Beacon Project "Common and Precious" (OFF air, 10.237,5kHz, 5W)

[Zazzle]

Hi everyone,

Firts of all: I'm not a native speaker of the english tongue. Please bear with me in case I screw up the grammar every now and then.

I toy with RF stuff since my childhood. (I have a passion for RF stuff on shortwave since I was eight. Back then my grand mother gave me her old valve radio as a birthday present and after I hit the SW button one day in the evening I could not let go. For me, SW is the modern way of the 'travel bug'). I've build quite some stuff so far and operate a small "wide range radio thermometer grid" in my area (µC controlled, solar powered, 1W 70cm transmitters. The farest one is 8km).

Now I want to build a CW beacon. Why? First, there is this a certain curiosity on the technical side: tinkering and the question if someone will pick it up, ever. Then there's this romantic side. A CW beacon is some kind of light house in a dark, stormy night. It's there when no one's around, sending it message to whoever may hear it.

So, enough smalltalk. Lets start with some technical stuff

1) The concept:
The whole beacon will be solar powered (by a 20-40W PV panel - overdone for the summertime but will also offer enough charge during winter) and controlled by an AVR controller. It will charge during day and as soon as the sun has set it'll start transmitting a message. The mikrocontroller will monitor the battery and shut down the device to keep itself in operation with 20% of battery capacity remaining as a buffer. The Battery, charge controller, mikrocontroller and transmitter will be mounted inside a water proof box. It's a low budget project because I guess sooner or later someone will find and steal/disassemble or damage it. The problem is that I live in a rather large city near its limits and so far I couldn't find a proper place that would allow to mount such a transmitter high above the ground without being noticed. Therefore I have to make a tradeoff between performance and security.

Picture

2) Location
The main problem is the place where I live. It's a big city. There are no hills around. But there's lot of destructive people that like to damage stuff and there's a lot of background noise. The surrounding land is more or less flat (and used for agiculture purposes). My focus is on three primary locations:

A) Old Airfield

Picture

It's possible to mount the module close to the ground so it won't be visible from some distance. There's an old fence with massive wooden post running across the field. I could mount long wooden slants to them, holding the transmitter an Antenna.

PRO
- No buildings around
- Wide open space
- If installed clever, not visible from the distance
CONS
- The Antenna will only be 2-3m above the ground
- Easy to reach / disassemble / damage

B) Old bridge

Picture

I could install the PV module and transmitter easily in the middle of the bridge and run the antenna wires left and right towards surrounding trees.

PROS
- Transmitter and module easy to hide
- Not easy to reach
CONS
- The construction is made of steel
- The top of the bride is just 3-4m above the ground since its socket is below the ground level.

C) Old depot house roof

Picture

Except of the (thin) antenna wires the whole setup is easy to hide. Climbing into the roof is hard and requires acrobatic stills. Therefore the setup is not prone to be stolen or suffering from damage.

PROS
- Not easy to access
- Antenan wires will be 10-12m above the ground
CONS
- Still in the city, higer building on one side, an open air field on the other side
- The roof construction is made of steel

3) The frequency.
I'm completely not sure about the frequency I want to use. My circuit is cabable of tuning about -/-20kHz around the fundamental crystal ferquency. I have crystals for:

10.00000 MHz
13.25583 MHz
16.48555 MHz

For me it's important to not disturb the HAM bands or any other service like nautic communication and alike. But of course I want my beacon to be noticed. Therefore I'm looking for a frequency that's not prne to cause trouble but is still likely to be scanned frequently.

My favorite is the 10MHz one. Because it's close to the 30m CW band.

4) The Signal.
This is primary about battery lifetime. Like told before, the whole device runs on Solar power and I also want it to operate during winter, even if not all night long. I have to make a decision between the TX-Time (the more often the more possible the beacon will be noticed) and battery lifetime (there is no use in having the transmitter shut down after just one hour of operation).

At the moment the Concept looks like this:

- Every 10 seconds a 1s long dash (to get attention)
- Each 10 Minutes the following message will be transmitted:

"<LOCATOR> SOLAR NIGHT ONLY UBAT 125" (where UBAT 125 means BATTERY VOLTAGE = 12,5V.)

5) Antenna
I think I go for a classic wire dipole. It will be quite long (23 ft. 4.8 inches / 7.1 mtr. per element) but on the other hand it's easy to build and mount. The box holding the transmitter will be mounted in it's middle. Anything I should pay attention to?

