This should be called Vers III, since I've ripped up the layout twice now and started over, each time with a different design. I'm about to send this gerber off to Seeed, wait 3 weeks, then get to soldering. I'll be doing the software during the wait time. Wish me luck!

For those interested, what this does is to simulate a rotating beacon for a miniature lighthouse. The LEDs are pulse modulated so that as one LED is increasing in brightness, the one next to it is at full brightness, and the one next to that is decreasing in brightness. I plan to allow programming the overall brightness and rotation rate by shinning a spotlight on the solar panel (at night) in a series of flashes that the processor will interpret as programming commands, kinda like morse code only really simple. The lighthouse will be floated on my pond on top of a simulated rock - a well pump cover with 1-gallon plastic bottles at each corner for floatation. My last try at this resulted in my contraption sinking but I managed to salvage the lighthouse before it went bye-bye.

I'll be back to this thread to report on the results, with pictures & movies. Comments/criticism/suggestions welcomed.

 

The LEDs are 25mA max and a PIC chip will deliver this without any driver transistors and the circuit will be much simpler.

I have done the same thing with 16 LEDs and a PIC12F629 and the dimming etc will not be noticeable as POV takes over when the LEDs are rotating.
See my Rotating Beacon project on talkingelectronics.com website.

 

The LEDs are 25mA max and a PIC chip will deliver this without any driver transistors and the circuit will be much simpler.

I have done the same thing with 16 LEDs and a PIC12F629 and the dimming etc will not be noticeable as POV takes over when the LEDs are rotating.
See my Rotating Beacon project on talkingelectronics.com website.

Hi Colin. Good to hear the PWM dimming won't be noticeable. I'm using a PWM frequency of 125 Hz with 32 brightness levels so I think no one can see that even if the rotation is stopped. And I considered using the port pins to drive the LEDs directly, but with a maximum of 100ma total port current I would be getting mighty close to the S08's limits. The LEDs are rated at 30ma continuous and 100ma peak, so to give myself an escape route I chose to drive them with discreets just in case I needed to up the current beyond the 30ma that I'll be using initially. I just didn't have the room for a driver chip(s) for the LEDs unless I used something so small I couldn't solder it, or put them on the back of the board... neither of which sounded like a path I wanted to go down.

I looked for your Rotating Beacon project at that web site but all I found was a 3-LED beacon for model RR. In your project, how many LEDs did you have on at one time, and did that look just like a mechanically rotating beacon? I'd love to see some pictures of your project, and better yet, a movie.

 

It is called Robot Beacon

http://www.talkingelectronics.com/projects/RobotBeacon/RobotBeacon-1.html

 

Very interesting. So tell me, does it look like a mechanically rotating beacon or can you detect that the LEDs are blinking on/off in sequence like in the picture you posted? What I wanted to do was to make it indistinguishable from a real rotating beacon and I thought that 1 LED on at a time wouldn't get me there. So I'm having 1 LED increasing in brightness while the one next to it is at peak and the one next to that is decreasing in brightness. It may be overkill but I don't charge for my time. Like this:


My first version of this thing was to multiplex the LEDs in a 4x4 matrix. My 2nd was to use 2 shift registers to drive the LEDs. The 3rd (and I hope final since Ive already sent the gerbers to Seeed) version is what it is now.

 

My 20,000mcd LEDs (5 cents each) are designed for 17mA and looking at them is toooooo much for the eyes. They will "wash-out" your eyes. You have to keep the current much less. The speed of rotation creates POV and this is about 200mS. Simply build the kit and see the results. If you want a lighthouse, use my circuit. If you want an emergency truck yellow flashing light, use less LEDs in the sequence.

 

The human eye is very sensitive to changes within the visual field, so LEDs that go from full-off to full-on in nanoseconds or microseconds will be very noticeable. I vote for Wingsy's approach, although it is a large increase in the programming required. We went through this in another thread with a guy who wanted LED tower lights to fade on and off like the big incandescent lights on TV towers. He didn't have a moving pattern to contend with, but the optical requirements were essentially the same.

ak

 

Now THIS is exactly the effect that I want. I've only got 8 LEDs on the breadboard but you can imagine 16 in a circle. I think that from a little distance you won't be able to detect that there are multiple lights.

 

... LEDs that go from full-off to full-on in nanoseconds or microseconds will be very noticeable.

I'm not sure what you're saying. I'd say the opposite is true. It's how persistence-of-vision works. You need changes slower than 100Hz to have any hope of seeing it. That's 10 milliseconds, not nano or micro.

