Mornin' Guys. New member here so don't trash me too bad on my first day.

I'm a retired military electronics technician but it's been many years since I've been actively engaged in electronics. Nonetheless, I still understand the theory, at least up until the mid-1980s when all of the new and amazing technology started showing up! In a nutshell, my electronics education and experience ranges from vacuum tubes to CMOS digital ICs.

The reason why I'm here is because I need a cheap obstacle ("ob") light flasher for a 40ft tower about a quarter mile away from a small private runway. However, "flasher" would be a misnomer in this case since ob lights don't really flash anymore (at least not at night anyway) but, instead, exhibit a slow rise and fall as if fed by a very low frequency sine wave of about 1 cycle every four or five seconds. I'm looking for a circuit capable of driving a pair of common 60W incandescent bulbs in parallel. My electronics design skills, being as rusty as they are, have me thinking along the lines of a circuit based around a sine wave-controlled triac.

Of course, if inexpensive ob light flashers are already available for the amateur radio market, that would save me a lot of trouble (but also deny me a bit of fun).

Your opinions are solicited.

Many thanks!

Harvey

 

I would think that, to avoid having to climb up the tower and replace the bulbs, you would use a longer lasting form of light such as LED bulbs rather than incandescent.

You could control the brightness with a controller such as this with the DC brightness level signal provided by a couple of 555 timers circuits.
Does that sound like something you could put together if the circuit schematic for the timer was provided for you?

 

I agree with Wally - no matter what the color, LEDs are the way to go. New white tower strobes and runway strobes are all LED now (a friend designs them for Siemens). You can get the ramp-up / ramp-down look with a couple of opamps driving an ebay PWM dimmer.

If you want to go retro, a TRIAC dimmer driven with a sinewave will get you the look you want with incandescent bulbs.

HOWEVER - My guess is that the FAA has something to say about this. Your tower height / runway length / traffic load might fall below their rules cutoff for mandatory lighting performance, but you really should check with them.

ak

 

Hi Guys, thanks for your replies!

I have no problem going with LEDs as long as they'll fit a standard light socket. However, aren't dimmable LEDs still a bit expensive?

And A-Kid, thanks for the heads up about the FAA but they don't get involved with private airports other than insisting that recognized standards (green & white beacon flashes for civil runways, white runway edge lights, blue taxiway edge lights, etc) be maintained.

As Paul Harvey used to say, here's "The Rest Of The Story" (See the attached pic.)

Harvey

 

Here is a dimmable (60W equivalent) LED bulb for a little over U$3 each.

 

You could control the brightness with a controller such as this with the DC brightness level signal provided by a couple of 555 timers circuits.

Does that sound like something you could put together if the circuit schematic for the timer was provided for you?

I'm interested.

Harvey

 

I'm interested.

Okay
Below is the LTspice simulation of a 555 timer circuit with an LM324 op amp buffer that I think will do what you want.
The output from the U2 op amp is a simple rising and falling RC integrator signal, which should give you a reasonable looking beacon signal from the bulbs driven by the Velleman circuit.
Changing the value of R7 will affect the shape of the Out signal.

The power can be provided by a 15Vdc regulated-output AC adapter (wallwart) type supply.

Note that you need to look at the data sheets to determine the pin connections for U1 and U2.
The circuit can be built on a small perf board.
I suggest using sockets for the ICs if you can.

 

Wow, nice artwork and the circuit is well within my capabilities to build. While the rise period looks like what I'm thinking, I'd prefer that the fall period mimic the rise. Your notes said that the shape of the waveform could be changed by varying the value of R7 but would it change it that much?

I've taken the liberty of altering your waveform to illustrate what I'm talking about but I'm having trouble posting an Xcel or Word doc.

Harvey

 

I think I know what you are asking.
You want the fall to be a mirror image of the rise. But that's not so easy to do with a simple circuit.
I'll give it some further thought and get back with you.

 

I'm having trouble posting a picture of my modified waveform.

Harvey

 

Careful, you could step into a pile here. If the FAA does not require you to light your tower it is probably best not to do it. The FAA requires that tower lighting be of a certified type, they won't approve of a DIY setup. If the FAA notices your homebuilt light they can require you to install an approved type ($$$)

 

I typically do a screen grab of the picture using the Microsoft Snipping Tool, copy that in the Snipping Tool Window, and then paste it into the edit area.
Takes about 10 seconds.
See, here's a quick screen copy of your avatar.

 

Brevor,

There's no threat of stepping into a pile here. My Houston FSDO is fully aware of, and is fully supporting my 84 year old beacon tower's erection. (Which, by the way, was not originally equipped with ob lights of any kind.) The FAA only requires "approved" lighting systems when used on public-use airports and even then, they're nothing like what's required for commercial airports.

Harvey

 

I typically do a screen grab of the picture using the Microsoft Snipping Tool, copy that in the Snipping Tool Window, and then paste it into the edit area.
Takes about 10 seconds.
See, here's a quick screen copy of your avatar.
View attachment 110081

I just recently got switched (not by choice) to Windows 10 and I'll be darned if I can find the snipping tool (or anything else for that matter!)

Harvey

 

I just recently got switched (not by choice) to Windows 10 and I'll be darned if I can find the snipping tool (or anything else for that matter!)

Isn't it fun learning where everything is in a new operating system? (Not).
Click on the search icon (little magnifying glass) and type "Snipping tool".
After then launching it, you can pin it to the taskbar so it's always handy (right click on its icon in the taskbar).

 

Is there no way to post a Word or Excel picture to this?

Do a screen print and post that as a .jpg. .png or .gif. Keep the file size down, to allow for folk having dial-up or poor/metered broadband.

Edit: +1 on pinning the snipping tool to the taskbar.

 

Okay guys, I finally got my images to post. Thanks for your help! (I'm really not so clueless when it comes to computers but this new Windows 10 is a bit aggravating!)



Anyway, Crutschow, I've overlaid two versions of my desired ob light modulation waveforms (in red) on top of the one that you posted yesterday. The top one shows a simple sine wave ramp up and down, and the bottom one shows a modified version with a longer full-on period. This second one is very similar to what the ob lights look like on a cell tower that I saw recently. This version is preferred but either are acceptable. Note that both versions produce a modulation profile that's a bit smoother than a simple triangular waveform.

Harvey

 

Unfortunately simple sine wave shapes are not easy to generate without going to a more complex sinewave oscillator.
Below is the 555 circuit modified to give a rough waveshape to what you want.
I doubt that it will look significantly different to your eye than the second wave you posted.
The value of R7 affects the shape of the rising waveform and the value of R3 affects the falling waveform.
U4 adjusts the output amplitude.

 

How about this?

Trim1 sets the wave period, Trim2 sets the plateau width. C1 and C2 are back-to-back electrolytics but preferably would be replaced by a 4u7 non-polarised cap.

@crutschow
Do you have a good model of the LM324? The version I've got throws a wobbly if there are more than 2 instances on the schematic.

 

Here's the model I use and it seems to work ok with more than two instances in a schematic.