out all those years ago.
I don't really mind if you are rivetted to your lounge room chair in
front of your PC, but when I ever about an idea which might be good I
rush out and try it out to see if I can replicate claims by the
inventor.
OK, what exactly is your claim for inserting a cathode follower between the
secondary of the IFT and the detector diode? *On your web page you make two
claims.
Claim #1 is that "this CF isolate the loading effects of the diode detector from
the secondary of the 6BA6", however you don't say why this isolation might be
desirable? *The only reason I can see is to increase the Q of the secondary of
the 6BA6, however increasing the Q doesn't seem to be your goal as you also talk
about adding 100k damping resistors to reduce the Q. *You could save the cost of
the cathode follower and the loading resistor by simply using neither.
OK, RDH4 talks about the ill-effects of diode + R&C detection loading
and although the book was written to promote tube use it didn't go far
enough as I have in this area. Loading should be resistive only, and
then you buffer to the following processes.
Claim #2 is that the cathode follower "provides a low impedance output to drive
the 1N914 silicon diode detector." *Again you give no hint why this might be
useful, although it does seems to have some potential usefulness, on the other
hand a higher impedance output also has some advantages in detector
design/operation.
Si diodes for signal detection at highish F should be driven by low Z
sources lest you get all manner of distortions you don't want. Just
build and measure, and you'll be pleased. If you just won't try such a
tiny little idea with just a tiny handful of parts to do something
that is so simple because its veracity cannot be proved in advance in
words here at this group, then you do indeed possess a moribund mind
and one possessing a high amount of sloth.
If you can't manage to build my detector circuit which uses 1 x 12AU7
tube with a few R&C then maybe you've lost your touch.
Looking at the schematic of your detector it appears to use only 1/2 x 12AU7 for
the cathode follower between the secondary of the IFT and the diode, your design
doesn't really use a cathode follower after the diode, instead depending on a
voltage divider, 330k/100k, in conjunction with a relatively high AC coupled
load of 890k, to minimize negative peak clipping. *You do use cathode followers
as the audio stage, but they are not connected in a way that they have any
influence on negative peak clipping.
Your design has given me one idea that I have missed up until now. *My design
uses a cathode follower directly coupled to the diode detector to eliminate
negative peak clipping. *The down side of this approach for those of us who have
been influenced by bean counters is that it requires a negative supply for the
cathode of the cathode follower, impacting both cost and reliability to some
extent. *On the plus side one advantage of this approach is that the AGC voltage
can be easily obtained from a tap a little way down the cathode resistor of the
cathode follower.
The idea you have given me is to float the secondary of the IFT and the diode at
a positive potential allowing the direct coupled cathode follower to operate
without a negative supply. *The complication is that a separate AGC rectifier is
required, which has the potential to negatively affect the sound, especially
when a delay circuit is used.
The generation of AGC voltage may be done easily and off the CF output
with a cap and diode with its cathode grounded, with the generated -
Vdc drained off with 1.5M to bias points. It can also be done from the
anode of the IF amp and thus from the primary of IFT. C used is
usually 33pF, and loading effect is negligible.
The anode IF signal is higher than secondary do more AGC is available
and usually there is less sibilance during tuning when the anode is
used for AGC dervivation.
The rules about invention are anything but strict; just do what works
better than you read about in text books or that you see being used in
commercial junk designed bt accountants.
And I have extensively discussed all this stuff before with you here
at r.a.t and I refuse to do it all over again.
Discussed it yes, unfortunately you have never offered an explanation of the
usefulness of the cathode follower between the IFT and the detector diode,
beyond the subjective claim that it provides the "best sound". *Near as I can
tell the only purpose this cathode follower serves is to assuage your prejudices
against "bean counters".
I give a damn if all I've said misses the point asfayac. Those capable
of invention and who insist that AM radios and audio amps should sound
better and measure better than much of the dumbed down junk that's
sold in shops will understand my website and the attitudes I have.
Patrick Turner.
--
Regards,
John Byrns
Surf my web pages at, *http://fmamradios.com/