On Mon, 15 Nov 2010 09:17:30 -0600, "amdx" wrote:

C shape laminated core
with a
small gap. The gap will be large enough that a small magnet will fit in it.
The magnet
is connected to a lever and on the other end is a diaphragm.

This doesn't sound like a Crystal radio project at all. You describe
nothing that comes close to even 1KOhm, much less 1MOhm in load. Your
descriptions all use appeals to sensitivity, not impedance.

Trying to maximum signal for contest situations want a longer antenna.

This confounds your desire for higher Z. In the extreme (antenna of
several wavelengths and necessarily close to ground) will be less than
1KOhm. In the mid-range, could be hi-Z IFF it is a halfwave long. In
the conventional lengths, some may pose a moderately hi-Z (maybe
KOhms). None will exhibit the Z you anticipate for your Tank.

As I said, start thinking backwards from the power delivered to your
ear. Can you express that as a number? Not much point in the rest of
this if you cannot.

73's
Richard Clark, KB7QHC

"Richard Clark" wrote in message
...
On Mon, 15 Nov 2010 09:17:30 -0600, "amdx" wrote:

C shape laminated core
with a
small gap. The gap will be large enough that a small magnet will fit in
it.
The magnet
is connected to a lever and on the other end is a diaphragm.

This doesn't sound like a Crystal radio project at all. You describe
nothing that comes close to even 1KOhm, much less 1MOhm in load. Your
descriptions all use appeals to sensitivity, not impedance.

Trying to maximum signal for contest situations want a longer antenna.

This confounds your desire for higher Z. In the extreme (antenna of
several wavelengths and necessarily close to ground) will be less than
1KOhm. In the mid-range, could be hi-Z IFF it is a halfwave long. In
the conventional lengths, some may pose a moderately hi-Z (maybe
KOhms). None will exhibit the Z you anticipate for your Tank.

As I said, start thinking backwards from the power delivered to your
ear. Can you express that as a number? Not much point in the rest of
this if you cannot.

73's
Richard Clark, KB7QHC

Hi Richard,
I have probably confused things, I have 4 or 5 threads running at this
time.

The C core EI core thing is a starting attempt to build a speaker with a
high impedance
to eliminate the losses of a matching transformer. The whole excercise is to
build a crystal
radio that will eack out the most sensitivity.

1 picowatt to the earphone is a good number.

For sensitivity the starting point has got to be the tank circuit, you
want to build
an inductor with very high Q and then mate that to a good quality capacitor.
A Q of 1000 is possible over much of the AM BCB.
Can we agree on that?

Now you need to couple in energy from an antenna. If this is adjusted for
maximum power transfer, we have reduced the Q by 1/2 or Q=500.
Assuming a 240uh inductor and frequency of 1 Mhz the XL is 1507 ohms,
multiply that by the antenna loaded tank Q of 500 and we have an Rp =
753,500 ohms.
Does that work for you?

I think I found a good site for the antenna matching;
http://www.crystal-radio.eu/enantunittest1.htm


Now we need to detect and tranfer the signal to a transducer

I'll stop here till I get some feedback, I don't know my question anymore? \
:-)
Mike.

amdx wrote:


The C core EI core thing is a starting attempt to build a speaker with a
high impedance
to eliminate the losses of a matching transformer. The whole excercise is
to build a crystal
radio that will eack out the most sensitivity.

1 picowatt to the earphone is a good number.

You may want to make sure this enough to be audible. An iPod may output 30
milliwatts which is probably more than you would ever need/get. 1 picowatt
to a headphone may not be sufficient.


For sensitivity the starting point has got to be the tank circuit, you
want to build
an inductor with very high Q and then mate that to a good quality
capacitor. A Q of 1000 is possible over much of the AM BCB.
Can we agree on that?

Have you considered what Q=1000 does to the bandwidth of the received
signal?



