John Byrns wrote:

In article , Patrick Turner
wrote:

John Byrns wrote:

You have still haven't enlightened us with some concrete information about
how much, if at all, your biased diode detector really helps reduce the
distortion of the diode peak envelope detector.

It should be *obvious* from the circuit!

It's not, at least not to those among us, such as myself, who are not so clever.

My circuit is as simple as it gets.

See where the DC flows in my circuit, even without any IF signal present.


Hve the cathode of the CF at +50v, and have a 1M R to drain 0.05mA
through the diode. Much more current could be used.

This method means that detection of weak signal lower than the forward voltage
of the Ge diode of 0.27v peak approx are not subject to the non linear turn on
of the diode, ie, there is no clipping by the diode.

Is what you are saying is that the diode in your circuit is always "turned on"?

Indeed it is turned on all the time, and this stops the huge variation in 455 kHz
ripple
voltage with low level signals when the IF signal is very low, and the R of the RC
detector circuit is just taken to ground from a C which has a very low charge in it
when the signal is very low, as it is in a conventional circuit.



Quantity not quality was what dominated radios in old days.

If you wanted better sound, you bought a Quad AM tuner, which only
rich folks could afford.

I have one of those "rich folks" Quads right here next to my computer, and
I guess those "rich folks" got taken as the Quad uses a common vacuum
diode detector, like a common kitchen radio, nothing special, the only
special care taken in the detector design seems to be that the audio
output is tapped way down on the diode load resistor to minimize negative
peak clipping.

I don't recall what Quad have in their box, but on strong reception, but the mainly
conventional
circuit within does work fairly well enough for many people.
I like things to work better than "conventional", because this stands for lowest
common
denominator, which is plain substandard, imho.


You could better make your point if you posted a couple of graphs for
distortion vs. signal level for a diode detector, with and without bias,
and for several modulation levels, maybe 80% and 100%.

My biased SS diode has lowest thd at high levels of signal.

One would expect you to be an advocate of your on circuit, but how low is
the thd at high modulation levels?

Very low, compared to many other circuits.

Its difficult to measure because other thd effects swamp the thd of the detector.

To measure the accuracy of the detector, you need an AM signal
from a generator that itself has less than 0.01 % of thd in the envelope shape.

If an AM signal at say 1 MHz is applied to the RF input of a set,
then the mixer and IF tubes will impose some thd, and finally the detector
then adds its thd.
But from tests using a dual trace CRO with an input signal from a 455 kHz source,
so that the recovered audio is overlaid onto the wave form of the envelope,
I found it was impossible to see any thd with my detector idea up to several volts,
with any level of % modulation.
This method of inspecting the detector performance removes the need to
have a low distortion enevelope shape to begin with.

I might add that the CF buffer used for the detector was a 12AU7,
and its thd would be around 0.1% at 10 vrms output, simply
because of the thd of such a tube arranged the way it is.
If a 12AT7 were used, the open loop gain is higher, and the gain reduction much
more than 12AU7, so the thd would be perhaps 0.03% at 10v output.


But if you have a normally gronded last IF coil feeding a diode to 100pF
with a typical 1M discharge R to ground, then with low signals on weak
stations,
the thd is appalling, and I thought such issues would be obvious to anyone
familiar with diode operation. No need for me to copy out
the wave form analysis I did, build a detector like I have and you won't be
dissapointed!

Actually your wave form analysis would be very interesting to see indeed.

I drew up the wave forms I thought would appear, and sure enough they did appear
when I tested a typical circuit.

At very low signal levels, or where the % modulation approaches 100%, the 455 kHz
ripple voltage is very low, and when the 455 kHz signal is stringer, the ripple
voltage becomes much higher.
There is distortion in these goings on if the diodes Ra changes a lot, which is
does,
so a germanium diode which is turned on at all times by a DC flow prevents the
variations
one sees with a tube diode.



The impedance of the IFT is quite low at the harmonic frequencies in the
diode current wave form so the diode sees a low impedance drive at those
frequencies even without a cathode follower, and the filtering effect of
the IFT prevents the harmonics in the current pulses from reaching the
plate of the IF amplifier.

And if a CF buffer is used, the low impedance caused by diode current flows
are not seen by the IFT and IF pentode amp, so no distortion occurs.
The CF removes the thd caused by the diodes.

See the discussion on page 1077 of the RDH4
relative to loading of the IFT secondary by the diode,although much more
comprehensive discussions can be found elsewhere.

Indeed more exists elsewhere, but after thought about the problems,
I went my own way.

Patrick Turner.



Regards,

John Byrns

Surf my web pages at, http://users.rcn.com/jbyrns/