Security light powered from mains adaptor instead of batteries.

[Dave Plowman (News]

Obvious answer is your power supply isn't replicating battery power. Wrong voltage, too much ripple, etc.

[newshound]

With your observation of the meter going offscale at 500mA it sounds to me like that is the issue. 150mA at 6 volts is less than a watt, which is not really going to give you much useful light. My standard small torches draw 800mA from 1.5 volts, or 1.2 watts.

Maybe you want to put rechargeables in there, and use that power supply to keep them charged (although it is probably not the best sort of supply to give the batteries long life).

[wasbit]

I'm sure I had this working as it should before I resorted to working outside. Now, whatever I try the light just cycles every 30 seconds. I am going to revert to batteries. Thank you for all your input.

[newshound]

Seems like a good idea to me.

Given that you were going to run a wire to it from a power supply, you could always run a wire from it to a battery box inside containing 4 D-Cells for longer life, and easier battery replacement. But then if you are running a wire, why not just go for a mains light instead? Under £15 for Screwfix basics.

[Adrian Caspersz]

Ripple and noise (switch mode supply?) from the AC supply will trigger whatever circuitry is in there (monostable?), switching the light on for its predetermined period and repeating the same forever.

Have a look at the circuit board and see if you can identify the chip used, and get an application note from the webs. There may be a circuit diagram, showing the filtering components that they didn't bother to fit (it being battery operated).

Or you could play trial and error with a electrolytic capacitor, soldered across the chip power connections. 100 microfarads maybe?

Access to a oscilloscope would be useful.

[wasbit]

Thanks, but way over my head.

[Adrian Caspersz]

maybe not? You can hold a soldering iron :)

It's how much of this Chinese stuff works, and can be hacked about in the name of DIY.

Have a watch of some of Big Clives youtube videos sometime.

Entertaining some of them. Just avoid mucking about with mains power.

[wasbit]

Change of plan. Left the security light on batteries. Found a used £5 PIR LED tape out of the spares box. Can't find a link but Big Clive stripped one down on Youtube -

on 5 to 24 volts so lots of power options. Not sure how well the LED tape will stand the weather. Only time will tell but it's working well for now. Cats & hedgehogs have both been detected.

[Paul]

One thing to consider, is the circuit has two halves. A 3.3V section driven by a linear regulator. Nice smooth clean power. I've not bothered to make a representation of the control block in the diagram below, leaving the open gate lead as a convention for "control drive goes here".

But the load circuit connects directly to whatever power source you are providing. If the power source just happens to be 5V and the LED strip has its R_resistor set for 5V, then your design is happy.

If however, you connect the 5V LED ribbon, then blast 24V in on the left, there is a bright flash of light from the strip and smoke (somewhere). That's because a 24V source would cause too much current to flow in the load. The 5V LED ribbon may not be able to handle a "wide range input".

+24V -------------------------X ??? | \ R \ | \___ Load design must match input LED / power source, as input power | / source receives no further regulation. 5V ribbon, 24V src = No! X | |--+ ----| MOSFET switch (5A max but poorly cooled) |--+ | GND --------------------------+

Now, the LED ribbon, could consist of any number of designs, even including its own constant current source circuit in series with the load LEDs. So anything is possible.

But as the designer, you should at least not be under any illusion this is "plug and play". This circuit is not defensive in nature. Connecting wrong supply with wrong load, equals problems. If you understand the characteristics the load needs, you'll be fine. A 5V ribbon combined with a 5V wall adapter, should give a relatively tame result. You'll notice Big Clive powered his demo off some sort of 5V wall adapter with USB connector. He did not reach for his 24V adapter, for fun.

Hopefully, the device has a spec, for what kind of current the designers think it is safe to switch. The channel resistance is 0.050 ohms or so, I think I heard mentioned, but you might want to roll back the tape and listen to that part of the Youtube.

There's nothing wrong with what you're doing, but don't get too adventurous without digging out your notebook and doing some calcs.

Paul