Reader [pscmpf] really digs the scrolling light look of old marquee signs and as soon as he saw some Christmas lights with G40 bulbs, he was on his way to creating his own vintage-look marquee arrow.
We must agree that those bulbs really do look like old marquee lights or small vanity globes. [pscmpf] started by building, varnishing, and distressing the wooden box to display the lights and house the electronics. He controls the lights with an Arduino Pro and an SSR controller board. The 24 lights are divided into ten sections; each of these has its own solid-state relay circuit built around an MC3042 as the opto-coupler, with a power supply he made from a scrap transformer.
[pscmpf] shares some but not all of his code as it is pretty long. There are five patterns that each play at three different speeds in addition to a continuous ‘on’ state. In his demonstration video after the jump, he runs through all the patterns using a momentary switch. This hack proves that Arduino-controlled Christmas lights are awesome year-round.
+1
Point the way… ;^)
Looks a bit like this awesome arrow: http://hackaday.com/2014/01/05/retro-chaser-sign-lights-up-your-life/ . I actually first thought that this post was a repeat of that, but no – this is a new creation. :-)
Two of the same on Hackaday in the same year. Time to invest in the antique bulb sign market ;-)
Arduino?
Sigh, back in the old days you just use a 55 with a 4017 decade counter (or two) while listening to Gramps telling you all you needed was a bunch of cams & contacts (which is how they’re still made)
You don’t really need the zero-crossing opto either.
Someone should point out that the bulbs last longer if you keep them warm, you have it so there always just a little bit of current flowing through the bulbs. Makes them turn on faster too (not that you’d notice).
Well, some of us aren’t engineers (yes, this might be shocking news) or have studied something electronic related.
Do you just blast them with current when they’re on and that raises the average temperature?
You add a resistor to each bulb so there is always a small amount of current going through it, that keeps the filament warm.
The relay bypasses the resistor allowing the full current to flow.
It’s the initial inrush of current that blows incandescent bulbs, when cold they have low resistance and so the starting current is much higher than what it would normally run at. When hot the filament resistance increases and so the current drops to spec.
That’s why bulbs ‘pop’ when you first turn them on, and why people think they suck (yay LEDs).
Advantage of a microcontroller is that you can have arbitrary blinking patterns. Also, you can use PWM to keep the filaments warm like you suggested, and still have the option to turn them off completely.
Arbitrary control is a bit pointless when their shaped like an arrow, but yes, that can be handy.
PWM and AC current don’t get along easily. In this case that zero-crossing opto is going to screw up PWM (he mentions he tried PWM and it didn’t work in the article – that’s why).
You’d need to add circuitry to detect when the mains crosses zero, and time your PWM to match that. That’s how light dimmers (the good ones) work.
I see a transformer, so the extra circuitry for the zero crossing detector could be done with a single resistor. And of course, the SSR controller needs to be replaced with a random turn on type.
Use 2 resistors, and remember they’re probably not mains rated. I prefer a bit of isolation between my mains & uC, you can get chips for this or simply use and opto, using the mains to drive its LED.
But once you have it you simply turn the triac on when you detect the zero-crossing. You don’t really need to, but it does eliminate the noise.
Resistors don’t have to be mains rated if you take the AC from the transformer that is already in the circuit.
Dunno why I assumed you’d use the primary, brain-fade I suppose.
Thinking about PWM I’m not sure it’s worth the trouble. Say line frequency is 50Hz, that means the shortest period the bulb can be on is for is 1/100th of a second (with switching at the zero-crossing points).
You’d have to work out how many pulses per second are needed to keep the bulb warm, say 1 in 10. Even then I’m not sure the bulb would be happy with that, it might not heat up enough or cool too rapidly to prevent the inrush current shock.
Might be worth playing with, though resistors are much simpler (and probably better). That doesn’t stop you from using PWM in addition to that to set the bulb brightness, no doubt someone has tried it at some point.
this would have been the ideal opportunity for an animated gif :)
animted png’s are the wave of the future. Nice build. Always interesting reads in the comments, in this case the discussion regarding some current going into the bulbs to extend bulb life. I was not aware of that, either, and seems a pretty quick add-on to this project to lengthen bulb life (and keep maintenance costs down!) :)
Keeping the filament warm is one way to extend bulb life.
There are other ways of extending their life, though not really relevant for a sign like this.
Bulbs eventually wear out when run at spec, You can run them at a lower voltage (eg 10v not 12v), use a high voltage bulb (eg 130v in a 120v socket), add a resistor in series to lower the current slightly etc.
The ‘bulb savers’ were often just a diode, this means the bulb is only ‘on’ half the time. That doesn’t mean you use half the power nor get half the light. During the ‘off’ cycle the filament is still warm, and so still putting out light. Since it’s warm, the next ‘on’ cycle doesn’t have the high current as the resistance is higher than at cold.
You maybe get 80% of the light. You might need to swap bulbs, eg replace the 40w one with a 60w to get the equivalent brightness. (This won’t save much power, but it’s the bulb itself we’re trying to save).
Some versions used a thermistor instead, I’ve never played with that.
If you use PWM then you don’t need any of that, on/off is simply changing the rate.
Hi Tony – yes, I actually used the diode approach to half wave rectifying the AC line voltage. This is a trick I learned from my dad as a kid – he would use them to to make the one lamp we left on at night “half” as bright – I know – not sure that’s exactly true. He would cut one side of the cord, slip the diode’s leads into each side of the wire insulation and then use black electrical tape to seal it up. After seeing how bright the lights were on the marquee – I remembered his approach … and truly it made the lights much easier to look at – not so blinding!
Great comments – thanks for contributing!
I was thinking that if you added BLE and wifi you could have the arrows direct specific people in specific directions.
I had this exact thought to set up arrows for a friend’s bar – but I wanted illuminated lettering and I needed to point two directions (arcade and patio). Glad to see my core concept was solid and this guy has some excellent execution!
now to turn it into a giant game spinner for the most epic game of ! xD
supose to have {insert spinner-based game here} before the !
Why the solid state relays ? Are these running on mains voltage ?
Yes.
The article does mention that a few times.
Ok. I live in a country with 220V mains, and it’s rare to see small light bulbs that work off mains voltage. They are typically only used in ovens and fridges, and have very long and delicate filaments. Not something you’d want in a sign. If I were going to make a sign like that, I’d probably grab low voltage bulbs, and use DC with a small mosfet for switching. Much simpler circuitry, and added bonus of allowing high frequency PWM.
They’ve pretty much vanished everywhere, replaced by LEDs. You can still get them, Google “incandescent string lights”.
You don’t really need PWM, and if you did it only needs to be a low duty cycle (light bulbs are very slow to react).
Are you a kid? Up until about 20 years ago all illuminated signs were either incandescent or neon.
Are you illiterate?
Sure, 20 years ago incandescent bulbs were everywhere. Today not so much.
Looks like an arduwinner to me :D