Need help with garden lights project

[abhi37]

HI,
As part of my school electronics project im doin a project which is - controlling existing solar garden lights with a remote as they are not always needed to be lit. I was wondering if I could use radio transmission to activate the lights when ever i wanted to. I recently did a radio transmission project. The problem with this is that how could i change the existing garden light circuit and if the garden light received enough power to drive a micro-controller. The transmission would be from inside the house.
Any help would be good coz im a newbie at electronics and this is my first year taking it.

 

[HarveyH42]

Well, I've modifying solar landscape lights for about a year now. When I find some cheap and on sale I stock up. The batteries and solar cells are usually worth the $3-4 price.

Past few months the lights on sale have a single AA and drive a white LED, so they also have a boost converter. Haven't messed with one of those yet, still have some older ones to work with, and I know the circuit well enough.

The ones I got here:
https://www.allelectronics.com/cgi-bin/category/565/Solar_Panels_and_Photovoltaics.html

are pretty simple, and a ATTiny13L works great for color fading and sequential flashing. Just take the (+) directly, hole already there, but need to scrap it clean before soldering. The (-) is switched, there are holes for 3 LEDs, best to just use the neg. hole you pulled the light from. The sequencer with red LEDs is fine off the 2 AA supplied, the RGB needs 3 AAs. The solar panel is 2X3 volts in series. Not sure of the current, but will do 4 AAs 2100 mAH, and still be going strong at 2 AM (leave for work).
Sorry, I'm don't do radio (yet), so no idea if it'll work off 4.8 volts (probably the max. you can expect to charge of the panels).

 

[Hero999]

Use infrared or even better PIR so the come on when you walk past.

 

[HarveyH42]

What type of garden lights do you use? The I just guessed at solar. The low-voltage type have a big Xformer, timer, and CDS, on/off switch, so a remote doesn't seem to useful. There are other kinds of lights I didn't consider, but basically the same deal as low-voltage.

Putting a reciever and MCU in each solar light would be expensive. The thing with solar, is that the power is basically free, and letting the batteries discharge is good, otherwise you risk overcharging them.

Must agree that IR recievers would be good. Not sure how you would build an IR transmitter to cover a large area.

I have a future project for controlling multiple lights, but still a few details to figure out. Basically, I want use IR to comunicate with other nearby lights, so when one light is one, the other stay off. The light only stays on for a short period. This way I could do sort of a light-chaser thing across my yard, or other moving/flashing light stuff. Basically, path lighting is okay, but a little visually boring. Color changing and flashing/flickering helps, but I think giving them motion would be cool.

 

[things]

i have seen little ir activated relay circuits somewhere. maybe you could use an ir activated relay (or transistor) to turn it on and off with a ir led as a remote?

 

[gootee]

For simple on/off control, you shouldn't need anything too fancy, inside the lamps, to do it with any sort of transmitted/received signal, whether it's radio or IR or ultrasound or whatever. Once you have a way to transmit, and a way to receive with a tiny, cheap, low power receiver, you just need some type of latching, bi-stable switching circuit, such that if a signal is received, the switch should change to the opposite state. i.e. Click the remote once and the lights go on. Click it again and they go off. Click it again and they go on. Etc.

You'd need to find some device that can switch the actual power to the light on and off. And that device shouldn't consume too much power, in either state. You might also need something to control that switching device, if it needs to be held on or off by a voltage or current, which hopefully wouldn't consume too much power, either. You'd need some circuit that interfaces to the receiver's output, which might also perform the bi-stable latching part. Maybe an opamp-based peak detector circuit could drive a logic input for a two-state flip-flop circuit (no processor or large-scale IC should be needed, certainly), which might be able to control the actual switching device (or its control circuit if one is needed). You'd also probably want to make the logic circuitry "bounceless", maybe so it would only respond once every second or two, to avoid having it flicker on and off multiple times because of mechanical switch effects, or if something else was a bit dicey.

Maybe a small "logic level MOSFET" would work well for the switching of the actual power to the lamp.

