<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>If a crystal set can pump milliwatts of electrical power into a
>headphone, something very similar should be able to light up an LED.
>No mixed fruits are involved.
<jfie...@austininstruments.com> wrote:
>On Mon, 08 Oct 2012 07:12:06 -0700, John Larkin
><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>If a crystal set can pump milliwatts of electrical power into a
>>headphone, something very similar should be able to light up an LED.
>>No mixed fruits are involved.
>---
>So build one.
I might some day. I already have a lithium-battery night light that I
built, glowing on the bookshelf near my bed. I figure it will last
20-30 years, always on. It has an interesting hands-free variable
brightness feature, where I can find it by its glow, but crank it up
as needed for finding my way around, after an earthquake or whatever.
Regular LED flashlights should do that, namely glow a little all the
time, so you can find them in the dark.
Given the energy density of lithium batteries, "energy harvesting"
rarely makes sense.
> "Billyb97113" <billyb97...@yahoo.com> wrote in message
> news:k4dohb$at9$1@dont-email.me...
>> To quote John Larkins:
>> "More practical would be to get usable LED lighting from ambient RF"
>> I tried several different types of diodes and Ferrite rods and coils and
>> could never get a measurable voltage across a .01 uf cap. I used a 10X
>> scope probe as the only load. I would be happy if the LED occasionally
>> blinked using any of the inductive circuits found on the internet.
>> Any ideas?? -bill
> People apparently don't build crystal sets any more,
> On Mon, 08 Oct 2012 10:11:36 -0500, John Fields
> <jfie...@austininstruments.com> wrote:
>> On Mon, 08 Oct 2012 07:12:06 -0700, John Larkin
>> <jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>> If a crystal set can pump milliwatts of electrical power into a
>>> headphone, something very similar should be able to light up an LED.
>>> No mixed fruits are involved.
>> ---
>> So build one.
> I might some day. I already have a lithium-battery night light that I
> built, glowing on the bookshelf near my bed. I figure it will last
> 20-30 years, always on. It has an interesting hands-free variable
> brightness feature, where I can find it by its glow, but crank it up
> as needed for finding my way around, after an earthquake or whatever.
> Regular LED flashlights should do that, namely glow a little all the
> time, so you can find them in the dark.
> Given the energy density of lithium batteries, "energy harvesting"
> rarely makes sense.
The first LED flashlight I ever bought was like that. It had a faint
glow so you could find it in the dark.
It came in handy, I bought it at the Orlando Hamfest, on the way home that night I had a flat tire, my kids had been playing with the light and didn't know where it was. I turned off the lights in the van and found it glowing under the rear seat.
The flashlight has slightly larger than the 9 volt battery that powered it. It had a pretty developed surface mount circuit that developed 3 different intensities for the LED.
Mikek
<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>If a crystal set can pump milliwatts of electrical power into a
>headphone, something very similar should be able to light up an LED.
>No mixed fruits are involved.
The 30 to 100pF zero bias capacitance of an LED is going to look like
a short circuit compared to the less than 1pF capacitance of a decent
detector diode. However, if the capacitance is known, it can be
resonated with a series inductor to light up the LED:
<http://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-...>
Here's an RF energy scavenger experiment that successfully lit up an
LED (Fig 15) using a voltage multiplier:
<http://www.hindawi.com/journals/apec/2010/591640/>
However, he cheated and used a nearby transmitter to power the device.
Still, the use of a voltage multiplier might be useful.
I don't think a crystal radio does not deliver "milliwatts" to the
headphones. See:
<http://www.crystal-radio.eu/enluidsprekertest.htm>
which offers a table of minimal earphone power levels commonly used on
crystal sets. Operating levels will be somewhat higher.
Headphone Sennheiser Model HD433 9.6 pW
Headphone Sennheiser Model HD330 0.78 pW
2x 2000 Ohm headphone Telefunken EH333 0.022 pW
2x 2000 Ohm headphone Omega 0.033 pW
Crystal earplug "Taiwan" 0.13 pW
Driver unit Adastra Model: 952-207 0.0078 pW
Looks more like fractions of a pico watt, which is what I would expect
from the signal levels found on a crude antenna.
At the 2.5mV/m contour a fair size dipole antenna will pickup about
100uV into 75 ohms.
<http://www.nd2x.net/calculators/FS.html>
Notes:
1. Plenty of typo errors on this page, but the numbers seem correct.
2. A simple dipole still has 0dBd gain down to about 1/10th
wavelength. The impedance becomes small, but the gain remains at
about 0dBd.
