I wish I could afford to have some boards made, with 4 ounce copper on them instead of 1 ounce.
Sorry, getting off topic.
It the polarity correct for the bottom 2 emitters in the diagram in Post 11? I wouldn’t expect that to work.
Good catch, missed that myself, the emitters must all be + to - with the output wires from the driver feeding the correct ends of the string.
Also good info, I don’t do much with 7135’s any more, a Zener mod FET driver could handle 3s cells just fine (feeing a string of 3s emitters).
Good catch. Looks like wight just edited the black line. Another way to do this is to run power to all three LEDs from all three cells but power to the board from just one. Since the cell driving the board as well should run down slightly sooner, lvp should still work.
Could you make a diagram for that?
Not easily from my ipad but draw a wire from the + of the bottom cell to the driver and + of the top cell to the led string. Led - to board and board ground to B-.
Yeah, my mistake. I should have just drawn a new diagram.
Won’t be easy to tap a single 18650 without a carrier.
I’ve thought about this many times, I have decided it may be doable using some sort of insulator and two wires running to each meeting point of cells, it wouldn’t work in an 18650 tube cause there’s no room for wires but you could fit it in a 26/32mm host using a modified spacer to get room.
Still, not easy to load. I like to drop cells in and go when possible. No real gains here either IMO.
sorry for confusing you guys
i will explain
all i need is to use the user Download to drive 3 XM-L2 with one nanjg driver by a 3 cells in series
so i tried it but the driver fried i investigated and found that the driver at low mod is geting almost 7 volts so i said as long as the zener mod that we do for the MT-G2 allows the driver to take more volts lets give it a try then i did it and assembled everything like the user Download draw well it worked but at high on tail cap i am getting only 1.45 A
so for temp solution i disabled all the mods on a new Nanjg Driver and made it high only mod and used the user download method without zener mod or anything else and it worked perfect . but still if we could do it with the full mods it will be much better
I think something is be getting lost in translation… You probably need to switch over from 7135’s to a FET, 7135’s will not be happy with over 6v input (even with the zener mod, all the zener diode does is regulate the Vin to the MCU, not the whole board, the rest of the board is seeing full input voltage).
These are linear drivers meaning they cant step you or step down, the voltage out is the same as voltage in, to run off 3 series cells you need 3 emitters in series as well.
Also off 3s cells 1.45A is a pretty normal tailcap reading, as the voltage increases amp draw decreases. You need to measure current to the emitters, not current in (aka tail cap current) to know whats going on.
For us to help you more can you please make your own drawing/diagram of EXACTLY how you have it wired, since neither of the pictured diagrams in this thread are correct if your following one of them of course it isnt working. You dont have to use the computer, just draw it by hand on a sheet of paper and take a picture of that hand drawing.
He specifically stated in post #30 that he wants to drive 3s emitters on 3s battery. I corrected my pic in post #11 with an edit. I say there is no reason not to wire it that way, do you disagree?
I’m not sure what you are getting at with an FET recommendation - if he wanted to run 1 emitter on 2s then that would smoke it. That’s not his goal however, so it’s no big deal.
That’s your problem. As I said earlier, you cannot wire the zener mod that way. Look around and see if you see anyone else wiring a zener mod that way. You should find plenty of people wiring it like the diagram now shown in post #11. I will explain in detail if you need me to.
As has been pointed out the 7135’s max voltage is 6v, I recommended switching to a FET because they arnt limited to 6Vin. What I’m getting at is switching to a FET would make it so he wasnt over driving components by 200% (which he’s doing trying to regulate current with the 7135’s at 12.6 input).
I know he’s not trying to run 1 emitter off 3s cells, of course that wouldn’t work but in his 3s emitter / 3s cell setup the voltage is way to high for 7135’s even tho technically it somewhat works, thats why I make the FET recommendation.
I asked for him to post a diagram so we could check his work (since he obviously has some issue still) but also so there could be a correct diagram here in this thread, as of now there isnt a good diagram, what if people find this thread searching in the future, they wouldnt be getting a correct answer from either of the two diagrams posted.
Here is a correct diagram to run 3 3v emitters in series from 3 cells in series.
The dotted line at the top shows alternate LED+ wiring, it can bypass the switch to help lower resistance, in that way the switch would only interrupt power to the driver, the LED + of the first emitter in the string would pull direct from the batt+, no need to go threw the switch (but either way is possible) second diagram shows this way
Same way but with LED+ direct to batt+
Cereal_killer, your diagrams seem to be electrically the same. They also both show an unnecessary switch bypass. It will work, but there is the possibility of a parasitic drain with your wiring.
As I said previously, my diagram in post #11 is correct.
An FET is unregulated, he may not want to beat on his cells and emitters in that way. Plus, it sounds like he’s already got all the parts to build a working light without waiting on anything else… It’s already been extensively demonstrated here that 7135’s work fine above 6v as long as the delta between Vin and Vout is low enough.
Ok yes now its edited to be correct, I wasn’t tho, sorry I didn’t go back and read every post in the thread before posting, I’ve kept up with this and originally the emitter wiring was wrong.
As for the switch bypas causing a parasitic drain, that’s not the case, for it to the driver would have to has a short and the emitters would glow all the time were it to. Especially in a high power mag build its always better to not pull + threw the driver but instead straight from batt+. Here is a link to CPF showing it, its been done that way for years now.
I did specifically refer you to the correction and even asked you whether you disagreed in post #32…
This is definitely not a high-current build and therefore does not require a switch bypass. I’m not sure you are correct about the parasitic drain. There does not have to be a short, that is for sure.
After re-reading the datasheet, our usage with the Zener mod is clearly not a mystery. It is actually within spec driving any voltage of LEDs as long as the voltage delta doesn’t exceed 7v with respect to GND.
that would be nice of you and please if you can draw it
OK, since I drew it hopefully you won’t need a huge explanation. The “download” wiring is not intended for a zener modded 105c. AFAIK his wiring is intended for a driver with no MCU, only 7135’s. When you put the two LEDs in series with BAT+ you get a voltage drop across them. Therefore the voltage seen by the zener mod is decreased, maybe something like 4.3v-5.8v depending on the voltage and charge state of the cells. Assuming you start at something like 4.1v per cell you’ll probably see about 4.6v between BAT+ and BAT- on the 105c. This will decrease a bit over time of course.
With the lower input voltage, current across the resistor added for the zener mod drops. With lower current, zener voltage drops. If you started with a 4.3v zener, you are now getting less than 4.3v. Possibly as low as 3.3v. Uh-oh! The AMC7135 spec says 2.7v Vdd for turn on. The ATtiny13A spec shows us dropping about 0.7-1.0v. So something like 2.3-2.6v on the ATtiny13A’s PB1 (the PWM pin which feeds Vdd on the AMC7135).