You'd still need to figure out what transmit/receive method would be best (i.e. easiest and cheapest to design/buy/build/use, while still being effective-enough and reliable-enough, and, small-enough with low-enough power consumption, for the receivers at least). I'm guessing that some sort of RF (Radio Frequency) might be the best, overall. Basically no information needs to be transmitted. So the transmitter and receiver should be able to be dead-simple. Maybe you could cheat and find some small, ultra-cheap remote-controlled toys and use their transmitters and receivers. Otherwise, if you don't have to design them from low-level discrete components, there are some pretty cheap and tiny ICs that do radio transmitter and receiver functions. You would probably only need the most-basic carrier transmission and reception capabilities. i.e. Yes, there's a carrier signal, or, no, there's not. You could probably even find some receiver IC that would do that ("yes or no") part for you, so you wouldn't have to mess with designing a peak detector. Maybe you should look into what ICs they use for key-fob remotes for car door locks, and for the in-car receivers. There are enough of those made that they ought to be dirt-cheap.

If I were trying to do something like this, I'd download LTspice (aka SwitcherCad), from linear.com, so I could design and simulate it at the same time. I'd also get the printed catalogs from mouser.com, alliedelec.com, and digikey.com , and get familiarized with searching their websites, and those of major IC manufacturers, too.

Good luck!

- Tom Gootee

**broken link removed**

-

 

[HarveyH42]

Almost forgot... Wireless Doorbell. Been wanting to pick one up, but never see them. Remembered they are cheap and don't use a lot of power. Never think of it at the hardware store, usually focused on some tool or material to be used for something other than it's intended purpose...

 

[things]

how about a latching relay, i don't think they use power once they have latched so all you would have to do is get a simple circuit to send pulses to a relay when the signal is recieved.

 

[gootee]

how about a latching relay, i don't think they use power once they have latched so all you would have to do is get a simple circuit to send pulses to a relay when the signal is recieved.

things,

That sounds great, and would remove many of the design problems!

One such relay, for which I found a datasheet at mouser.com by searching for "latching relay 5vdc", says it needs a minimum pulse width of 10 ms. So it seems like maybe he could just let the pulse width be controlled by how long he holds down the button on the remote, unless that might be able to waste too much power, e.g. if the button was accidentally held down for a long time, or maybe even in comparison to having a circuit at the receiver that would just produce a short pulse of 20ms or so, no matter how long the button was held down, rather than a typical button press length of maybe 0.5 sec. With one relay I found, it would use about 100mW for however long the pulse was, each time (see below). He would have to decide if it was worth adding the pulse-length-control circuitry, to save that power.

The relay I happened to look at, which was just the first 5V latching one that came up and may not be the best one for this application, was an OMRON G6JU-series surface-mount type, and has a 10mm x 5.7mm footprint, and a height of 10mm. It can carry up to 1A of current. The coil triggers at >= 75% of 5V (i.e. 3.75V or more), for both Set and Reset, and has a resistance of 245.5 Ohms and a rated current of 20.4 mA. So power consumption when pulsed would be about 100mW.

There are also similar relays that work on 3V instead of 5V, although they use about the same amount of power, when pulsed. And there are through-hole versions for most of them, too.

It might be better to use a relay with a 3V coil, because it could work even for lower (partially-depleted) battery voltage levels, and would make the coil-driver circuitry easier to design, by giving more headroom if the available supply voltage was only 5v, for example.

So, with a latching relay like the 5V one I mentioned (modify the following if a 3V relay is used), it seems like at the receiver end he'd only need a way to produce 4V to 5V quasi-DC that could push at least 21mA, whenever the receiver got a signal from the transmitter, and 0v (or even up to 3V or so) when no signal was present (and maybe limit the pulse length, if desired, as mentioned above). There are probably at least several different ways to do that. But it would probably mostly depend on exactly what types of transmitter and receiver were used.

- Tom Gootee

**broken link removed**

-

 

[Sceadwian]

**broken link removed**

There's a good 'oneshot' both positive and negative edge triggered there that are good for this kind of application.