By design, when the light is shut down in a low voltage state, it does the blink to indicate low voltage, as you say, but does it only for 6 minutes, then shuts down the MCU. So after 6 minutes, it should go into a "deep sleep" (power savings) mode. After the 6 minutes is up, it should turn off the Indicator LED before going into power savings mode.
So, did you wait the 6 minutes to see that occurs or not?
I'm pretty sure I tested this in the older Narsil version, but not in NarsilM, nothing was changed in this area.
However if your cells are pretty low but not in a low voltage state, and you leave the light with the main LED off and Indicator LED on for several months, or maybe weeks, depending, then yes - there is no smarts in the MCU running to determine low voltage, therefore it cannot shut the LED off, so it will over drain the cell(s). This parasitic drain is pretty low though, depending on what LED and LED resistor value you use. I've left several flashlights in this state with the Indicator LED on for months without noticing much battery voltage drop.
Here's the code:
//---------------------------------------------------------------------
// Be sure switch is not pressed and light is OFF for at least 5 secs
//---------------------------------------------------------------------
word wWaitTicks = 310; // 5 secs
if (LowBattState)
wWaitTicks = 22500; // 6 minutes
#ifdef ONBOARD_LED
// If the battery is currently low, then turn OFF the indicator LED before going to sleep
// to help in saving the battery #ifdef VOLT_MON_R1R2
if (byVoltage < ADC_LOW) #else
if (byVoltage < BATT_LOW) #endif
if (locatorLed)
TurnOnBoardLed(0); #endif
_delay_ms(1); // Need this here, maybe instructions for PWM output not getting executed before shutdown?
SleepUntilSwitchPress(); // Go to sleep
}</code></pre>
OK I didnt wait more than a minute in my Test
If it shuts down after 6 minutes thats OK
As I wrote with a medium resistor and 0603 LED the parasitic drain gets really low when the battery is almost depleeted which takes months to be a problem
+1 with Lexel. I think I had the same issue. For 2S, you really have to have the "good" 7135's, and even still, it's risky. The XHP70.2 can be driven at high amps, so maybe it's more likely to fail?
the problem is there are many cheap AMCs that are not even the original manufactor with different quality
What I can dicover sometimes on drivers if I do thermal calibration when I heat the driver up they start to leak and when cooled down they are fine again
often its way below moonligh level, just a few are that bad like you see it
Yes, they still usually work above that voltage, but every once in a while you get one that seems to stay on. (According to other people that have run into this problem)
Yea, I didn't know about this one - definite bug, though it's hard to speculate what 5% and 10% mean on a particular light mod or build, since LED's and FET's make such a huge difference. For example in a 7X parallel XPL2 light, maybe you hit 25A is 100% at 10K lumens, so 10% being 0.35A is much lower than 10%...
I'll have to fix this/change this to something reasonable. Think I was using ~1,500 lumens max (typical max single XM-L2 or XPL moderately configured) as a guideline for rating the percentages.
I'm working on NarsilM v1.1 now (GT Buck driver support merging), so will get it in before I forget, so when it's released, it will be there.
Really, all these mode sets should be tweaked build/mod by build/mod since there's so many variations of build/mods. Unfortunately this is what you get when working with fixed amp output (7135's) and variable amp output (FET's) combined in one driver.
It was discussed to not use percentages, but just state what it is: one max 7135, 20% of a 7135, 45% of the FET, etc.
Yup completely understand the percentages are guesstimates. Just wanted to bring this to your attention since it seems obvious that two modes shouldn’t be identical.
I try to make another use of the NarsilM firmware driver.
Not working properly still.
After installing the battery, direct light high, and then will not light up again.
7135 using A705NGT-350 SOT-89
Flash firmware on Attiny85
FET 404
85fuses.bat:
REM BOD disabled:
Wp hvc
REM BOD enabled at 1.8V: Avrdude -p t85 -c usbasp -Ulfuse: w: 0xe2: m -Uhfuse: w: 0xde: m -Uefuse: w: 0xff: m
85M.bat:
Rem 85NarsilM - downloads NarsilM (Tiny85 Multi-channel e-switch UI configurable)
Rem Avrdude -p t85 -c usbasp -u -Uflash: w: NarsilM.hex: a