From the above calculator:
Receive-Power = E^2/R = (100*10^-6)^2 / 75 = 133*10^-12 watts = 133 picowatts
That's not anywhere near enough power needed by the LED (2.8
milliwatts). However, it's more than enough for the <1 picowatt
required to drive a crystal radio earphone.
<http://www.crystal-radio.eu/enluidsprekertest.htm>
The impedance of the LED is 1.4V / 0.002A = 700 ohms. 30pF of diode
capacitance at 1MHz is about 5K ohms which isn't going to have a big
impact on the 700 ohms. I'll call it 750 ohms because it makes the
numbers come out neatly. A tapped coil resonant at about 1MHz should
suffice for matching the 75 ohm antenna to the 750 ohm LED. Turns
ratio is:
sqrt(750/75) = 3.2
With 100uV at the antenna, the coil will deliver an inadequate 320uV
to the LED. The LED needs 1.4V, not 320uV.
On Mon, 08 Oct 2012 12:39:52 -0700, Jeff Liebermann <je...@cruzio.com>
wrote:
>On Mon, 08 Oct 2012 07:12:06 -0700, John Larkin
><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>If a crystal set can pump milliwatts of electrical power into a
>>headphone, something very similar should be able to light up an LED.
>>No mixed fruits are involved.
>The 30 to 100pF zero bias capacitance of an LED is going to look like
>a short circuit compared to the less than 1pF capacitance of a decent
>detector diode. However, if the capacitance is known, it can be
>resonated with a series inductor to light up the LED:
><http://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-...>
Why not put the LED right across the crystal set LC tank? The tank
needs capacitance anyhow. We have some high-efficiency LEDs in the low
10s of pF. The classic crystal set cap was a 365 pF variable.
>Here's an RF energy scavenger experiment that successfully lit up an
>LED (Fig 15) using a voltage multiplier:
><http://www.hindawi.com/journals/apec/2010/591640/>
>However, he cheated and used a nearby transmitter to power the device.
>Still, the use of a voltage multiplier might be useful.
>I don't think a crystal radio does not deliver "milliwatts" to the
>headphones. See:
><http://www.crystal-radio.eu/enluidsprekertest.htm>
>which offers a table of minimal earphone power levels commonly used on
>crystal sets. Operating levels will be somewhat higher.
> Headphone Sennheiser Model HD433 9.6 pW
> Headphone Sennheiser Model HD330 0.78 pW
> 2x 2000 Ohm headphone Telefunken EH333 0.022 pW
> 2x 2000 Ohm headphone Omega 0.033 pW
> Crystal earplug "Taiwan" 0.13 pW
> Driver unit Adastra Model: 952-207 0.0078 pW
>Looks more like fractions of a pico watt, which is what I would expect
>from the signal levels found on a crude antenna.
Those are audibility threshold powers. They say nothing about how much
power a crystal set can deliver.
>Here's the consensus on the original question:
>"Crystal radio to power an LED?"
><http://answers.yahoo.com/question/index?qid=20071201183427AAwom9C>
>Accumulating the scavenged energy and using it to build an LED flasher
>might be a good way to make it work.
Seems silly. A good 50 KW AM station might provide 100 uw/sq meter
roughly five miles away. Collect some of that with a longwire antenna.
1 uw is enough to light a good led visibly in room light.
>>I wasn't thinking electrostatics (hard to get DC from) but picking up
>>energy from an AM station.
>Well, let's do the math.
>An LED will barely light at 1.4VDC and 2ma.
> 1.4V * 0.002A = 2.8mw
>That's how much power needs to be produced by this contrivance.
The Osram LEDs that we use are often too bright at 2 mA. They are
clearly on in normal room lighting at 1 uA. Dark adapted, under
optimal conditions, I could just barely make out light at about 800
picoamps.
My night light is an Avago green LED, a Tadiran lithium cell, and a 1
meg resistor. It should last 20 or 30 years.
>At the 2.5mV/m contour a fair size dipole antenna will pickup about
>100uV into 75 ohms.
><http://www.nd2x.net/calculators/FS.html>
>Notes:
>1. Plenty of typo errors on this page, but the numbers seem correct.
>2. A simple dipole still has 0dBd gain down to about 1/10th
>wavelength. The impedance becomes small, but the gain remains at
>about 0dBd.
>That's not anywhere near enough power needed by the LED (2.8
>milliwatts). However, it's more than enough for the <1 picowatt
>required to drive a crystal radio earphone.