Now you need to couple in energy from an antenna. If this is adjusted for
maximum power transfer, we have reduced the Q by 1/2 or Q=500.
Assuming a 240uh inductor and frequency of 1 Mhz the XL is 1507 ohms,
multiply that by the antenna loaded tank Q of 500 and we have an Rp =
753,500 ohms.
Does that work for you?

Won't both the antenna and the load serve to reduce the Q? You need to be
looking at how power is transferred from the antenna to the earphone. Trying
to optimize one part at a time may not yield the best result.



I think I found a good site for the antenna matching;
http://www.crystal-radio.eu/enantunittest1.htm


Now we need to detect and tranfer the signal to a transducer

I'll stop here till I get some feedback, I don't know my question anymore?
\
:-)
Mike.

"joe" wrote in message
...
amdx wrote:


The C core EI core thing is a starting attempt to build a speaker with a
high impedance
to eliminate the losses of a matching transformer. The whole excercise is
to build a crystal
radio that will eack out the most sensitivity.

1 picowatt to the earphone is a good number.

You may want to make sure this enough to be audible. An iPod may output 30
milliwatts which is probably more than you would ever need/get. 1 picowatt
to a headphone may not be sufficient.

I can only go by what I have read on the crystal radio groups and they
say 1pw
is audable with very sensitive headphones..

For sensitivity the starting point has got to be the tank circuit, you
want to build
an inductor with very high Q and then mate that to a good quality
capacitor. A Q of 1000 is possible over much of the AM BCB.
Can we agree on that?

Have you considered what Q=1000 does to the bandwidth of the received
signal?

Ya, it would limit bandwidth. But to much Q is easy to solve and hard to
get.
That is only unloaded Q of the tank, adding the antenna brings that down to
Q=500.
Then comes the load from the detector and then the audio transducer
assembly.



Now you need to couple in energy from an antenna. If this is adjusted for
maximum power transfer, we have reduced the Q by 1/2 or Q=500.
Assuming a 240uh inductor and frequency of 1 Mhz the XL is 1507 ohms,
multiply that by the antenna loaded tank Q of 500 and we have an Rp =
753,500 ohms.
Does that work for you?

Won't both the antenna and the load serve to reduce the Q? You need to be
looking at how power is transferred from the antenna to the earphone.
Trying
to optimize one part at a time may not yield the best result.

Yes, absolutely. I think the optimization is in the tank circuit after
that
it is impedance matching.

MikeK


I think I found a good site for the antenna matching;
http://www.crystal-radio.eu/enantunittest1.htm


Now we need to detect and tranfer the signal to a transducer

I'll stop here till I get some feedback, I don't know my question
anymore?
\
:-)
Mike.

On Mon, 15 Nov 2010 17:56:16 -0600, "amdx" wrote:

I can only go by what I have read on the crystal radio groups and they
say 1pw
is audable with very sensitive headphones..

Hi Mike,

You should consider their being very good at this then. 1 pW is the
lowest limit of hearing at 1KHz. No speaker/transducer is 100%
efficient. So, automatically, this claim you have read is suspect in
the highest degree.

Let's work these numbers further. The transducer elements you have
been describing may be sensitive, but that is not the same thing as
efficient. Given that they are ancient magnet and diaphram
constructions, they would be pushing the limits at 10%.

Further, your chosen power level of 1pW would not be heard but in a
very special anechoic chamber, and even then your heart and your
breath would be in competition as QRN. Speach as perceived to be at
normal quiet talking levels would be 40dB more powerful (and, again,
in a quiet surrounding like a library). This would be a normal
expectation of program content sound level.

Being generous (good efficiency and whispers at 1 meter in the
library), I would suspect that your receive power level would be
closer to 1nW and this would be straining things for a real listening
experience.

For sensitivity the starting point has got to be the tank circuit, you
want to build
an inductor with very high Q and then mate that to a good quality
capacitor. A Q of 1000 is possible over much of the AM BCB.
Can we agree on that?

Have you considered what Q=1000 does to the bandwidth of the received
signal?