6) Operation times
Given that the signal travels way more far during night I forsee from running the transmitter during day. What's your opinion on that?

7) RF-Power.
Currently the Transmitter delivers 1,6W into Z=50R. Is that too much? Since it's CW I guess it's quite a lot. On the other hand, I have some background noises to overcome, the antenna isn't mounted very high and I still want to bridge some distance.

Yours,
~Zazzle

[ff]

Nice project Zazzle!  I have read your post several times and still have yet to digest it all.  I agree with you about using the 10 MHz range.  Since your unit will be in charging mode all day, you should stick with lower bands for more reliable nighttime operation, when you will actually be transmitting.  It is imperative to keep clear of ham bands and anywhere that is being used.  If it were me, I would start monitoring several different frequencies, looking for a little used place.  Cheap crystals can be found on most electronics websites.  Most of these are defunct microprocessor clock freqs; many of these would work well for you.
I think your power level is fine - IF your location is at least a hundred meters or so from where someone would be using phones or radios or TVs.  In that situation, anything up to 5 watts seems OK to me.  Regarding that, my choice would be the old depot, IF you don't hurt yourself getting an antenna up there.  I have many people around me that like to destroy things too, but the same people tend to be lazy enough that they won't mess with something that takes a lot of effort to get to.  Have fun and good luck with your project Zazzle!

[Zazzle]

Hi there, ff!

First of all: thank you for the answer: I was already fearing to go without one

Heh, yes. That, it was quite a bunch of text, and there's more stuff I keep pondering in my head. But pondering leads to nothing, doing stuff does. So here we go. This weekend the Transmitter will, if nothing goes wrong, on air.

I'd  settle on 9.995Mhz. Around here it wasn't in use each time I checked in the evening hours every now and then. Tomorrow I'll spend my day adjusting the oscillator and power stage for the best performance. Friday is assembling day and on Saturday late evening the baby is set free. When it comes to the climbing... luckily I'm quite acrobatic and may also use a climbing harness. Man, I'm so exited. I hope no one will damage or steal it.

I also had some more thoughts about the signal. So far I wanted to send dashes each 10 seconds to get attention. Each 10 Minutes I wanted to transmit the message. But that's stupid. Who's listening to boring dashed for, like, more then 5 minutes? Instead, I decided to do it different.. I send the message all teh time but slow. It has some advantages. Firts, possible audience will know pretty fast that it's a message and will listen to it until the point it starts to repeat. Second, I can send the message more slow, making it easier to read when with a bad SNR. I got to have a close eye on the available charging capacities during dark winter days.

Maybe I won't stick to "space is three times the time of a 'dat'". Hell, since the message is simple I guess everyone can tell the words apart.

Anyway, I got a 12V/7,2Ah battery. It offers enough buffer for really bad days when, for example, it snows a lot. For those interested, here's the way I calculated my signal timing in regard to the energy consumption.


MESSAGE =
XXXXX SOLAR POWERED NIGHT TX ONLY 1WTX UBAT 125
     

TX active time (per message)
----------------------------------
. = 61 (* 1s) =~ 61s
- = 63 (* 3s) =~ 189s
==================================
TX(on) = 250s
==================================


Message duration
----------------------------------
+Sign AND letter spaces (5s each) = 124 (* 5s) = 620s
+ TX(on) = 250s
==================================
Duration = 870s
==================================


Message per hour
----------------------------------
3600s / 870s =~ 4,20
==================================
Messages / h = 4,20
==================================


mAh / h consumption
----------------------------------
(100% / 3600s) * 250s * 4,20 = ~ 30%/h
(240mA/100%) * 30% =  72mAh

+ 16h (winter) * 72mAh = 1152mAh
+ 24h * 5mA (system) = 120mAh
==================================
(Winterday) = 1272mAh
==================================


Worst charge condition
(The case of a snow covered module is ignored)
----------------------------------
Sun radiation in winter = 50w/m2
= 5% from Summer (1000w/m2)

5% from 80W = 4W
I(charge) = 4W/17V = 235mA
Charge hours per winter day = 5
5h * 235mA = 1175mAh
==================================
Charge per winter day = 1175mAh
==================================

[ff]

I love your attention to detail, Zazzle!  I have never had much luck figuring things so closely - Mother Nature always shows me my folly of thinking I can figure everything out to the decimal point.  One thing I urge you to do: you said in your first post you can bend your crystal freq +/- 20 KHz.  I would bend as far LOWER of 10 MHz as I could get it.  If you can get down to 9980, that's great.  If not, get as far from 10MHz as you can.  There's several time standard stations around the world operating on 10 MHz and you want to get out from under their interference.  For my thinking, 9995 is just too close.  I think you could find several cheap crystals in the 9-10MHz range that might be an even better way to go.

