Nano Waves

[Budd Churchward]

At the Digital Group Meeting tonight I mentioned a project that I am starting that uses light a very bright LED as the carrier frequency for sending Data and Voice communications over remarkably long distances.

I will take some time over the next few weeks to document my progress in building a couple of these transceivers. For now, I will share a few links to the sites that got me excited enough to tackle this project.

The first is the 6th entry in VE7SL's blog where he and a friend actually succeed in running a two way QSO over about 20 miles. In addition to a short video of the actual transmission you will see a map of the link between the two stations. It was this map that got me excited. I did not realize while I was following this blog that these experiments where actually going on outside my window. Well, almost anyway. You see, with due respect to Ms Palin, I CAN see Canada from my porch. If you find number 6 interesting, work you way back and read the others.

View VE7SL's Blog Here

And click here for an article that tells more about the project.

Steve and Marcus are using a dual timer chip, the 556 (basically two 555 timers in on IC) to generate two audio tones. The 556 drives a high powered LED through a MOSFET that pulses the LED at 600 Hz. I plan on driving my LED with a BuckPuck LED driver and an Arduino. The BuckPuck is capable of flashing an LED at up to 10 KHz and can be driven from a simple I/O pin on any micro-controller. Click the highlight link above. I ordered two of these BuckPucks from the website at LED Supply and they arrived today. They are not the ones I need, they are not even the ones pictured on the web page. I received the 3021-D-N-700 ... what I need is 30212-D-E-700. I will have to call them and sort that out. There doesn't seem to be a way to choose the 'E' version from their drop down menus, however they offer the one I need on Amazon for a few dollars more. The N version doesn't have the control and reference pins that I need to run this thing with my Arduino.

It is my belief that using these parts I will be able to make a very simple transmitter without the need for a dozen other components. The BuckPuck not only provides a constant current source for the LED but also provides the regulated 5V needed to run the Arduino. So it is simply: 7 - 28 V input into the BuckPuck. An LED hooked up to the output ... no current limiting resistor needed cuzz the PB does all that. And an Arduino, hooked up to the BP's control pin, running code to send my beacon. The fun part will be building the optics to focus the beam of light. All of those parts arrived today as well.

There is a simple lens called a collimator that narrows the wide angle beam of light from the LED down to about 6 degrees. This is set up at the focal point of a cheap Fresnel lens (they are used to magnify pages of text for old farts) . And batta-bing! we have a nano-transmitter.

So what I think would be really really cool would be to get these Crazy Canucks to set up a beacon at the Grouse Mountain Ski Slopes. There are lots of places around Bellingham that should be able to receive that signal. Are there any skiers in our group?

73 for now,

Budd - WB7FHC

[Budd Churchward]

I tried my first experiment late last night with a simple little LED flashlight and a 8.5 x 11 inch Fresnel lens. I purchased a packet of two lenses from Amazon (see the item here) for a penny less that $8. That's $4 a lens. I will need one for the transmitter and one for the receiver. The flashlight was one of these simple little gift versions that you see around. I got mine from Wells Fargo Advisors. It has their logo on it. The light has nine LEDs arranged in a circle, or hexagon, of eight with a ninth LED in the center.

 

At around 11:00 last night I took them outdoors and aimed the flash light at the tops of trees behind houses in my neighborhood. There was no indication that the light was reaching them. Then I held the Fresnel in front of the flashlight so that the beam went through it. I focused by moving them back and forth changing the distance between them. Not only was I able to light up the tree tops, but when I got the focus just right, I could clearly see the pattern of nine bright perfect circles projected on the evergreens. I tried this at trees even further away and was still able to clearly see the circles, even at about 100 yards. I had no trouble lighting up a road sign down the street from me. I aimed my beam out over the water in front of my house and spooked several hundred ducks that noisily few off into the darkness. No doubt they settled back in over near Birch Bay Village where the human's are much less rude and disruptive. It's hard to be a duck.

 

The Fresnel is listed as being 3X magnification. It is a flexible plastic sheet, probably larger than I need. I think I will be able to trim it down by cutting back from all four edges. I need to come up with a way of making the lens flat and rigid and then holding it a fixed distance from the LED on the transmitter and the Photodiode on the receiver. I've read where some have used picture frames to tame the flop. I should give that a try. 

