aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* fixed SP10 ramp flickering properly, using double-buffered registersSelene ToyKeeper2021-11-131-4/+5
| | | | | | | instead of my janky forced phase-reset workaround (gchart found the solution but couldn't reproduce the issue, so I tried his method and confirmed it seems to be fixed)
* enabled manual memory and timer by default;Selene ToyKeeper2021-11-131-0/+6
| | | | | | | reset to ~6 lm (level 50/150) after being off for 10 minutes This sets the factory-reset default settings and affects Simple UI, so it will likely need confirmation from Sofirn.
* increased SP10 PWM speed as much as possible without making ramp bumpy, and ↵Selene ToyKeeper2021-11-134-6/+20
| | | | | | | | | | | | | | | | | made party strobe pulses much faster Reduced max PWM TOP to 3072, because 2048 wasn't enough and 4096 was more than necessary. Also, Ch1 lumens / 256 / ch2 lumens = 6, so 256 * 6 * 2 is the lowest value which allows ch1 to start at half of ch2's power. I tried 1536 initially, but it made the ramp visibly malformed at the channel boundary. However, 3072 seems about right. Implemented a non-linear PWM_TOP ramp-down in level_calc, to allow it to converge faster and reduce the number of levels with visible pulses. Added an option to keep the regulator chips on between strobe pulses, by keeping the LEDs at moon instead of turning completely off. This allows the SP10 party strobe to use much shorter, more consistent pulses.
* reduced SP10 firefly/moon power usage slightly, by keeping CPU underclocked ↵Selene ToyKeeper2021-11-061-1/+1
| | | | more until ch2 activates
* reduced SP10 downward ramp flicker even moreSelene ToyKeeper2021-11-061-4/+4
| | | | | | (reduced PWM_TOP minimum timing window to 32 cpu cycles, to allow TOP value of 64 to work better)
* greatly smoothed out SP10 ramp, and configured defaultsSelene ToyKeeper2021-11-062-16/+25
| | | | | | | also fixed random ramp stuttering, by adding phase-reset register to hwdef (though it still has a brief stutter sometimes while ramping down across the channel boundary, at least it always seems to be smooth while going up now)
* merged recent anduril2 branch updatesSelene ToyKeeper2021-11-0618-81/+218
|\
| * fixed KR4-tintramp build (switch wasn't working) and added it to MODELSSelene ToyKeeper2021-11-033-8/+8
| |
| * reset tint while changing tint ramp styleSelene ToyKeeper2021-11-031-1/+5
| |
| * added runtime config to choose tint-ramping or tint-toggleSelene ToyKeeper2021-11-038-32/+54
| | | | | | | | | | (Off->9H option 1, 0=smooth, 1=toggle)
| * added KR4-tintramp buildSelene ToyKeeper2021-11-033-1/+63
| | | | | | | | | | | | (is basically identical to D4Sv2-tintramp, but with the switch on a different pin, and no button LED)
| * documented number entry 1H action (add 10) in UI tableSelene ToyKeeper2021-10-311-0/+1
| |
| * made d4sv2-tintramp-fet work the same as no-fet version, but with added FET ↵Selene ToyKeeper2021-10-312-9/+17
| | | | | | | | | | | | | | from ramp step 141 to 150 (0 to 100% power from 1 to 130, 101% to 200% from 131 to 150, and +DD FET from 141 to 150) also calibrated candle mode a bit better
| * fixed flicker on inactive tint ramp channel during rampingSelene ToyKeeper2021-10-311-2/+8
| |
| * added "200%" turbo on tint-ramping lights: D4S and LT1Selene ToyKeeper2021-10-315-37/+71
| | | | | | | | | | | | | | | | | | | | Normal ramp from 0% to 100% power on levels 1 to 130, then 101% to 200% power at levels 131 to 150 using both channels at maximum for turbo. When either channel would go over 100%, the extra spills over to the other channel.