Ya, it would limit bandwidth. But to much Q is easy to solve and hard to
get.
That is only unloaded Q of the tank, adding the antenna brings that down to
Q=500.
Then comes the load from the detector and then the audio transducer
assembly.

In the middle of the band, it would give you poor telephone audio
quality. However, many hams find it suitable for DX work.


Now you need to couple in energy from an antenna. If this is adjusted for
maximum power transfer, we have reduced the Q by 1/2 or Q=500.
Assuming a 240uh inductor and frequency of 1 Mhz the XL is 1507 ohms,
multiply that by the antenna loaded tank Q of 500 and we have an Rp =
753,500 ohms.
Does that work for you?

Good enough to pitch back and forth, much as your power level.

73's
Richard Clark, KB7QHC

"Richard Clark" wrote in message
...
On Mon, 15 Nov 2010 17:56:16 -0600, "amdx" wrote:

I can only go by what I have read on the crystal radio groups and they
say 1pw
is audable with very sensitive headphones..

Hi Mike,

You should consider their being very good at this then. 1 pW is the
lowest limit of hearing at 1KHz. No speaker/transducer is 100%
efficient. So, automatically, this claim you have read is suspect in
the highest degree.


Hi Richard,
What do you think of this guys numbers and methodology?
He says he can hear .0078 pw with a Adastra Model: 952-207
http://www.crystal-radio.eu/enluidsprekertest.htm
MikeK


73's
Richard Clark, KB7QHC

On Mon, 15 Nov 2010 19:22:37 -0600, "amdx" wrote:

What do you think of this guys numbers and methodology?
He says he can hear .0078 pw with a Adastra Model: 952-207
http://www.crystal-radio.eu/enluidsprekertest.htm

Hi Mike,

Where did the search for hi-Z go when this 16Ohm speaker was hauled
out for listening? OK, sure, it is all a matter of making a match - I
can go with that.

Let's do the math and see where that leads us for the specification
offered:
SPL @ 1W/1m: 112.5dB

When driven by .0078 pw we find ourselves 140dB below the 1 Watt that
yields 112.5 dB SPL heard at 1 meter. That translates to -27.5dB re
the absolute lowest level of hearing.

OK, supposing you are not 1 meter away from that speaker? I can well
anticipate that you would expect the stethoscope lead comes in to
rescue this claim. Does it get us to within 1mM of the cone to make
up the difference? Your ear can not get that close (maybe a cM) and
the volume of air in the tube makes it worse (unless we are using an
Hemholz resonator, and at that, the program material goes out the
window).

Being generous and saying the claim is off by 1 decimal place still
has us sitting in an anechoic chamber. No one has that kind of bucks
for a hobby pursuit except Bill Gates. Even then, this is about the
threshold of hearing for a juvenile. Is your scribbler 17 years old?
I can well imagine you, like myself, even that age out - 3 to 4 times
over. Program content is going to depress these readings by roughly
5dB for age and another 5 to 10dB for frequency variation.

If you want to copy 1WPM CW at 1KHz, this may fly (if you are buried
alone in a cave in South America). Who transmits A3 modulated CW
(yes, a contradiction in acronyms where CW commonly means morse code)
these days?

So, on the commonsense side of this, no that myth is busted.

The author explores efficiency and states:
The efficiency is 7.03µW / 56.8µW = 0.123
Which was my generous offering in an earlier posting (however, the
author stipulates this is a total conversion efficiency for both
speakers).

Going further we observe:
The efficiency is quite varying with different frequencies, at 1 kHz there was a peak.
At other frequencies the efficiency is lower.
This can be caused by resonances in the speakers, because this situation with two speakers connected is quite different from the normal use.
Normal use indeed (what I call listening to program content). There
is every chance that the coupled speakers were driven at a hemholz
resonance. Using the scope probe as a crude ruler, the volume of air
looks to be close to a half wave long.

Note the leading stipulation again:
The efficiency is quite varying with different frequencies
.... indeed.