[Budd Churchward]

There seems to be a bug when posting a reply to Google Groups. When you click 'reply' the box appears. I entered my reply. When I click on 'POST' I see an error that says: "You cannot create a post that contains no content or attachments." What? I entered three paragraphs. How is that 'no content'?

I searched for a solution. The text box appears with grey pipe in the upper left corner. You have to press 'Enter' twice and then 'Back Space' twice to remove it. After this quirky little nonsense, enter your reply and the post will succeed.

[Budd Churchward]

I just got off the phone with support and LEDSupply and got my driver problem sorted out. My problem was that I had selected 'non-dimming'  from the drop down menu and I should have selected 'dimming' .... of course, I don't want to dim the LED, but that's the version that will let me control it with the Arduino. I'm sending the wrong ones back and have ordered the ones I need. (I am building two of these transceivers so that I can talk to myself).

If anyone is interested in trying this out with me, here are the parts that I have ordered from LEDSupply .... You can use my coupon code: 5CS as long as you order before OCT 31.

Luxeon Rebel LED ... Deep Red 660 nm Part# 07040-PDRED-C

Carclo Lens Holder Part #10732

Carclo Lens - Plain Tight Spot Part# 10193

BuckPuck DC LED Driver (7-Pin SIP/Dimming/700mA Part# 03021-D-E-700

All total a single set of this stuff runs about $25 with the BuckPuck being more than half of it.

You can find Fresnel lens everywhere. I think I will check one of the dollar stores.

The BuckPuck may not be needed. There are schematics around to build your own driver which can probably be done for one-third the cost. I just want to give it a try, its only $14.

[Gayland Gump]

Very cool Budd.  Glad to hear you are proceeding. 

Gayland

KG7GCF

--
You received this message because you are subscribed to the Google Groups "Northwest Washington Digital Amateur Radio" group.
To unsubscribe from this group and stop receiving emails from it, send an email to nwwdigi+u...@googlegroups.com.
To post to this group, send email to nww...@googlegroups.com.
Visit this group at http://groups.google.com/group/nwwdigi.
For more options, visit https://groups.google.com/d/optout.

[Budd Churchward]

Since my last post I was able to get the right version of BuckPuck. Art gave me a couple of heat sinks and I have attached the LEDs to them. These puppies get very hot with 700mA of current running through them. I have lit them up and let me tell you! They are bright. I have been experimenting with sending signals to the control pin on the BuckPuck. Using an Arduino, I experimented with sending a audible tone which is really just a square wave. Not sure exactly what's happening, but my HF rig which happened to be on at the time picked up the audio and I could hear it on all bands. Didn't expect that at all.

The BuckPuck is an LED driver. It requires an input voltage between 7 and 32 Volts. I usually have my Morseduino CW decoder hooked up to the accessory jack on the Icom 7410. That provides a switched 8 V which I decided to use for the BuckPuck. My guess is that the audio is finding its way into the receiver through that. But I'm not sure.

I played around with PWM (pulse width modulation) also. I was able to dim the LED with that. It is interesting that the PWM also produced an audible tone in the HF receiver. With PWM, the frequency stays the same, so the tone did not vary as it did when I was using my Arduino to output tones of specific frequencies. PWM varies the length of ON time vs. OFF time which in the end varies the RMS current going to the LED. It is a standard way to dim LEDs and vary the speed of DC motors. I use PWM to control the speed of my electric car.

I played around with some lenses. The collimating  lens concentrates the width of the beam quite nicely. I have been holding the Fresnel lens above it and can project a tight image of the LED on the ceiling in my ham shack.

I will hook up my oscilloscope and see what kind of a trace I get on the output to the LED.

I am still hoping that I will be able to modulate the LED with the BuckPuck because it is designed to provide a constant current of 700 mA s to the LED.

Next up will be to figure out how to make sense of the photodiode. And see if I can demodulate the same frequencies that I am pumping into the LED. I am going to experiment with an Adafruit Class D stereo amplifier that I picked up some time ago. I don't know if this can be done, but I am thinking of running the signal from the photodiode into the left channel of the stereo amp and then swinging that back through the right channel so that it works like a two stage amplifier. I have absolutely no idea whether that can be done or not, but what the heck! Let's give it a try.

Budd

[Robert Sears]

Oh I forgot to add square waves are rich in harmonics, and fast rising and falling edges cause lots of EMI and all the little wires on your breadboard make great little antennas as the noise in your HF rig can attest to.