* | merged gchart's recent changesSelene ToyKeeper2021-11-0613-74/+298
|\ \
| * | Merge TKs changes through her commit 628Gabriel Hart2021-10-297-67/+137
| |\|
| | * Clarified Ramp->2C behavior a bit more.Selene ToyKeeper2021-10-281-4/+4
| | | | | | | | | | | | | | | (old wording was confusing people)
| | * fixed too-slow thermal regulation bug introduced in d4sv2-tintramp branchSelene ToyKeeper2021-10-251-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | (2021-09-10 r619.1.11) The gradual_target var was getting clobbered and causing thermal regulation to stop until the next thermal warning event, every time it reached a new ramp step. So... save/restore it to prevent it from getting clobbered.
| | * don't blink when toggling tintSelene ToyKeeper2021-10-141-1/+1
| | |
| | * don't hardcode 2.9V for aux LED shutoffSelene ToyKeeper2021-10-091-1/+1
| | | | | | | | | | | | | | | (use the same value as LVP; easier to configure if it's in only one place)
| | * added attiny type to the MODELS file, rewrote how MODELS file gets generatedSelene ToyKeeper2021-09-223-60/+128
| | | | | | | | | | | | | | | (because it's kind of important to know which MCU each light uses, and because this will be helpful later when the build system is rewritten)
| * | Add thefreeman's DAC linear driver (beta); also update Sofirn bilds to use ↵Gabriel Hart2021-10-296-7/+161
| | | | | | | | | | | | the new 2C_STYLE defines
| * | SP10 Pro: lower AA LVP to 0.7vGabriel Hart2021-09-202-2/+2
| | |
* | | merged gchart's SP10 Pro branchSelene ToyKeeper2021-09-1914-84/+172
|\| | | |/ |/| | | | | (minus the change to version.h)
| * SP10 Pro: use dynamic PWM, add hidden firefly, adjust candle amplitude, and ↵Gabriel Hart2021-09-193-15/+25
| | | | | | | | tweak party strobe & LVP level
| * Merge TK changes thru rev 623 (2021-09-17)Gabriel Hart2021-09-1932-190/+1112
| |\ | |/ |/|
* | clarified where brightness level comes from in beacon/sos/momentary modesSelene ToyKeeper2021-09-171-6/+6
| |
* | merged K9.3-tintramp / D4Sv2-tintramp branchSelene ToyKeeper2021-09-1518-87/+764
|\ \ | | | | | | | | | | | | | | | (which also changed the way tint ramping is implemented, to make things generally cleaner and more flexible)
| * | added Noctigon K9.3-tintramp build targets for FET, reduced FET, and noFETSelene ToyKeeper2021-09-153-0/+157
| | | | | | | | | | | | | | | | | | | | | (apparently Hank liked the D4Sv2-tintramp demo/experiment so much that he decided to use it in production for the K9.3... so it need its own builds since it's a pretty different light even if the driver is almost identical)
| * | fixed wrong value on D4Sv2-tintramp ramp calculator (was off by one)Selene ToyKeeper2021-09-151-3/+3
| | |
| * | fixed builds which failed due to being too big (rot66, mf01-mini)Selene ToyKeeper2021-09-102-0/+4
| | | | | | | | | | | | | | | | | | by turning off one more of the recent extra features (can turn it back on later if the build size goes down)
| * | fixed K9.3 builds (oops, new update_tint() needed to be compiled out for K9.3)Selene ToyKeeper2021-09-102-2/+2
| | |
| * | fixed blf-lantern-t1616 build... I thinkSelene ToyKeeper2021-09-102-16/+109
| | | | | | | | | | | | | | | (can't test because I have no hardware, but it at least compiles)
| * | fixed ancient bug which caused a weird up-spike in thermal regulation during ↵Selene ToyKeeper2021-09-101-4/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | initial turbo drop, on 3-channel drivers like FW3A and ROT66 (as shown here: http://toykeeper.net/torches/fw3a/therm-2019-05-22.1.png ) What happened was... the FET would start to drop, but gradual adjustments noticed that the Nx7135 channel needed to go from 0 to 255, so it would then slowly ramp that up, and then afterward, the FET drop could finally continue... because the code didn't jump straight from 0 to 255 like it was supposed to. Simple, easy fix: Make channel 2 go up immediately just like channel 1 does. This makes the thermal response several seconds faster than it was before, so it doesn't get as hot, and is less likely to overshoot and bounce later.