I think its wise that you are going to run a continuous message/ID.  Leaving spaces will lessen your chances of others hearing you.  Also, beaconeers that I am familiar with here use slower code.  As you say, it improves readability, and at your power level, your signal will be a challenge to copy under most conditions.

Are you using a PWM charge controller or a reflexive (switches on and off at preset battery voltage levels) unit?  Why I ask is because although PWM is more expensive, they tend to better use the available hours of sunlight, leading to a better charge.  From your calcs it looks like you are keeping the charge/discharge rates well below C20, which is good.  Also, you might want to consider a time out feature that would cut off your beacon after 6 hours of on time.  That would be helpful, especially in the winter sun months.

Finally, may God bless you on your planned climb.  I was still climbing tall trees in my early 40s but now that I'm almost 60, I realize I was running on stupidity and dumb luck.  Climbing is for kids... please be careful Zazzle!  We radio nuts need to watch out for each other.  Good luck with your project... 73!

[Zazzle]

Hi there!

I love your attention to detail, Zazzle!  I have never had much luck figuring things so closely - Mother Nature always shows me my folly of thinking I can figure everything out to the decimal point.  One thing I urge you to do: you said in your first post you can bend your crystal freq +/- 20 KHz.  I would bend as far LOWER of 10 MHz as I could get it.  If you can get down to 9980, that's great.  If not, get as far from 10MHz as you can.  There's several time standard stations around the world operating on 10 MHz and you want to get out from under their interference.  For my thinking, 9995 is just too close.  I think you could find several cheap crystals in the 9-10MHz range that might be an even better way to go.

I'm a perfectionist. Sometimes it's a bless (like in my old job as an engineer) but sometimes it's also a pain because I can't stop thinking (what causes problems when it's about overthinking social stuff). Heh!

Understood. But it makes me a wee bit unhappy because I'd need to re-calculate most parts and I also need to re-wind inductors. AFAIK there are some odd 7.xxx MHz crystals in my xtal box. Or I use some 3...4MHz one along with a frequency doubler. I'll check that. But yes, if nothing works I'll try to tune it away from 10M as far as I can. I just checked frequency charts. 10MHz seems to only go for the US and not Europe. But anyway, interfering with government services is a NoGo.

Edit: "Crystal penning" is going to be my friend :)

Are you using a PWM charge controller or a reflexive (switches on and off at preset battery voltage levels) unit?  Why I ask is because although PWM is more expensive, they tend to better use the available hours of sunlight, leading to a better charge.  From your calcs it looks like you are keeping the charge/discharge rates well below C20, which is good.  Also, you might want to consider a time out feature that would cut off your beacon after 6 hours of on time.  That would be helpful, especially in the winter sun months.

Naa, as simple as possible. I think I have a cheap reflexive charger somehere on the attick (just the board, without chassis). If I can't find it I'll go with the charger circuit attached (just scribbled it down, should work fine). Of course PWM would be the best solution when aiming for the best efficiency. On the other hand, the project should be as cheap as possible (since it's prone to get lost). I decided to kill the energy problem with a 80W module (got a verrrryy cheap one from ebay, one cell has a hot spot. I short circuit that one. The remaining 35 cells should work fine). A PWM charger is about 80+ bucks over here. Too much for the project. The panel was less then 15 bucks, shipping included.

C20 goes only for discharging. Lets say the battery is really drained during a winter day and the next day will be sunny, then it charges with like 3-4A. That's okay since I'm okay with, like, 1Ah missing to full charge due to the fast charging of the first 90%. In case I use my own circuit it won't cut off the charging but goes into "CU (constant voltage) charging mode", what allows the battery to keep charging at a lower current.

About the cut of: I go with a different solution: The µC will monitor the battery voltage and if it drops below 11,5V it'll not start another transmission after tehe running has finished. When the battery is back to 12,5V it starts transmitting again. The battery will wear of when staying not fully charged for a longer time but I'm aiming for broadcast-time, not a as-long-as-possible battery life. I can change the battery once a year if needed. The battery I use isn't a new one anyway.

So far I'll try to keep the costs as low as possible. It will keep me from cursing if the project gets's lost... much. Most time I'd spend on the software and circuit. And that I've on my PC so a second transmitter can be build way faster.