><http://www.crystal-radio.eu/enluidsprekertest.htm>
>The impedance of the LED is 1.4V / 0.002A = 700 ohms. 30pF of diode
>capacitance at 1MHz is about 5K ohms which isn't going to have a big
>impact on the 700 ohms. I'll call it 750 ohms because it makes the
>numbers come out neatly. A tapped coil resonant at about 1MHz should
>suffice for matching the 75 ohm antenna to the 750 ohm LED. Turns
>ratio is:
> sqrt(750/75) = 3.2
>With 100uV at the antenna, the coil will deliver an inadequate 320uV
>to the LED. The LED needs 1.4V, not 320uV.
According to my good'ol red Radiotron book, one might expect roughly
100 uW/sq meter a few miles from a 50KW AM station. You'd only need a
couple of uW to light an LED. So it's just an impedance matching
problem. It looks doable.
On Tue, 02 Oct 2012 02:29:39 -0700, miso <m...@sushi.com> wrote:
>I vaguely remember something about fluorescent tubes glowing near TV >transmitters. Maybe Sutro tower.
I used to drive past Sutro Tower every day in my ratty old Ford
Fiesta. The speakers would make awful sounds, even with the radio off.
I guess the output transistors were rectifying the RF picked up by the
speaker wires.
Jan Panteltje <pNaonStpealm...@yahoo.com> wrote:
> On a sunny day (Tue, 02 Oct 2012 02:29:39 -0700) it happened miso
> <m...@sushi.com> wrote in <k4ec7i$ha...@speranza.aioe.org>:
>> I vaguely remember something about fluorescent tubes glowing near TV >> transmitters. Maybe Sutro tower.
> Fluorescent tubes also glow when stuck in the ground under a HV power line.
> I used to walk around with a glowing neon next ot my 7 MHz xmtr antenna.
I need to try something like that up the road, where there are no houses,
but feed from power station goes through. Three phase might screw it up.
On Mon, 08 Oct 2012 10:24:49 +0300, upsided...@downunder.com wrote:
>On Sun, 07 Oct 2012 17:21:20 -0700, John Larkin
><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>On Sun, 07 Oct 2012 16:55:51 -0700, Jim Thompson
>><To-Email-Use-The-Envelope-I...@On-My-Web-Site.com> wrote:
>>>On Sun, 07 Oct 2012 16:42:36 -0700, John Larkin
>>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>>On Sun, 07 Oct 2012 15:28:35 -0700, Jim Thompson
>>>><To-Email-Use-The-Envelope-I...@On-My-Web-Site.com> wrote:
>>>>>On Sun, 07 Oct 2012 17:10:59 -0500, John Fields
>>>>><jfie...@austininstruments.com> wrote:
>>>>>>On Tue, 02 Oct 2012 09:12:59 -0700, John Larkin
>>>>>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>>>>>On Mon, 1 Oct 2012 20:54:20 -0700, "Billyb97113"
>>>>>>><billyb97...@yahoo.com> wrote:
>>>>>>>>To quote John Larkins:
>>>>>>>>"More practical would be to get usable LED lighting from ambient RF"
>>>>>>>>I tried several different types of diodes and Ferrite rods and coils and >>>>>>>>could never get a measurable voltage across a .01 uf cap. I used a 10X scope >>>>>>>>probe as the only load. I would be happy if the LED occasionally blinked >>>>>>>>using any of the inductive circuits found on the internet.
>>>>>>>>Any ideas?? -bill
>>>>>>>I have no idea how much power can be extracted from a ferrite rod
>>>>>>>antenna.
>>>>>>---
>>>>>>Why, then, is that admission of ignorance germane?
>>>>>>---
>>>>>[snip]
>>>>>Larkin took another crap, and his hat covered his eyes ?:-)
>>>>> ...Jim Thompson
>>>>You keep repeating the same childish attempts at insults, and you
>>>>can't post circuits that work. Sounds like senility to me.
>>>You have no idea of "effective area" ??:-)
>>More word salad.
>The effective area of a half wave dipole is about 0.12 square
>wavelengths, thus at 1 MHz with 300 m wavelength, the capture area is
>about 10000 m² or 1 ha.
So at 100 uw/sq meter (a few miles from a 50 KW AM station) that
computes to 1 watt! Of course, that's a pretty big antenna.
>The typical ferrite rod antenna has a gain of -30 to -60 dB below the
>full size dipole, thus the effective area is somewhere between 10 m²
>and 1 dm². Multiply this by the local power density [W/m²] and you get
>a ballpark figure of the available power.