I have had a hearing test in specially designed chambers, employing a
test that eliminates guessing when the sound is, or is not there. I've
even designed testing systems that use that methodology for measuring
Army helicopter pilot alertness. The psychological pressure of
expecting to hear a faint sound can drive results that are impossible
to replicate without that testing protocol.

Think you could follow the chain of reasoning here to cross-check the
other transducers' performance? If it is on par, then you can trust
the testing methodology. If my back-of-the-napkin calculations are
off, this will reveal it.

73's
Richard Clark, KB7QHC

Hello Mike,


On 15 nov, 22:54, "amdx" wrote:
"Richard Clark" wrote in message

...



On Mon, 15 Nov 2010 09:17:30 -0600, "amdx" wrote:

C shape laminated core
with a
small gap. The gap will be large enough that a small magnet will fit in
it.
The magnet
is connected to a lever and on the other end is a diaphragm.

This doesn't sound like a Crystal radio project at all. *You describe
nothing that comes close to even 1KOhm, much less 1MOhm in load. *Your
descriptions all use appeals to sensitivity, not impedance.

*Trying to maximum signal for contest situations want a longer antenna.

This confounds your desire for higher Z. *In the extreme (antenna of
several wavelengths and necessarily close to ground) will be less than
1KOhm. *In the mid-range, could be hi-Z IFF it is a halfwave long. *In
the conventional lengths, some may pose a moderately hi-Z (maybe
KOhms). *None will exhibit the Z you anticipate for your Tank.

As I said, start thinking backwards from the power delivered to your
ear. *Can you express that as a number? *Not much point in the rest of
this if you cannot.

73's
Richard Clark, KB7QHC

* Hi Richard,
*I have probably confused things, I have 4 or 5 threads running at this
time.

*The C core EI core thing is a starting attempt to build a speaker with a
high impedance
to eliminate the losses of a matching transformer. The whole excercise is to
build a crystal
radio that will eack out the most sensitivity.

*1 picowatt to the earphone is a good number.

* For sensitivity the starting point has got to be the tank circuit, you
want to build
an inductor with very high Q and then mate that to a good quality capacitor.
A Q of 1000 is possible over much of the AM BCB.
* Can we agree on that?

Now you need to couple in energy from an antenna. If this is adjusted for
maximum power transfer, we have reduced the Q by 1/2 or Q=500.
Assuming a 240uh inductor and frequency of 1 Mhz the XL is 1507 ohms,
multiply that by the antenna loaded tank Q of 500 and we have an Rp =
753,500 ohms.
*Does that work for you?

I think I found a good site for the antenna matching;http://www.crystal-radio.eu/enantunittest1.htm

Now we need to detect and tranfer the signal to a transducer

The transducer will be the difficult part (the antenna is quite simple
compared to this). As you may know, below 150mVRFpk across the
junction, rectified output voltage drops quadratically with RF input
voltage, hence detection efficiency.

So to get maximum voltage across the junction, you need a high
impedance rectifier (that means low "Is"). Disadvantage of this is
you need a transducer with same (very high) impedance (as I assume
you don't want to use electronic LF amplification). I did my
experiments with LF electronic amplification.

If you can't find / make one in the several hundred kOhms range, you
will probably need to use rectifiers with higher Is. If so, you also
need to transform the RF impedance to a lower value to get best RF
power transfer to the rectifier.

Regarding diodes, years ago I did experiments around 7 MHz with tuned
detectors where the diode capacitance is no longer small with respect
to the tuning capacitance.

When using hybrid schottky rectifiers (that are the ones with relative
high reverse blocking voltage), strange hysteresis effects occurred in
the DCout versus RFinput voltage curve. When using real schottky
rectifiers (like BAT15, 14, etc), this effect wasn't present. I didn't
document it (only some notes), so I can't share the full details with
you.


I'll stop here till I get some feedback, I don't know my question anymore? \
:-)
* * * * * * * * * * * * * * * * * * *Mike.


Good luck finding/designing the best transducer,


Wim
PA3DJS
www.tetech.nl
without abc, PM will reach me, very likely.