My LT1 went on a round-trip from China to The Netherlands, back to China and now on it’s way back to The Netherlands. Been traveling since the 11th of November :,)
Any idea how to handle calibration for lights which use a voltage divider instead of measuring VCC internally?
FSM supports two methods for measuring voltage:
Direct from the VCC pin. Uses the “fudge factor” thing for calibration, since it’s usually pretty accurate aside from needing to add a constant value to make up for voltage drop across the RPP diode.
Voltage divider. Works on a wider variety of hardware designs, but uses up an extra pin. Calibrated by using 2 data points at opposite ends of the charge spectrum, then estimating the rest with a line between those 2 points. Those were 8-bit ADC values until recently, when I switched it to 10-bit.
Both are currently configured per build target, and the UI code doesn’t really know or care which is used internally. I’m not really sure what kind of UI would work for calibrating the voltage divider, since it needs two numbers from 0 to 1023.
Would there be any difference for the 3rd batch of the LT1?
I) I understand the first batch had these “issues”:
a) the 7135 solder points = some of them had solders on all, and then cleaned up (supposed to come with 7135x5, but it seems maybe the factory was mistaken and soldered all 7135x7, but then cleaned up to make it back to 7135x5, which resulted in some having not so well cleaned-up solder points)
b) the top handle is not “aligned” to the button (it’s diagonal)
c) uses an older firmware
II) While the “second batch” fixes the above issues
— 7135x5 is soldered properly (and the extra “6” and “7” are “clean”)
— top handle is now centered/aligned to the button
— a bit later version firmware is now used (but likely not yet the latest)
III) How about the 3rd batch? any minor improvements or the like? (maybe uses the very latest firmware now?)
That’s pretty cool. Any idea how to handle calibration for lights which use a voltage divider instead of measuring VCC internally? FSM supports two methods for measuring voltage: * Direct from the VCC pin. Uses the “fudge factor” thing for calibration, since it’s usually pretty accurate aside from needing to add a constant value to make up for voltage drop across the RPP diode. * Voltage divider. Works on a wider variety of hardware designs, but uses up an extra pin. Calibrated by using 2 data points at opposite ends of the charge spectrum, then estimating the rest with a line between those 2 points. Those were 8-bit ADC values until recently, when I switched it to 10-bit. Both are currently configured per build target, and the UI code doesn’t really know or care which is used internally. I’m not really sure what kind of UI would work for calibrating the voltage divider, since it needs two numbers from 0 to 1023.
[/quote]
That post do not fill me with confidence, the programmer shouldn’t be asking forum members to correct & improve the programmers software.
Don you chose nicer words than I was going to [thumbsup]
I think it’s great that two programmers can have a meaningful conversation about how to improve something they worked on, in the same thread consumers read. Shows humility and self confidence.
That’s how free software works… it’s a participatory community — distributed non-hierarchic collaboration. Instead of dividing people into producers and consumers, people try to work together as peers.
Not everyone is a programmer, but that doesn’t mean they have nothing to contribute. Lots of people around here design lights and circuits, manage projects, track interest lists, write documentation, answer questions, keep each other in good spirits, run tests, measure hardware performance, prototype new concepts, etc. Even if it’s just ideas and suggestions, everyone is welcome to contribute what they can… and I try to factor in the needs of many different people.
But for some context, Tom E is one of the most experienced modders here, and one of the most prolific flashlight firmware authors, and his code is used on quite a few popular products. He created Narsil / NarsilM, which is what inspired Anduril, so I take his suggestions very seriously.