| * | added D4Sv2-tintramp-FET build, which uses a FET on one set of LEDs for ↵Selene ToyKeeper2021-09-104-15/+86
| | | | | | | | | | | | | | | | | | | | | | | | | | | turbo modes (it's weird, but Hank wants it) also reworked gradual tint adjustment a bit, so some complex parts go in set_level() instead (probably needs more testing)
| * | added -mrelax to build.sh; saves ~330 bytes on t1634 buildsSelene ToyKeeper2021-09-101-1/+1
| | | | | | | | | | | | | | | | | | (suggested by solrize) (doesn't seem to reduce size of t85 builds though)
| * | changed D4Sv2-tintramp from 255 to 511 for lowest PWM TOP value,Selene ToyKeeper2021-09-082-5/+5
| | | | | | | | | | | | | | | | | | to make its thermal regulation smoother and generally increase tint ramp resolution in middle modes (also, tried thermal regulation, and it works)
| * | got BLF LT1 working again, using new tint-ramping codeSelene ToyKeeper2021-09-083-31/+50
| | |
| * | fixed D4Sv2-tintramp PWM_TOP (should be set to the value it needs during ↵Selene ToyKeeper2021-09-081-2/+2
| | | | | | | | | | | | thermal regulation)
| * | copied Q8 hwdef to LT1 so I can improve itSelene ToyKeeper2021-09-081-0/+21
| | |
| * | fixed outdated comment in KR4 hwdefSelene ToyKeeper2021-09-081-1/+1
| | |
| * | got D4Sv2 tint ramping to work, with dynamic PWM (PFM)Selene ToyKeeper2021-09-084-80/+121
| | | | | | | | | | | | | | | | | | | | | | | | | | | and (maybe) thermal regulation (untested) (also broke BLF LT1 in the process; need to fix that now) Rewrote how tint ramping works, so it provides a virtual "PWM1_LVL" for other code to use, and it translates that internally into actual hardware controls. This should, in theory, allow smooth thermal regulation (gradual_tick) to work on tint-ramp lights.
| * | copied k9.3 config for new d4sv2-tintrampSelene ToyKeeper2021-09-071-0/+85
| | |
| * | adjusted hwdef settings for tint-ramping D4Sv2Selene ToyKeeper2021-09-071-17/+25
| | | | | | | | | | | | | | | (initial rev, will almost certainly need changes later)
| * | copied hwdef from K9.3 for new tint-ramping D4Sv2Selene ToyKeeper2021-09-071-0/+164
| | |
* | | documented "hold" action in number entry stateSelene ToyKeeper2021-09-141-0/+5
| | |
* | | added a hold shortcut in number entry menu: click for +1, hold for +10Selene ToyKeeper2021-09-142-2/+14
|/ /
* | merged in a bunch of new features...Selene ToyKeeper2021-08-3115-95/+320
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | + runtime option for smooth ramp speed + runtime option for whether hold-from-off should ramp up or stay at the floor + runtime option to select turbo style / 2C style in Advanced UI + same thing, but for Simple UI * sped up the auto-reverse timing window; it was too slow
| * | fixed builds which failed to compile after recent changesSelene ToyKeeper2021-08-314-2/+8
| | |
generated by cgit v1.2.3 (git 2.25.1) at 2026-03-28 19:11:06 +0000