On Apr 20, 7:34*pm, David wrote:
Looking for Crystal Radio Receiver Circuits, that have a voltage
quadrupler to increase headphone volume without batteries or AC power.

Hi David - I think this idea will not work well for the reasons others
have already given - (a) the forward voltage drop of the diodes (2
instead of 1) which means that you must have an rf signal of around
1.2v instead of just 0.6v before you hear anything at all, and (b) the
question of the availability of power. A quadrupler will not increase
the very small amount of power picked up by the aerial.

However, all is not lost. Other enthusiasts have been at work with a
crystal set which uses a 'zero gate-threshold' FET as a form of
passive synchrodyne detector. The incoming signal opens and closes
the conducting channel of the fet at the same frequency as the signal
applied to the earphones through the channel (they are both the same
signal). This give synchronous mixing and hence detection of the rf
envelope. The advantage is that the fet has low and high impedances
in the right places (low channel impedance, high gate impedance) and
so is highly efficient.

Forgive the long posting, but here is something copied from the GQRP
Yahoo group, with references. Andy G4OEP

Hi to everyone on the list,

For those who have been following the "High sensitivity Crystal Set"
and "More ham xtal set DX heard tonight on 80m" threads on this list I
should like to share my recent experiences/observations using the
ALD110900A device as a synchronous detector/crystal receiver.

The high sensitivity crystal set was featured again in the current
Rad-Com (June 2007 edition, page 60) and the RSGB have made available
a limited number of the ALD110900A devices on a
first-come-first-served basis via RSGB sales. I obtained a small
number of the ALD110900A devices so I could compare the performance to
the 2N3819 jfet which I had been using in my version of the receiver
based on the design by David Cripe (NE4AM) in his article "Nostagia
For The Future" (Amateur Radio Today, Dec. 1995 pages 14-16) a copy of
which can be found on this web page...

http://members.wideband.net.au/gzimm...tors/FET1.html

The article referenced above was the inspiration for the more recent
article in the January edition of QST, a copy of the QST article
appears here...

www.arrl.org/qst/2007/01/culter.pdf

Observations/Comments:
I found the best results are obtained using the design described by
David Cripes (NE4AM) in his 1995 article. His use of two tapped tuned
circuits (band pass filter arangement) permits precise matching of
both the antenna and the load, both of which are vital for best
crystal set performance. In operation I found the receiver to perform
very well with resuls comparable to a BC receiver, seperation of
stations was readily achieved with good volume and loud enough for
this "deaf old coot" to be able to listen to news broadcasts etc and
understand what was being said. The antenna used is about 60 feet of
wire and a central heating system counterpoise. My receiver shares the
two 4 inch diameter coils used in the NE4AM design, it also shares the
jfet (2N3819) with gate bias as described by David in his article.

The June Rad-Com article does not seem to place enough emphasis on the
importance of matching the ALD110900A sorce/drain impedance and load
(headphones) to the tuned circuit. My opinion is that best results are
obtained with a good selection of tappings on the coil as described by
NE4AM. The ALD110900A (or jfet plus negative gate bias supply)
circuit demonstrates its superior performance over the OA81, OA91 etc
style detector circuits when detecting weaker signals. Signals which
are simply inaudible to the OAxx style detectors are rendered audible
by the ALD or jfet synchronous detector design. The ALD or jfet design
definately offer better performane for "crystal set DX for 80 Mtrs" or
other similar applications.

I note that in both the January QST and June Rad-Com articles they
used a low impedance telephone earpiece via an autotransformer for
matching. In the Rad-Com article they said that "an old pair of Hi-Z
headphones" was substituted for the low-Z earphone/autotransformer
with no noticable improvement. I can only conclude that the "old pair
of Hi-Z headphones" must have been very poor quality or had weak
magnets, in my tests substituting my pair of "old Hi-Z headphones"
into the circuit made a very significant difference. I tested several
low-Z telephone earpieces (all worked) but the Hi-Z phones (4000 Ohms)
consistently outperformed the low-Z units. Its also worth pointing out
that the difference was most notable on weaker signals. On strong
(local) signals the difference was much less obvious. I currently do
not have a crystal earphone to test with but I suspect a crystal
earphone would also outperform most low-Z earphones/matching
transformer configurations.

And finally:
I would strongly advse anyone who is thinking about experimenting with
this form of synchronous detector to first read David Cripes (NE4AM)
article as a starting point and also give serious consideration to
using a 2N3819 jfet plus adjustable negative gate bias supply in
preference to the ALD110900A device. The only significant advantage I
can see with the ALD device is that its zero Volts gate threshold
avoids the requirement for the seperate negative gate bias supply
required by the jfet arangement and thus saves a few components.

The RSGB are currently offering the ALD110900A devices for 4 UK pounds
each (including UK postage) but if you already have a 2N3819 (or
similar N-ch' jfet) in the junk box then try that first with an
adjustable negative gate supply and evaluate the circuit for yourself
before spending your 4 pounds. One additional advantage of the jfet
plus negative gate bias approach I have found is that by making the
bias adjustable (front panel control) you can improve
performance/sound quality on strong signals. The bias point for best
sensitivity and lowest distortion (on stronger signals) are not the
same so the ability to make bias adjustments "on-the-fly" is an
advantage.

Conclusion:
If you are building a crystal set for the "local" station(s) then I
see no real advantage in bothering with the syncronous detector
approach but if you want a crystal set with highest possible
performance then the syncronous detector plus tapped coils (for
matching) gives outstanding performance.

I wonder, has anyone else on this list experimented with the ALD
devices and if so what are your experiences?

73,s to all on the list.

Des (M0AYF G-QRP 9788)