Even the low end 10 m^2 times 100 uw/m^2 is a milliwatt! A good LED is
visible at a couple of microwatts, so even a ferrite rod antenna
should visibly light up a good LED a few miles from an AM station.
John Fields <jfie...@austininstruments.com> wrote:
> On Tue, 02 Oct 2012 09:12:59 -0700, John Larkin
> <jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>> On Mon, 1 Oct 2012 20:54:20 -0700, "Billyb97113"
>> <billyb97...@yahoo.com> wrote:
>>> To quote John Larkins:
>>> "More practical would be to get usable LED lighting from ambient RF"
>>> I tried several different types of diodes and Ferrite rods and coils and >>> could never get a measurable voltage across a .01 uf cap. I used a 10X scope >>> probe as the only load. I would be happy if the LED occasionally blinked >>> using any of the inductive circuits found on the internet.
>>> Any ideas?? -bill
>> I have no idea how much power can be extracted from a ferrite rod
>> antenna.
> ---
> Why, then, is that admission of ignorance germane?
> ---
>> Googling produces vague results.
> ---
> GIGO?
> ---
I was just comparing results googling vs blekko. Google just returns me to
here.
I'm just trying to learn blekko. No ehow wiki on blekko.
gregz <ze...@comcast.net> wrote:
> John Fields <jfie...@austininstruments.com> wrote:
>> On Tue, 02 Oct 2012 09:12:59 -0700, John Larkin
>> <jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>> On Mon, 1 Oct 2012 20:54:20 -0700, "Billyb97113"
>>> <billyb97...@yahoo.com> wrote:
>>>> To quote John Larkins:
>>>> "More practical would be to get usable LED lighting from ambient RF"
>>>> I tried several different types of diodes and Ferrite rods and coils and >>>> could never get a measurable voltage across a .01 uf cap. I used a 10X scope >>>> probe as the only load. I would be happy if the LED occasionally blinked >>>> using any of the inductive circuits found on the internet.
>>>> Any ideas?? -bill
>>> I have no idea how much power can be extracted from a ferrite rod
>>> antenna.
>> ---
>> Why, then, is that admission of ignorance germane?
>> ---
>>> Googling produces vague results.
>> ---
>> GIGO?
>> ---
> I was just comparing results googling vs blekko. Google just returns me to
> here.
> I'm just trying to learn blekko. No ehow wiki on blekko.
>>>I wasn't thinking electrostatics (hard to get DC from) but picking up
>>>energy from an AM station.
>>Well, let's do the math.
>>An LED will barely light at 1.4VDC and 2ma.
>> 1.4V * 0.002A = 2.8mw
>>That's how much power needs to be produced by this contrivance.
>The Osram LEDs that we use are often too bright at 2 mA. They are
>clearly on in normal room lighting at 1 uA. Dark adapted, under
>optimal conditions, I could just barely make out light at about 800
>picoamps.
Ok, let's use your numbers and see what happens.
I didn't know LED's would work at 1uA. Do you have the Osram part
number handy? I found some "low current" LED's on the Osram web pile,
but the spec sheets were for 2ma.
For the LED:
1.4VDC * 1uA = 1.4 microwatts
1.4VDC / 1uA = 1.4M load impedance
>According to my good'ol red Radiotron book, one might expect roughly
>100 uW/sq meter a few miles from a 50KW AM station. You'd only need a
>couple of uW to light an LED. So it's just an impedance matching
>problem. It looks doable.
The ambient level in US metro areas is about 50 uWatts/sq-meter from a
study that I can't seem to find.
For the antenna:
100 uWatts/sq-meter is the energy density. To convert into detected
energy, the effective aperture of the receive antenna will be needed:
<http://vk1od.net/antenna/concepts/Ae.htm>
<http://vk1od.net/software/fsc/index.htm>
For a 1Mhz dipole (143 meters long), that would be about 7000
sq-meters effective aperture.
100 uWatts/sq-meter * 7000 sq-meters = 0.7 watts
That will work, but who is going to install a 143 meter long half wave
dipole just to light up an LED? Using a more reasonable 0.05
wavelength dipole, detected voltage will be 1/10th of the dipole,
resulting in 1/100 the detected power as 0.007 watts = 7 milliwatts.
That's considerably more than the 1.4 microwatts needed for the Osram
LED, so it's quite possible that it will work.
Assuming I add loading coils to the 1/10th wavelength antenna to bring
the dipole back up to 75 ohms, the input tank will need a turns ratio
of:
sqrt(1.4*10^6 ohms / 75 ohms) = 136:1
which is buildable but will probably need a big air core coil.