Robert

[Budd Churchward]

Thanks for the comment, Robert. I am aware that square waves cause a lot of harmonics, but I thought that was mostly at RF frequencies. Not sure what effect it has at AF. The good news is that the BuckPuck LED driver is putting out a real nice sine wave to the LED. I was able to see it clearly on the Oscilloscope. 

Here is a cool image that I took with my phone. This proves that the LED is flashing when I output a tone from an I/O pin on the Arduino. I am creating a 440 Hz tone that is attached to the control pin on the LED driver. I took a business card and flapped it quickly in the light of the LED. I tried different frequencies and it changed the width of the visible card edges and the gap between them. The spacing is always equal. That is the off time and the on time are the same. Don't be confused by the lines in my window shade behind the scope. They are straight and parallel. The card edge is much brighter and slightly curved.

The cool thing here is that my LED appears to be on steady, It is only when I quickly move an object back and forth within the beam that the strobe effect becomes visible.

I am satisfied that I can modulate the LED with an Arduino at any audio frequency.

Now it is time to teach myself how to use the photodiode and demodulate the LED strobe.

 

[Robert Sears]

Some reason this did not get through last night.

But something that tripped me up when using a photo diode was that photo diodes work like Zener diodes, so they go in the circuit reverse of what you would think. The Anode connects to ground and the Cathode would be connected to the current limit resistor that is connected to positive voltage.

Robert

Next up will be to figure out how to make sense of the photodiode. And see if I can demodulate the same frequencies that I am pumping into the LED. I am going to experiment with an Adafruit Class D stereo amplifier that I picked up some time ago. I don't know if this can be done, but I am thinking of running the signal from the photodiode into the left channel of the stereo amp and then swinging that back through the right channel so that it works like a two stage amplifier. I have absolutely no idea whether that can be done or not, but what the heck! Let's give it a try.

Budd

[Budd Churchward]

Yes, a photodiode has to be reverse biased to create the depletion zone. When a photon hits the depletion zone it splits off an electron-hole pair that results in the flow of current. The more light that hits the zone, the more current you get. And there is something called 'dark current' .... I just love that term. It almost sounds like some sort of Sci-Fi power. Even though the diode is reverse biased, there is still some current flowing through it when no light is there. Hence the term. Because they are used in devices that actually measure the amount of light, such as in photographic purposes, you need to know the amount of dark current so you can subtract it from the total current to get an accurate reading. Lucky for me, I don't need to know how much current is there, I only need to detect the changes.

For this reason I am a bit less enthusiastic about my last night's experiments than I was when I called it quits and posted my results. Although I can easily change the frequency of the strobe, I am always in a 50% duty cycle. I think this might mean that the current will be the same regardless of the frequency. This is even more probable because the BuckPuck is specifically designed to deliver a constant current even though the voltage may vary. What I need to know is whether or not I can recreate a square wave with the photodiode. If I was using a laser instead of an LED, I would have to change the duty cycle of the PWM in order to modulate the light. Lasers have a very narrow current range that limits amplitude modulation. If the current is too low, you get no light. If it is too high, you never get light again. I am looking into some advanced and creative ways to program the ATMega so that I can manipulate the duty cycle without using the standard analogWrite() function. AnalogWrite() will probably work for data communication, but I don't think it has the speed or the range needed to send voice.

So far, all the documentation I can find on Light Wave Communication uses hardware to solve the problems. Everything is done with circuit components. They use 555 timers to create their square waves. I am looking for software solutions wherever they might work. I have shown that I can create the transmitter with software and very few components, albeit the LED driver is chock full of its own circuits ... I didn't have to build them. The receiver won't be so easy.

On Friday, October 30, 2015 at 6:50:18 AM UTC-7, Robert Sears KF7VOP wrote:

Some reason this did not get through last night.

But something that tripped me up when using a photo diode was that photo diodes work like Zener diodes, so they go in the circuit reverse of what you would think. The Anode connects to ground and the Cathode would be connected to the current limit resistor that is connected to positive voltage.