Finally, may God bless you on your planned climb.  I was still climbing tall trees in my early 40s but now that I'm almost 60, I realize I was running on stupidity and dumb luck.  Climbing is for kids... please be careful Zazzle!  We radio nuts need to watch out for each other.  Good luck with your project... 73!

Friends say I have that special grin on my face that tells people I'm up to no good. But well, since I still have one year until I hit the 30 (and start to get officiall old)... lets do this! :)

Yours,
~Zazzle

[ff]

It all sounds really good Zazzle!  Please keep us posted on your progress and final frequency choice.  Many here will want to put your beacon freq on their check list.  And since you're 29, you really ARE a kid, at least for a little while longer.  Still, be careful!  You don't want to scare this old man, do you?    Good luck to you and your project...

[Zazzle]

Hey,

I'm not dead! Nor is the Project.
 I just got the hell busy with RL stuff and that made my interests shift a bit towards other projects.

Anyway. There's progress:
- The Software (for an Attiny25) is almost done
- The HF part is done (see picture)
- The µC board is also almost done

Question: I got a Cystall for 10.240MHz. Suitable? I found no services I might affect nor seems this frequency crowded. I also have 10.110 Mhz Crystalls but I like to forsee from annoying the CW'ers.

I also changed te type of Antenna I'm going to use. I've build a MicroVert for 10.240Mhz. Works fine and solves my problem with long wires.

Best Regards,
~Zazzle

[Zazzle]

Hey,

I got a question. With about 1W TX-Power I have to keep an eye on my Battery balance, especially during winter. I aim for long operation hours, therefore I have to make a trade-off between the following parameters:

- Length of signs: the longer the sign the more power is consumed.
- Space between signs: the longer the space the more likely that the Beacon gets overheard.
I can't stick to the usual ratio of 1:3 between DIT and DAT because the power consumption would be the same, regardless of long or short signal duration.

So I need your opinion on the following timing:

DIT = 0,7s = 5
DAT = 2,1s = 15
SIGNSPACE = 5s = 50
WORDSPACE = 10s = 100

My timer allows 100ms steps.

What's your opinion?

With best regards,
Zazzle

[ff]

I need your opinion on the following timing:

DIT = 0,7s = 5
DAT = 2,1s = 15
SIGNSPACE = 5s = 50
WORDSPACE = 10s = 100

My timer allows 100ms steps.

Good to see you're still with it Zazzle and nice looking build!  I have precisely ZERO experience with beacons so please regard my comments as opinion.  To me, your dats are a bit fat.  I'm guessing that dit=5 and dat=10 are long enough.  The wordspace=100 will probably be OK but I'd shorten up the signspace, maybe to 20 or so.  Just my take on it...

[Zazzle]

Hey everyone!

ah, f# this. I finally got this project done.

Of course. I wanted it Solar powered. I wanted it to not only transmit the Locator but also the Battery Voltage and Temperature. And it should include a good charge controller and a low-battery protection function and so on. In the end I got frustrated because it always felt like spending too much work on a project that gets liekly stolen/damaged anyway. Also, after observing the mount of sun (and therefore energy available during winter during the last monts) I came to the conclusion: even a 240W Module would not provide enough energy during winter to have the transmitter operating during night under every conditions.

I reminded myself of a simple but remarkable quote from the movie Contact. Ted Arroway: "Small moves, Ellie, small moves."

So I started programming the AVR controller doing just one thing: providing the key-ing signal. I got my motivation back after that and spend my last Saturday getting everything ready.

I forsaw from using a Solar powered system (but I have plans for a new one. 7MHz this time. Solar powered. But I'll disassemble it during winter so it works fine with a 20W module during Summer. Small enough to hide it).

My location of choice is now an old industrial ruin around the corner. I know every corner of it and it's close by. Perfect. AND there's grid power available from a pumping station that decontaminates the groundwater. The cables run all across the areal.

So, it's on air since Sunday afternoon. It broadcasts: "<LOCATOR> LIFE IS COMMON LIFE IS PRECIOUS <LOCATOR> 1W TX ALL DAY LONG" [RETURN]

The Frequency is 10.240MHz (okay, 10.238,5MHz exactly)

So I called this station "Common and Precious #1". Now, let me spam this thread a bit so I can post more pictures

The pictures show the transmitter in its weather proof housing.