There are some additional losses, which haven't been considered. Balun
loss at the antenna, half wave rectification only recovers half the
power, decrease in antenna gain due to proximity to the ground, and
resistive (Q) losses in the tank circuit.
>My night light is an Avago green LED, a Tadiran lithium cell, and a 1
>meg resistor. It should last 20 or 30 years.
Incidentally, for finding things in the dark, I use phosphorescent
paper and a UV LED flashlight. I paste various shapes cut from the
paper to my calculator, cell phone, TV remote, door knob, light
switches, etc. They're not self lighting, but when hit with a UV
flashlight, they're VERY bright and easy to find.
<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>On Mon, 08 Oct 2012 10:24:49 +0300, upsided...@downunder.com wrote:
>>The effective area of a half wave dipole is about 0.12 square
>>wavelengths, thus at 1 MHz with 300 m wavelength, the capture area is
>>about 10000 m² or 1 ha.
>So at 100 uw/sq meter (a few miles from a 50 KW AM station) that
>computes to 1 watt! Of course, that's a pretty big antenna.
Sounds about right. That 100 µW/m² is in good agreement with old CCIR
(now ITU-R) field strength diagrams above average soil for a few
kilometers at 1 MHz.
With an outdoor antenna, just tune out the capacitively reactance with
some loading coil and use some 1:100 step up transformer and you might
get some usable LED currents.
I remember seeing some articles (long before the Internet) about a
transistorized tunable receiver powered by rectifying the signal from
a strong local broadcast station :-)
>>The typical ferrite rod antenna has a gain of -30 to -60 dB below the
>>full size dipole, thus the effective area is somewhere between 10 m²
>>and 1 dm². Multiply this by the local power density [W/m²] and you get
>>a ballpark figure of the available power.
>Even the low end 10 m^2 times 100 uw/m^2 is a milliwatt! A good LED is
>visible at a couple of microwatts, so even a ferrite rod antenna
>should visibly light up a good LED a few miles from an AM station.
For those not so familiar with the metric system 1 m² = 100 dm².
At the low end, the available power is only 1 µW.
Just googled around and found a measurements of a small (5 cm)
loopstick with -80 dBi gain, thus, the available power would be 100 nW
or 50 nA LED current. Perhaps an eye, well adapted to darkness for
half an hour, might be able to see something :-).
>> I wasn't thinking electrostatics (hard to get DC from) but picking up
>> energy from an AM station.
> Well, let's do the math.
> An LED will barely light at 1.4VDC and 2ma.
> 1.4V * 0.002A = 2.8mw
> That's how much power needs to be produced by this contrivance.
Your figures are over 3 decades out of date. The best modern white and some green LEDs are just about visibly lit on the die in normal room lighting at 1uA. When dark adapted you can drop that by a factor 100 or even more. ISTR the voltage drop is nearer 2v though and only green or white ones are worth trying since you need peak scotopic sensitivity.
Actual power requirement is about 2V * 0.01uA = 20nW in total darkness.
It is getting the 2v potential difference that is hard.
Selecting the brightest diode from a batch would be worthwhile...
> At the 2.5mV/m contour a fair size dipole antenna will pickup about
> 100uV into 75 ohms.
> <http://www.nd2x.net/calculators/FS.html>
> Notes:
> 1. Plenty of typo errors on this page, but the numbers seem correct.
> 2. A simple dipole still has 0dBd gain down to about 1/10th
> wavelength. The impedance becomes small, but the gain remains at
> about 0dBd.
> From the above calculator:
> Receive-Power = E^2/R = (100*10^-6)^2 / 75 = 133*10^-12 watts
> = 133 picowatts
> That's not anywhere near enough power needed by the LED (2.8
> milliwatts). However, it's more than enough for the <1 picowatt
> required to drive a crystal radio earphone.
> <http://www.crystal-radio.eu/enluidsprekertest.htm>
> The impedance of the LED is 1.4V / 0.002A = 700 ohms. 30pF of diode
Actually for a sensibly chosen modern high intensity LED suited to the task it is more like an impedance match to 2V/0.01uA = 200M.
> capacitance at 1MHz is about 5K ohms which isn't going to have a big
> impact on the 700 ohms. I'll call it 750 ohms because it makes the
> numbers come out neatly. A tapped coil resonant at about 1MHz should
> suffice for matching the 75 ohm antenna to the 750 ohm LED. Turns
> ratio is:
> sqrt(750/75) = 3.2
Except it that should be sqrt(2x10^8/75) = 1700 turns
> With 100uV at the antenna, the coil will deliver an inadequate 320uV
> to the LED. The LED needs 1.4V, not 320uV.