Robert

Next up will be to figure out how to make sense of the photodiode. And see if I can demodulate the same frequencies that I am pumping into the LED. I am going to experiment with an Adafruit Class D stereo amplifier that I picked up some time ago. I don't know if this can be done, but I am thinking of running the signal from the photodiode into the left channel of the stereo amp and then swinging that back through the right channel so that it works like a two stage amplifier. I have absolutely no idea whether that can be done or not, but what the heck! Let's give it a try.

Budd

--
You received this message because you are subscribed to the Google Groups "Northwest Washington Digital Amateur Radio" group.

To unsubscribe from this group and stop receiving emails from it, send an email to nwwdigi+unsubscribe@googlegroups.com.

[Budd Churchward]

I built up a little Perm-Proto version of my Luxeon Rebel 660nm transmitter. As you can see in the photo, I have installed my own ATMega328 right on the board rather than hooking things up to my Arduino Uno as I did earlier in the breadboard version. Building your own Arduino is no big deal. You can almost count all the parts on the fingers of one hand and still have one left over to pick your nose. I usually throw in a few extras for convenience. I have added the reset switch (button) and three header pins connected to pins 1, 2 and 3 of the ATMega. These are in that order: RESET, RX and TX. I can pull the micro-controller chip out of my Uno and jump those same three points from the naked Arduino back to my pins and program the chip just as if it was sitting happy in its 28 pin Arduino socket.

The data sheet on the BuckPuck says that with a DC input of 7 to 32 volts the REF pin will put out a regulated 5V that can run a micro-controller.

What I discovered, though, was quite unusual. I wrote a very simple sketch to blink the LED three times quickly and then wait 3 seconds before doing it all over again and again. What it produced was an LED that simply blinked quickly over and over and never took the 3 second break. Frustrated I decided that the only logical explanation was that the BuckPuck could not drive the ATMega. I didn't think it was putting out enough current.

This morning I was talking to Peter, WA7FUS, about this on the 220 Mnt. Constitution repeater. I told him that I thought the Data Sheet lied to me. I've heard before that you can not always trust what the manufacture says. So I went back to the data sheet. And guess what. It tells me that the maximum output on the REF pin is 5V at 20mA. That didn't seem like much to me, so I asked Uncle Google: "How much current does an Arduino Uno draw?" And he came back with 80mA. Ha! Did the data sheet lie or not? Well the guy that wrote the data table probably didn't lie, must have been the marketing genius that said: "Let's tell all those DIY guys they can use this magic little cube as a power supply for their cool Arduino project." I mean really! The reason I went with the BuckPuck was because I thought it could also be my voltage regulator. And to think, I bought two of them.

So tonight, I am going to add an L7805C voltage regulator and a couple of filter caps to the Perma-Proto and see if that makes a difference.

BTW ... the Perma-Proto is from Adafruit. And although I thought they were a bit lame because they are set up like a normal bread board, I soon learned that unlike with a bread board, I can break the rules and cut traces. That's why you see in the photo that I am bringing the input voltage into the +/- rails and then using the same rails for my regulated 5 volts. Be careful with those Xacto Knives, they can be our friends sometimes and sometimes not so much.

73

Budd - WB7FHC

[Art]

that buckpuck is kinda lame in that its only 20ma at the ref pin.... I have a couple of heat sinks that would be good for the 7805 regulator if you need one.

fyi at some point for the 'final' product you might want to take the heat sink and polish it with 500 grit then 1000+ grit sandpaper... the smoother the surface the better the heat transfer.... also I still have a few of those cpu coolers that you can do the same thing to.

Looking good! can't wait to see the video =D

--
You received this message because you are subscribed to the Google Groups "Northwest Washington Digital Amateur Radio" group.

To unsubscribe from this group and stop receiving emails from it, send an email to nwwdigi+u...@googlegroups.com.

[Budd Churchward]

No Art, the BuckPuck isn't lame after all. I am the lame one! I ran jumpers between my board and the Arduino Uno while I program and debugging the sketch. I was ready to add the L5805 but wanted to get video of the LED blinking all crazy, but it worked perfectly. It did everything it was supposed to. There was no problem. So I started thinking about what I was doing and BINGO ... I didn't have the Arduino connected. It turns out that the 20 mA seems to run the ATMega just fine, but when I hang the extra Uno board on it and all the hardware it has running we have problems. If I disconnect the bridge between the two boards everything works just fine. Gads, I should have see that the first time.

