I am a fan of Anduril 2, but I’ve noticed for some time now that on occasion Batt check does something odd. First report might be 3.7V, then the next one is 3.6V. I’ve always assumed the actual voltage was near the threshold, and a bit of juice removed form the cell lowered the voltage enough to go below the threshold. I recently modded my TS10 to reduce the Aux light brightness, and I’ve been periodically running a Batt check to see if the voltage is decreasing. I’ve noticed yesterday on this charge of the cell, that I got 4.2V once when the light was warm from being in my pocket, and then next round the report was 4.1V. Well, today I ran Batt check (warm out of my pocket light/cell), and the first report was 4.3V! The next two Batt check readings were 4.1V.
Am I the only one noticing this, or has this been discussed elsewhere?
On a sort of related note, I’m wondering if anyone else has consider or had the desire to have the aux lights flash out Batt Check always as one of the options (i.e. L, H, Blink Batt Check, Off)
I find i can only get an accurate reading when checking the light on low. Presumably if you run the check on a medium or high mode you are getting a voltage sag reading.
I just tried 3 lights, TS10, SC21, and FW3A. I ran them all up to a higher output, turned them off, and then ran Batt Check. They all flashed at a relatively low level, independent of what the light was last turned on at. Not sure what you mean check on medium or high mode. I reran the SC21, ramping all the way to the top of the range. Batt check light level is rather low comparitively.
Battery voltage changes over time. The change is often significant when batteries are freshly charged. I have seen many cases where there were significant drops in voltage after batteries are charged even without use. Also, when a battery is heavily used, the voltage drops and later it creeps back up within seconds.
Anduril can only display up to one decimal place so it is very well possible the last digit can jump around if the voltage is near the borders. e.g. 3.45 can be 3.4 or 3.5, 4.15 can be displayed 4.1 or 4.2.
Finally, I don’t think the volt meter in our flashlight is accurate. I have seen a couple of cases where the measured voltage in Anduril flashlight is off by more than 0.1v. In some cases I have seen the difference was 0.3v.
On my 2 SC31T Pro’s the battery check when on high always gives me 3.3 when in Simple UI regardless of the battery’s charge state. At the lower levels it gives me the correct readings. This isn’t an issue when the light is in the Advance UI.
I agree with everything you wrote Limsup. And thanks again for helping me negotiate measuring current on my TS10.
None of it seems to apply to a cell charged 4 days previous, installed in a light and sitting in my pocket unused for hours, and then changing by 0.2V between one reading and another immediately following. No DVMs here, just Batt Check blinking out numbers. O.1V, sure, I can see the voltage drop from just above to just below. And, in my last post, I saw the batt check raise 0.1V after running the light on high, then immediately running batt check. But dropping from 4.3 to 4.1 on consecutive readings on installed in a light not used for hours previous? Seems really odd to me . . .
Problem seems to be that the voltage is measured even when the light shortly turns on at ceiling after the second button press of 3C. That causes a voltage sag and it needs a second to recover (the actual voltage and to update the measurement). This could be solved by better checking when the voltage is measured.
@roostre
Number of decimal places is not that important. Voltage reference, calibration and consistency has a much larger influence compared to the resolution of the ADC (10 bits for the 1634 for example).
One or two of my Anduril lights give me two different readings. First a lower reading then a higher and more accurate reading. I don’t know why. I’ve learned to watch two cycles of the bat check to make sure I get the right reading.
This is what I was trying to describe, it seems by most of the responses I did a poor job of it. My lights tend to be higher first, then settle out if/when I get two different readings. This thread was prompted by a difference of 0.2V, which I had never seen before. What sort of difference between readings do you see? What light(s)?
If that was the case then it wouldn’t matter which UI the light was in. There’s a definite differenced in the voltage flashes on my 2 SC31’s depending upon the UI.
Simple UI won’t repeat the output, thus you will never see the correct value. In advanced UI you can wait a few cycles and wait until the voltage has stabilized.
It’s true that the Simple UI only flashes once but the first flash in Advance UI is accurate so it’s not as if the voltage has stabilized. If voltage stabilization is the reason why the Simple UI is wrong then logically the first flash in Advance UI would also be wrong; at least to my way of thinking.
Hummm I just checked with one of my lights set to high and on Advance the first blinks were 3.9 and then 4.1 thereafter. Still not as bad as on Simple where it would flash out 3.3, so I guess I should say more accurate rather than just accurate
How does Anduril actually measure the voltage? I know that it is a popular choice to measure the VCC voltage to the MCU and therefor avoid a voltage divider and the use of an IO on the MCU. Does Anduril do this? If so, then how is the driver designed? Some drivers have diodes as reverse polarity protection, some even have a small resistor before that diode. All of these things have voltage drops over them, and it can very depending on what the MCU is doing, in particular if small LEDs are driven directly from IO on the MCU, because then the current draw from MCU is higher which results in different voltage drops over those input components which in turn leads to different voltage to the MCU. And different input components will give different results under different conditions. How many different types of reverse polarity diodes have been used over the years?
When I was measuring MCU voltage I had these issues and got tired of the need to calibrate each driver so I switched to a voltage divider and ADC IO on the MCU, and have never looked back. It’s much more accurate, I’ve never needed to calibrate since then and have ditched my entire calibration routine.
Disclaimer: I could be wrong about Anduril reading MCU voltage and the above is just my own experiences which could be wrong due to my own misunderstandings or my own poorly designed drivers.