With 100uV on the antenna this gives 0.17V still not enough on its own, but a clever boost converter might be able to store enough energy on a low leakage capacitor for the occasional flash.
If you take the average current drawn down to below 1nA then it looks to me like you would be in the right ballpark for ordinary transistors.
The 5000t resonant coil will be hard to make though.
<jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>On Mon, 08 Oct 2012 12:39:52 -0700, Jeff Liebermann <je...@cruzio.com>
>wrote:
>>On Mon, 08 Oct 2012 07:12:06 -0700, John Larkin
>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>If a crystal set can pump milliwatts of electrical power into a
>>>headphone, something very similar should be able to light up an LED.
>>>No mixed fruits are involved.
>>The 30 to 100pF zero bias capacitance of an LED is going to look like
>>a short circuit compared to the less than 1pF capacitance of a decent
>>detector diode. However, if the capacitance is known, it can be
>>resonated with a series inductor to light up the LED:
>><http://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-...>
>Why not put the LED right across the crystal set LC tank? The tank
>needs capacitance anyhow. We have some high-efficiency LEDs in the low
>10s of pF. The classic crystal set cap was a 365 pF variable.
The resonant circuit impedance in a typical MW receiver is about 100
kOhms, while a LED circuit impedance is in the order of 100-1000 ohms.
On Tue, 09 Oct 2012 11:01:49 +0300, upsided...@downunder.com wrote:
>On Mon, 08 Oct 2012 19:15:06 -0700, John Larkin
><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>On Mon, 08 Oct 2012 12:39:52 -0700, Jeff Liebermann <je...@cruzio.com>
>>wrote:
>>>On Mon, 08 Oct 2012 07:12:06 -0700, John Larkin
>>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>>If a crystal set can pump milliwatts of electrical power into a
>>>>headphone, something very similar should be able to light up an LED.
>>>>No mixed fruits are involved.
>>>The 30 to 100pF zero bias capacitance of an LED is going to look like
>>>a short circuit compared to the less than 1pF capacitance of a decent
>>>detector diode. However, if the capacitance is known, it can be
>>>resonated with a series inductor to light up the LED:
>>><http://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-...>
>>Why not put the LED right across the crystal set LC tank? The tank
>>needs capacitance anyhow. We have some high-efficiency LEDs in the low
>>10s of pF. The classic crystal set cap was a 365 pF variable.
>The resonant circuit impedance in a typical MW receiver is about 100
>kOhms, while a LED circuit impedance is in the order of 100-1000 ohms.
On a sunny day (Tue, 9 Oct 2012 02:33:00 +0000 (UTC)) it happened gregz
<ze...@comcast.net> wrote in
<112229436371442754.348458zekor-comcast....@news.eternal-september.org>:
>Jan Panteltje <pNaonStpealm...@yahoo.com> wrote:
>> On a sunny day (Tue, 02 Oct 2012 02:29:39 -0700) it happened miso
>> <m...@sushi.com> wrote in <k4ec7i$ha...@speranza.aioe.org>:
>>> I vaguely remember something about fluorescent tubes glowing near TV >>> transmitters. Maybe Sutro tower.
>> Fluorescent tubes also glow when stuck in the ground under a HV power line.
>> I used to walk around with a glowing neon next ot my 7 MHz xmtr antenna.
>I need to try something like that up the road, where there are no houses,
>but feed from power station goes through. Three phase might screw it up.
John Larkin <jjlar...@highNOTlandTHIStechnologyPART.com> writes:
> On Tue, 02 Oct 2012 02:29:39 -0700, miso <m...@sushi.com> wrote:
>>I vaguely remember something about fluorescent tubes glowing near TV >>transmitters. Maybe Sutro tower.
> I used to drive past Sutro Tower every day in my ratty old Ford
> Fiesta. The speakers would make awful sounds, even with the radio off.
> I guess the output transistors were rectifying the RF picked up by the
> speaker wires.
>>>>I wasn't thinking electrostatics (hard to get DC from) but picking up
>>>>energy from an AM station.
>>>Well, let's do the math.
>>>An LED will barely light at 1.4VDC and 2ma.
>>> 1.4V * 0.002A = 2.8mw
>>>That's how much power needs to be produced by this contrivance.