 

I'm saying that the human eye can detect very easily the difference between incandescent and LED turn signals, and all that that implies.

ak

 

Now THIS is exactly the effect that I want. I've only got 8 LEDs on the breadboard but you can imagine 16 in a circle. I think that from a little distance you won't be able to detect that there are multiple lights.

A variation of a Knight Rider display, which also depended on the thermal lag of an incandescent bulb for part of its visual effect.

ak

 

Yeah, my porch lights and the white beacons on radio towers are obviously pulsed. Just look left & right of the light and you'll see the strobe effect. You'll probably see that on the lighthouse too since it's pulsed at 125 Hz. But you have to try to see it. I don't think it will be obvious. I can't see a strobe effect on those LEDs on the workbench but they may be too small and clustered in one place.

 

The human eye is very sensitive to changes within the visual field, so LEDs that go from full-off to full-on in nanoseconds or microseconds will be very noticeable. I vote for Wingsy's approach, although it is a large increase in the programming required. We went through this in another thread with a guy who wanted LED tower lights to fade on and off like the big incandescent lights on TV towers. He didn't have a moving pattern to contend with, but the optical requirements were essentially the same.

ak

Actually the software only took 35 lines of assembly code in the main program and 25 lines in 2 interrupts, and a SIN table of 32 bytes.

 

Hello,

Do you know the kight rider style lights:
https://forum.allaboutcircuits.com/threads/project-knight-rider-style-sweeping-light.26977/
You could make it go in one direction and have the fading affect.

Bertus

 

Got my lighthouse PCB from China a lot sooner than I thought I would. This is it running on the bench:

And this is it at night on the bank of the pond:

I can't tell if it is a mechanical beacon or not. That's what I was after. It's running at full power since I haven't gotten around to testing the remote control of the rate and intensity yet.

 

I can't believe my last post was almost 2 months ago. Real life sometimes interferes with stuff I want to do.

It's all done, in the pond doing it's thing, and I'm happy with it. The flotation turned out to be way more of an engineering challenge than the electronics but it turned out to be quite stable once adjusted, and if the turtles stay off of it. Here's a few pictures of the final result.

Those pictures were taken before I made the final adjustments to its vertical alignment. If you look closely you'll see the lighthouse is mounted to a round piece of plywood (sprayed with black FlexSeal) and the plywood is supported by 3 long bolts from under the fake rock. The nuts under the plywood can be run up or down the bolt to adjust its tilt. (Which can create an argument between males and females as to which way it needs to be moved to get vertical.) The thing is so tall that just the slightest thing will get it off vertical. Like even rain water standing in low places on the fake rock will tilt it noticeably. Fixed by drilling holes in those low places. I found that if the jugs were not held rigidly in place they would move slightly and throw the balance way way off. I finally cured that problem by wrapping the jugs in chicken wire and the sewing the cage to the rock with aluminum wire.


For my next project I think I'll stay more on the engineering side of electronics rather than mechanical.

 

Really cool project! I really like how it turned out.

 

Well, damn. This morning at 5AM the lighthouse was working perfectly. Then at 8AM I see this:

Capsized!

I tested this thing for strong winds by tilting it at least 45 degrees and it snapped right back to normal. Plus, there was no wind this morning. So what could have happened? My guess is that some rather heavy critter tried to climb up on it and pulled it over. Groundhog? Fox? Coyote? Sea monster? A really really big turtle?

The battery was dead, probably discharged through the water. After drying everything out and a battery recharge the circuit still worked. But now, what to do? If I just put it back out there like it was, the same thing will happen sooner or later. I'm considering going back to a previous design using a 6ft rod, bricks on one end and a 5-gal bucket of air on the other, and lighthouse on top of that. A very unwieldy contraption, but it was working right up until I accidentally drilled a hole in the bucket and sunk the entire thing. (Which cost me 2 days in bed with a sprained back.)

This brings the score to Lighthouse:3, Richard:0. But I'm not even close to giving up.

 

Could you suspend some ballast under the base, maybe 2 bricks 2 ft. below base & supported by 4- 1/4 in. rods from corners ?
Add more flotation bottles if necessary..
What lighthouse is this modeled after, it is very similar to mine taken from a t-shirt & now also found on some VFW literature.

 

Yeah, it would take very little weight on a rod hanging below to stabilize against capsizing. Even a weight on a string would have probably prevented this but a solid rod would be best. The condition you found it in is properly called "titts up". As in, "It flipped ass over tea kettle and is now titts up."