First, let me say this. I am no expert. Just another hobbyist perhaps (in your words) a few steps ahead, perhaps. My solution grew over the years and learned as I go. So, it is just "one man's experience" and may not be the best technical solution.
You brought up some interesting questions, I will touch on them - arranged in differ order.
it takes more than 6volts to be bright enough. But then, maybe that's a current thing, not a voltage. Anyway, the strip is fairly long for my application and nothing below 9v has provided enough light to be worth it. Usually it runs at 12v for full brightness.
...
...
but all of it seems to be targeted a plugged in setup.
Your strip (from your link) is set up as units of three LEDs - look at the cut-line and you can see that as well. Each unit has 3 LEDs in series with a ballast resistor to run at 12V and all the
units in the strip runs in parallel.
Not that I suggest you do it, but the way it is made, that ballast resistor is set up for 12V. A different (lower) resistor would allow it to run at lower voltage.
You very well may want to use the plug-in set up, see reasons later in the reply.
So it seems it does what my strip already does - keep current constant so all the leds light up equally bright? That's built in to the strip. I just need to provide enough current for the whole strip to light up.
No,
the strip doesn't keep constant current. My solution doesn't either. As the LED warms up or voltage change, the strip will draw different amount of current producing different level of light.
Each unit of 3xLED can draw different current, but within the unit, all three LEDs will have the same current.
My solution keeps it at constant source voltage (3.3v) and not constant current. The 7135 doesn't control current per-se, it really just limits it to 350mA so you control it by adding another 7135 for another 350mA. Since I am drawing below 350mA, so the 7135 can't do a thing but be a mere on/off switch.
As LED warms up, there is a difference in current draw and brightness. My solution does have the
potential capability to produce constant brightness with minor code modification. Since I already have a photo-resister feedback (for now just used to decide on or off), and since I made it with PWM to control brightness already, I could use the photo-resister reading to keep the LED at constant brightness.
[re: the AMC7135]
So I think that's what has me a bit confused. That doesn't fit the led model I have This seems to be targeted towards single or "special" led applications like flash-lights. What I have is something like this: ...
The 7135 might have been designed for flashlights, I just found it a great little switch to use particularly since it is designed with low-power-waste, PWM & LED in mind as compared to a general purpose transistor for switching.
I choose to use single LED over a long strip for a couple of reasons - main reason is if I need to make it very bright, I can do so in very limited space, and I can aim the lights better comparing to a strip. The other is, as they age, they get dim. Individual LEDs allows me to manage it better (swapping out individual ones, or even change them so edge ones are lower and middle ones are brighter).
My solution is not a clean thought out implementation. It grew over the years. It was (a) to see who is eating the cat food outside, and (b) who is walking the dog on-to
my lawn to let the dog do its business on my lawn!
An earlier version of my solution (with very bright LED and two piezoelectric beeper) scared the dog-walker away. So, now my set up is reduced to cat-watching with the exterior piezoelectric beeper removed as well. At the earlier solution, I actually had the Arduino (with relay module) "press the garage button" to open my garage door and then within a second (before garage door opens more than a few inches) closes it. That simulates my action whenever I catch a dog-walker near my driveway. But concerns over forgetting to turn it off when I am out, I eliminated that as quickly as right after showing it does work. If I left it on while I am out, someone can take advantage of that and walk into my house via the garage - all it needs is just something obstructing the closing and the garage door will stay open. That "deployment" didn't last much longer than the test phase just to "have done it".
If you are not using Arduino, you can just use your PIR to feed the AMC7135's switch pin.
Because I want to log each time the PIR was triggered, I needed something more than just the PIR, ie: Arduino. With Arduino in the picture, I might as well do brightness control (PWM). So, what I did was to use the Ardunio's PWM output and feed that into the 7135's switch pin to turn the LED on/off depending on the PIR sense, and at specific % brightness controlled by the PWM. The Vin is fed from the same Vin for the Arduino, voltage regulated down to 3.3V. Since I did not need so much light (just enough for my USB camera), I get by with a single LDO voltage regulator and a single 7135. The LED+ is with the 3.3V regulator output, the LED- is the AMC7135's OUT pin.
Actually, I one more "if statement" there - before it turns LED on, the Arduino checks the external brightness via a photo resistor. If it is bright enough, it wont turn LED on. Just so I save the battery when it is running with battery power (which I rarely do).
Heh - yeah, you're quite a few steps ahead of me. That's some of the same features I was looking at implementing. But since you're regulating voltage down too (to 3.3v) it seems like I'll be in the same boat again?
Well, regulating to 3.3V is not the issue here. Strip or single LED, PIR will be just driving a switch whether the switch is a general purpose transistor or an AMC7135.
If you are thinking about logging, you will be using at least an MCU (Arduino, I assume) or perhaps even Laptop connection. Between the MCU and the PIR, your battery is going to be driven hard. An Arduino UNO + PIR + LED + RealTimeClock probably would use up the 9V battery within a
few hours if that long. If you are planning on viewing the log without a PC, you will need an LCD screen on your Ardunio...
So, that bring us back to power usage. You just may want to run it off a car-battery! A cheap auto battery would give you
5000mAH 50AH (50,000mAH) at 12V. The 12V transformer (for your strip) can very well supply the MCU/PIR setup as well. Car battery or transformer,
the switching part of the circuit doesn't have to run at the same voltage as the LED light circuit. You can use a
more power-efficient buck converter (as oppose to linear regulator) to step down the voltage to 7V-ish for your Arduino Vin and PIR. If you trust the buck enough, you can step-down all the way to 5V and by-pass the voltage regulator in the Arduino board's Vin. Directly connecting your buck converter 5V out to Arduino's 5V out is risky. If your buck converter fails, your Arduino (and everything else connected to it) may fry. But you do save power with buck as compared to linear regulation so bucking down to 5V instead of 7V will eliminate the
additional waste of regulating 7V to 5V.
My solution simply grew from initial simple deployment. There may be better solution to it such as an off-the-shelf solution with it's own lighting. My solution is not an ideal one. It just grew to be exactly what I needed.
EDIT: corrected typo above. Small/cheap car battery typically is around 40-50AH, I mistyped. Small UPS battery would be around 5000mAH.