Just a super quick random project before I go camping this week - trying to build an additional area floodlight. I have 2 super powerful ones already, but I want a high CRI light as well. Because I can
I'm trying to use odds and ends I have laying about. These are:
3 triple 219 emitter PCBs. They are wired in parallel.
3 NANJG 2.8A 7135 drivers.
Is there anything wrong with powering each triple star of a single NANJG driver, and connecting all 3 drivers in series to a 12V SLA car battery? Seen as BAT+ goes directly to LED+ on 7135 based drivers:
So in the case of DRIVER_1 and DRIVER_2, their GND connection will actually go to the LED+ of the next LED group (remembering the LEDs in each group are in parallel).
the first driver will see battery voltage - the first triple, so around 9V which may well kill it. Some people have run a 2S light off a nanjg driver and 2S li-ion, but the voltage seen by the driver controller is probably ~7V or so when sag is accounted for.
You’d be better off doing the Poorman’s method (using the LEDs to drop the voltage before it hits the driver), but you’d probably need to run 4 LED groups to make that work.
Probably not enough time before this weekend, but a nice buck driver would be a much better alternative
What is the total Vf of all 3 boards wired in series? If it's close enough to the supply voltage, just use one driver. If you need multiple modes use the zener/resistor mod to set the MCU's input to ~4v.
These drivers don't work like that, they are essentially two devices on a single board. First part is the current regulation, which is from the 7135s, and are only in the path between LED- and GND. Second part is the MCU to control modes, that is the only part of the driver that needs a connection to the positive side of the power source. They are not like DC power supplies where you can use the positive from one to serve as the ground reference for the next one and get a higher voltage than a single one can provide on its own.
If the total Vf of the string is 9.5v, and the supply is 12.4v, it can be done with the three 3-up boards in series and a single 105C. Especially if this isn't restricted to a cramped flashlight format and you can add some effective heatsinking for the 7135s.
Just use the exact same parts we're using with the 105C/MTG2 builds - remove polarity protection diode & replace with 200 ohm '0805' resistor, and place a 4.3v zener diode between MCU pin #8 (the power input pin) and ground; most convenient location is stacked on top of the ceramic capacitor on the spring side of the driver, though it can go anywhere in the circuit that's connected to MCU pin #8.
What kind of 12v battery is this? Does the light get used while the battery is being charged? What's the actual min-max voltage? ("12v battery" isn't good enough)
Ergh I should know all this stuff. This is what I get for coming up with ideas just before I go to bed lol.
It's a sealed lead acid car battery and I don't intend on running the light while the battery is charging. 12.8V charged, 11.8V flat. It's worth noting that I never let my SLA batteries discharge below 50% (12.2v) as I regularly check them with a DMM.
If each group of 3 LEDs is being driven at 2.8A (933mA each) then the Vf per group is 3.4V. If I string 3 of these groups together that's a Vdrop of 10.2V before it hits the driver. This leaves between 2.6 and 2V for the driver to run off. The Attiny13A can run down to 1.8V. It's close but it might work.
I don't quite understand why you are suggesting the 105C/MTG2 build method? If I drop the voltage far enough prior to the driver (as described above) then there won't be any voltage issues. Of course if I use lower modes the Vdrop in the LED string will also drop, although at 100mA per LED the string will still be 8.4V, leaving the MCU up to 4.4V to deal with...which it can handle. Regardless, I do have the required zener diodes and resistors although my resistors are only 0603 0.2W.
No, the MCU gets its power from BAT+ and BAT-, unlike the 7135s. It will see the full input voltage, the 7135s only see the diff between Vin and Vf. MCU has to be at between 3 and 6 volts to survive. A 4.3v zener will bleed off to ground anything above 4.3v, and adding a resistor in the MCU's power feed limits the current that will flow thru the zener to GND. It's just a crude voltage regulator.
EDIT: I'm forgetting about the 7135 chips here I think...they only run down 2.7V :(
[/quote]
Yes, correct layout. But...
...honestly, I would use one of those cheap buck regulator boards, set the current limiter and then ignore it, and control it with the voltage pot. In fact, I'd replace the onboard voltage pot with a bigger external/remote wired one, with a fixed resistor in series to set the high end limit when the pot's turned all the way up.
Yeah I think it's normally 14.4V when actively charging. An SLA battery fully charged should be around 12.8V.
I'm only doing this mini-project because I leave on a camping trip this Saturday. I don't have time to order any new parts so I'm trying to use what I have here. I just happen to have 3 Nichia stars, a handful of NANJG drivers and the parts required for the MTG2 mod (thanks to DBCstm for that lol).
Place the first two stars in series with the NANJG driver, then the 3rd star in parallel with the NANJG driver. Worst case is the NANJG will see 6V (12.8-3.4-3.4), so I can modify it using the MTG2 method to ensure it's safe operation.
:~ :_(