• The PSU (cracked open wall plug PSU. Modified from 12V to 14V.
• The Attiny25 based Key-ing unit.
• And the Transmitter with Oscillator- and PA-Stage.

Greetings,
Zazzle

[Zazzle]

Murphy is not my friend! After having it running with a dummy load for days (in the fridge and in the stove to check how it behaves with different ambient temperatures) it suddenly started to run postal and oscillate like mad after adding a shield. But some blocking capacitors and lowering the amplification of the oscillator stage solved the issue. I now have a small intermodulation because of stray capacitance but the double PI-Lowpass takes care of that. An FFT analysis shows: harmonics are within the limits and the output looks fine.

[Zazzle]

Finally, photos of the setup.

Sunday. Installation. First problems:
The signal strength around on the receiver side drops to S1 within ~300m. (Icom IC-90, Set ti AM, Narrow). At home,~5km air line, I got a S1 signal with my NRD-525 and a large wire dipole (in the garden) as well.

Wednesday. Checking for problems. Changing the position.
I went on the roof again and mounted the transmitter on a 3 meter long wooden slat The antenna is now installed as a "wannabe Inverted V". The angle is quite low.

The Signal got better. S3 300 meters away and S2 at home. I also noticed that I may have cut the Dipole wires too long while preparing everything at home. I've calculated 6,85 meters per side. It looks like I accidently cut them to 7.9m or maybe even 8.9m. Since I had no folding rule along... a check is on my list for today. But it would explain the low emission. Feeding point mismatch. I may also add another meter to the wooden slant.

The voltage at the output measures around 8,5Vrms into a 50R Dummy Load, which equals approx 1,5W HF.

Geeky me. I also took a handheld DSO along yesterday. The signal on the Transmitter output looks fine while it's working. Clear sinewave. Voltage stays around ~8Vrms

But I'll open another thread on the antenna issue since I have more question on that topic.

Greetings,
Zazzle

[Zazzle]

Morning!

So, I finally had a look at the Antenna last weekend. Indeed, I had cut the Dipoles way too long. Approx 9m instead of 6,85. No idea how that happened. The signal increased (especially in quality) after cuting both wires down. I also changes the support wires to Nylon.

Still, I'm not satisfied with the field. The next thing I'll try is a 1:1 BalUn. Feeding the Dipole with a balanced Signal may help.

Greetings,
Zazzle

[Zazzle]

Good evening!

It's been a while. Projects at work. And not to mention that spring's here. The garden wants attention as well. Anyway. Time for some updates!

I've been on the roof today night and disassembled the Transmitter. That was fun. Obviously, people have been up on the roof. There was a chair (which hasn't been there before) and a - believe it or not - big Dart disk. Funny thing. People have been playing Dart up there. That's why I love this city. Silly people over and over. On the other hand those people have shown no interest in damaging the Transmitter/Antenna/etc. Which I'm grateful for. Psysically and function-wise the Transmitter survived the first weeks of operation. Good!

There are some things on the list that I want to try out on the workbench and in my backyard. The latter regards possible Antenna options and a better impedance matching.

• I ponder to change the frequency to somwhat around 6.950-6.980Mhz.
• I'll redo the 5-Pole PI-Impedance matching and harmonics Filter completely. The goal is to add vari-caps which will allow me to adjust the output impedance of the filter without the need to disassemble the whole thing.
• The fixed antenna gets removed. The Transmitter and Antenna will be seperated from each other instead in future (BNC connectors). That will allow me to loop a SWR/Power-Meter into the Antenna line and read the value on the roof. I really want to know whether it's an impedance matching or location problem.
• Additionally, I'll try another type of Antenna. The German name is "Fuchskreis". I have no idea what the correct english term in is. It's basically a  λ/2 single pole Antenna without ground that's feed from a parallel resonant circuit. The key feature is the easily tuneable circuit.

Stay tuned.
~Zazzle

[Zazzle]

Hi there,

I came to the conclusion that my setup works properly. The signal looks fine. The field is strong (hence, yes, RF is emitted) and other readings look fine, too. So, I guess the Antenna position is too low and the city around interfers with the signal.

Last week I put the transmitter back into operation. There's no use in having it rotting at home in the basement. So it can rott on the roof and do some work meanwhile. Maybe, eventually, in case fortuna likes me, someone will hear it. ;p

Maybe, in case I'm bored and feel funny, I may try a 1/2 lamda Dipole.

The exakt frequency is 10.237,74MHz. Changes a bit with the ambient temperature.

So, done so far.

Greetings,
Zazzle