I thought about doing the sanding thing myself. The Luxdrive people have a series of videos on YouTube that show you how to use things things. They showed a technique they called 'tinning' where you apply the epoxy compound to the surface of the heat sink and scrape it back and forth, every which direction, until it looks like it is no longer there. It fills in the cracks and scratches. Then you attach the Luxion Rebel with the same compound.

Your heat sink is definitely working. I am not detecting any heat build up at all. Boy that BuckPuck heats up, though.

I'm working on the video now. Watch for it in the next few days.

Budd

To unsubscribe from this group and stop receiving emails from it, send an email to nwwdigi+unsubscribe@googlegroups.com.

[Robert Sears]

I have some samples of thermal pads of varying thickness and a bunch of 6''x 6'' sil-pad sheets that have one sticky side.

If you would like to try some.

To unsubscribe from this group and stop receiving emails from it, send an email to nwwdigi+u...@googlegroups.com.

[Art]

There is a low power setting for the arduino, you can shut off a lot of stuff that you aren't using... also I don't know what you require for outputs but I'm wondering if there is another atmega package that would work thats lower power...

On 11/2/2015 9:09 PM, Budd Churchward wrote:

To unsubscribe from this group and stop receiving emails from it, send an email to nwwdigi+u...@googlegroups.com.

[Budd Churchward]

This is why they call it experimenting!

I thought I was out of the woods on the current problem after I disconnected the programming board. I soon learned, however, that as my sketch became more complex the ATMega drug its heals and pooped out on me. I went ahead and installed the 7805 voltage regulator and a couple of caps and all is good.

The REF pin on the BuckPuck is supposed to provide a voltage that you can run through a variable resistor and then into the CTRL pin. This creates a dimmer control. These things were built to drive stuff like under the cupboard lighting strips in your kitchen. They can also blink those bright 'OPEN - CLOSE' signs you see in store windows. Instead of dimming you turn it on by sinking the REF voltage to GRN and you turn it off by sending it through.

I'm not using the REF for anything now. I PWM 5V from the ATMega to CRTL and the BP stobes the LED.

[Budd Churchward]

I'm liking what I see.

After putting my own power supply on the board I am having no troubles at all getting all sorts of exotic effects.

The image above is a shot of my scope. The board has been running continuously now for three hours. I have it cycling through 9 different audio frequencies: 110, 220, 330, 440, ... 990 Hz. And it just keeps chugging on. The blue trace is the output from D9 on the micro-controller. The yellow line is attached to the +/- leads of the LED. This image is at 550 Hz. It looks like I can pretty much cover the audio frequencies just fine.

Budd

[Budd Churchward]

I made an animated gif to demonstrate PWM (Pulse Width Modulation) ... I am flapping a playing card in the light beam from my Luxeon Rebel Deep Red LED while my ATMega328 micro-controller varies the duty cycle of a PWM signal linked to the BuckPuck LED driver. Notice the change in the thickness of the light and dark bands as the duty cycle changes. The PWM signal is running at about 500 Hz.

[Robert Sears]

Don't forget the meeting tonight, as scheduled I will be discussing
oscilloscopes and their proper usage.

If you have a scope feel free to bring it and I will answer any
questions you have on it's operation.

Robert/KF7VOP

[Art Miller]

I don't think I'm going to make it tonight, not feeling too well and bad weather don't mix :(

[Robert Sears]

Here is the XYX's of Oscilloscopes document I had tonight. It's a little
geared towards Tektronix "Digital Phosphorus" scopes but the basic
fundamentals are the same for every scope.

Robert/KF7VOP

[Budd Churchward]

I just sent this board off to be made up by OSH Park. I tried my hand at creating a tiny one. All the resistors stand up on end. It is only 2.25 by 1.25 inches. The schematic is one done up by K3PGP of a G3XBM design. http://ve7sl.blogspot.ca/search/label/lightwave

By making the board so tiny, I was able to order three of them for only $13.

I have been experimenting with putting photo-diodes in series and parallel. My PCB will take two. I have a jumper that will allow me to switch the second one in and out of the circuit. This version puts them in parallel which should increase the current, but not the voltage. I hope that's what I want. If I put them in series, it would increase the voltage but not the current. The current from the diodes drives a JFET transistor which then goes through three more amplification stages on it's way to head phones.

I plan to mount the photo-diodes on the back of the board where they will be in the focal point of my Fresnel lens.