>>The Osram LEDs that we use are often too bright at 2 mA. They are
>>clearly on in normal room lighting at 1 uA. Dark adapted, under
>>optimal conditions, I could just barely make out light at about 800
>>picoamps.
>Ok, let's use your numbers and see what happens.
>I didn't know LED's would work at 1uA. Do you have the Osram part
>number handy? I found some "low current" LED's on the Osram web pile,
>but the spec sheets were for 2ma.
>>According to my good'ol red Radiotron book, one might expect roughly
>>100 uW/sq meter a few miles from a 50KW AM station. You'd only need a
>>couple of uW to light an LED. So it's just an impedance matching
>>problem. It looks doable.
>The ambient level in US metro areas is about 50 uWatts/sq-meter from a
>study that I can't seem to find.
>For the antenna:
>100 uWatts/sq-meter is the energy density. To convert into detected
>energy, the effective aperture of the receive antenna will be needed:
><http://vk1od.net/antenna/concepts/Ae.htm>
><http://vk1od.net/software/fsc/index.htm>
>For a 1Mhz dipole (143 meters long), that would be about 7000
>sq-meters effective aperture.
> 100 uWatts/sq-meter * 7000 sq-meters = 0.7 watts
>That will work, but who is going to install a 143 meter long half wave
>dipole just to light up an LED?
To be precise 70 m tall (vertical polarization).
>Using a more reasonable 0.05
>wavelength dipole, detected voltage will be 1/10th of the dipole,
at what impedance level ?
>resulting in 1/100 the detected power as 0.007 watts = 7 milliwatts.
>That's considerably more than the 1.4 microwatts needed for the Osram
>LED, so it's quite possible that it will work.
>Assuming I add loading coils to the 1/10th wavelength antenna to bring
>the dipole back up to 75 ohms, the input tank will need a turns ratio
>of:
> sqrt(1.4*10^6 ohms / 75 ohms) = 136:1
>which is buildable but will probably need a big air core coil.
The radiation resistance drops inversely proportionally to the square
of wavelength below 1/4 wavelengths, thus the matching network not
only needs to tune out the antenna capacitively reactance, but also
transform the very low (a few ohms or less) to the standard 50/75 ohm
impedance levels.
>>On Mon, 08 Oct 2012 19:15:06 -0700, John Larkin
>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>On Mon, 08 Oct 2012 12:39:52 -0700, Jeff Liebermann <je...@cruzio.com>
>>>wrote:
>>>>On Mon, 08 Oct 2012 07:12:06 -0700, John Larkin
>>>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>>>If a crystal set can pump milliwatts of electrical power into a
>>>>>headphone, something very similar should be able to light up an LED.
>>>>>No mixed fruits are involved.
>>>>The 30 to 100pF zero bias capacitance of an LED is going to look like
>>>>a short circuit compared to the less than 1pF capacitance of a decent
>>>>detector diode. However, if the capacitance is known, it can be
>>>>resonated with a series inductor to light up the LED:
>>>><http://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-...>
>>>Why not put the LED right across the crystal set LC tank? The tank
>>>needs capacitance anyhow. We have some high-efficiency LEDs in the low
>>>10s of pF. The classic crystal set cap was a 365 pF variable.
>>The resonant circuit impedance in a typical MW receiver is about 100
>>kOhms, while a LED circuit impedance is in the order of 100-1000 ohms.
> AM radios receive femtowatts.
> Sheesh!
Definitively _NOT_
While the provincial US organization "IHF" tried to introduce the dBf
(femptowatt decibels above 1 W) in order to make some sense into
advertisement, those 10 dBf figures apply _only_ to receivers in the
100 MHz band with +/- 75 kHz FM deviation.
Due to the band noise around 1 MHz, those dBf figures are useless.
On Tue, 09 Oct 2012 12:14:58 +0300, upsided...@downunder.com wrote:
>On Tue, 09 Oct 2012 01:09:28 -0700, MrTallyman
><MrTally...@BananaCountersRUs.org> wrote:
>>On Tue, 09 Oct 2012 11:01:49 +0300, upsided...@downunder.com wrote:
>>>On Mon, 08 Oct 2012 19:15:06 -0700, John Larkin
>>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>>On Mon, 08 Oct 2012 12:39:52 -0700, Jeff Liebermann <je...@cruzio.com>
>>>>wrote:
>>>>>On Mon, 08 Oct 2012 07:12:06 -0700, John Larkin
>>>>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>>>>If a crystal set can pump milliwatts of electrical power into a
>>>>>>headphone, something very similar should be able to light up an LED.
