Hi everyone I want to post some info here about the most common issues with these MINIs and some solutions how to fix them.
1. Incorrect Voltage measurement and LVP kicking in
I asked Lexel about the problems with voltage measurement (since he is the one who designed the board) and he was very kind in walking me through the problem and showing me a solution. Some of these MINIs measure the voltage bang on, some measure up to 0.4V less. If they measure the voltage to be 0.3V less than it really is, the low voltage protection will kick in 0.3V sooner and turn off the light. This is caused by the fact that the manofacturer has changed some components in Lexels original design for cheaper alternatives: mosfet, inductor and very likely a diode (Lexel wasnt sure from the picture about the diode). The easiest way to correct this is to flash the flashlight. Manual for doing so can be found here:
You will need to do following changes to the code:
_in the hwdef of the Anduril files you needx to adjust this code
// average drop across diode on this hardware
#ifndef VOLTAGE_FUDGE_FACTOR
#define VOLTAGE_FUDGE_FACTOR 5 // add 0.25V
#endif_
each +1 adds 0.05V
I asked ToyKeeper if she plans on adding a feature that would allow voltage calibration by clicking in the menu. She told me its not realistic at the moment. As is the Anduril code barly fits onto the chip most flashlights use. Thats why there is eather check version or muggle mode option but not both. Some flashlight like emisar have switcher for larger chips, but most havent.
2. Flashlight throttling down after few seconds from max ramp
These lights have a tendency to throttle down from max ramp after 20-30sec to about half the visible intensity. This doesnt seam to happen as fast in turbo mode when the lights get much hotter. I asked Lexel about it and he gave me this explanation (I copied only the most important parts):
_It’s not a problem it’s a sign that the user didn’t calibrate the temperature setting.
Any Bistro 3 channel driver has this behavior and no one ever complained about it.
It’s a known fact that FET+n+1 drivers heat more in regulated mode than while FET only active (smaller drivers are always more affected than big ones).
FET+1 drivers have only a fraction of heat on driver compared to a full bank of AMCs, also a 7 LED light has low LED voltage
You can simply eliminate this by flashing a firmware not using the AMC bank stepping back to FET+1 2 channels.
Generally a setting of like 45°C is way too low with new temperature code, I usually use 55 or 60°C (LiPo is 70°C upper limit)._
Easiest solution here is to calibrate the light to the correct room temperature and set the temperature limit higher. Let it cool down to room temperature (like about 30min) and calibrate the temperature to the actual flashlight/room temperature. Than set the temperature limit to a higher value like 60C (max 70C). As I understood this some components on the driver get very hot during voltage/current regulation in ramping and on max ramp they get so hot that they cause the driver to lower output power. This effect is a bit more noticable on this light because is has very small and a bit packed driver board (they had to fit those components in somehow).
EDIT 30.11.2019:
User man of light designed a great little copper insert that replaces the original plastic driver cover and creates a thermal bridge between the AMCs and flashlight body. This copper MOD will greatly increase the time before the light starts throttling down in ramping once you set ramp level. Here is a link to his project, its simple jet very clever MOD and he is trying to secure a small batch of these custom made for us, go check his thread for more info on this project, prices etc. Link:
s. I am very gratefull to Lexel and Toykeeper for entertaining my silly question (especially to Lexel who walked me through details).