>>>>>>No mixed fruits are involved.
>>>>>The 30 to 100pF zero bias capacitance of an LED is going to look like
>>>>>a short circuit compared to the less than 1pF capacitance of a decent
>>>>>detector diode. However, if the capacitance is known, it can be
>>>>>resonated with a series inductor to light up the LED:
>>>>><http://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-...>
>>>>Why not put the LED right across the crystal set LC tank? The tank
>>>>needs capacitance anyhow. We have some high-efficiency LEDs in the low
>>>>10s of pF. The classic crystal set cap was a 365 pF variable.
>>>The resonant circuit impedance in a typical MW receiver is about 100
>>>kOhms, while a LED circuit impedance is in the order of 100-1000 ohms.
>> AM radios receive femtowatts.
>> Sheesh!
>Definitively _NOT_
>While the provincial US organization "IHF" tried to introduce the dBf
>(femptowatt decibels above 1 W)
On Tue, 09 Oct 2012 10:55:52 +0300, upsided...@downunder.com wrote:
>On Mon, 08 Oct 2012 19:38:47 -0700, John Larkin
><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>On Mon, 08 Oct 2012 10:24:49 +0300, upsided...@downunder.com wrote:
>>>The effective area of a half wave dipole is about 0.12 square
>>>wavelengths, thus at 1 MHz with 300 m wavelength, the capture area is
>>>about 10000 m² or 1 ha.
>>So at 100 uw/sq meter (a few miles from a 50 KW AM station) that
>>computes to 1 watt! Of course, that's a pretty big antenna.
>Sounds about right. That 100 µW/m² is in good agreement with old CCIR
>(now ITU-R) field strength diagrams above average soil for a few
>kilometers at 1 MHz.
>With an outdoor antenna, just tune out the capacitively reactance with
>some loading coil and use some 1:100 step up transformer and you might
>get some usable LED currents.
>I remember seeing some articles (long before the Internet) about a
>transistorized tunable receiver powered by rectifying the signal from
>a strong local broadcast station :-)
>>>The typical ferrite rod antenna has a gain of -30 to -60 dB below the
>>>full size dipole, thus the effective area is somewhere between 10 m²
>>>and 1 dm². Multiply this by the local power density [W/m²] and you get
>>>a ballpark figure of the available power.
>>Even the low end 10 m^2 times 100 uw/m^2 is a milliwatt! A good LED is
>>visible at a couple of microwatts, so even a ferrite rod antenna
>>should visibly light up a good LED a few miles from an AM station.
>For those not so familiar with the metric system 1 m² = 100 dm².
Oh, yeah. We engineers use powers of 3, mostly. Deci and hecta and all
those are less common here.
A longwire, or a big loop, sounds more promising.
>At the low end, the available power is only 1 µW.
>Just googled around and found a measurements of a small (5 cm)
>loopstick with -80 dBi gain, thus, the available power would be 100 nW
>or 50 nA LED current. Perhaps an eye, well adapted to darkness for
>half an hour, might be able to see something :-).
1 nw is about the threshold with a good green LED.
On Tue, 09 Oct 2012 11:01:49 +0300, upsided...@downunder.com wrote:
>On Mon, 08 Oct 2012 19:15:06 -0700, John Larkin
><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>On Mon, 08 Oct 2012 12:39:52 -0700, Jeff Liebermann <je...@cruzio.com>
>>wrote:
>>>On Mon, 08 Oct 2012 07:12:06 -0700, John Larkin
>>><jjlar...@highNOTlandTHIStechnologyPART.com> wrote:
>>>>If a crystal set can pump milliwatts of electrical power into a
>>>>headphone, something very similar should be able to light up an LED.
>>>>No mixed fruits are involved.
>>>The 30 to 100pF zero bias capacitance of an LED is going to look like
>>>a short circuit compared to the less than 1pF capacitance of a decent
>>>detector diode. However, if the capacitance is known, it can be
>>>resonated with a series inductor to light up the LED:
>>><http://www.edn.com/design/led/4368392/An-LED-s-intrinsic-capacitance-...>
>>Why not put the LED right across the crystal set LC tank? The tank
>>needs capacitance anyhow. We have some high-efficiency LEDs in the low
>>10s of pF. The classic crystal set cap was a 365 pF variable.
>The resonant circuit impedance in a typical MW receiver is about 100
>kOhms, while a LED circuit impedance is in the order of 100-1000 ohms.
Hey, we're engineers. We know how to match impedances.