| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
|
|
| |
delays + smooth steps
Anduril has gradually gotten faster over the years, apparently, so it
needed longer delays to get accurate-ish timing for beacon and other modes.
Adding DSM also changes the timing perceptibly, so I made it possible to
calibrate the delay fudge factor on a per-build basis.
|
| |
|
|
|
| |
(its DSM interrupt wasn't working at all, and it needed a few other tweaks)
|
| |
|
|
|
| |
(but it's not tested on hardware, since I have none)
|
| |
|
|
|
|
| |
(it's mostly the same as FW3A, but has an aux LED
so there's less room for other stuff)
|
| |
|
|
|
| |
(it's mostly the same as FW3A)
|
| |
|
|
|
| |
(and made DSM interrupt definitions a bit cleaner)
|
| | |
|
| |
|
|
|
| |
(this gives better, smoother low modes and reduced flicker)
|
| |
|
|
|
| |
(is almost identical to a wurkkos-ts10)
|
| |
|
|
|
|
| |
(since the Q8 driver was the direct predecessor to those Wurkkos drivers,
they're almost identical)
|
| |
|
|
|
| |
(it's almost identical to noctigon-dm11-boost)
|
| | |
|
| | |
|
| |
|
|
|
|
| |
(needs further updates though, to improve ramping, since this version is
basically a straight conversion of the old t85 code with 8-bit ramps)
|
| | |
|
| |
|
|
|
| |
(it seems t85 doesn't like having timer overflow interrupts enabled in standby?)
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
which required ... a few pretty significant changes:
- no dynamic underclocking (it isn't compatible with DSM yet)
- no tint ramping brightness correction (removed to save space)
- removed ramp blinks (to save space, and because they're annoying)
- removed momentary mode (to save space)
- removed SOS mode (to save space)
- removed (to save space) some other relatively recent features which
weren't present in the original production firmware
... but some other things improved:
+ added smooth steps
+ extended Simple UI
+ added stepped tint ramping
+ added 13H factory reset, to save wear on threads
+ lower lows
+ smoother ramp
+ much higher tint ramp resolution in low modes
I'm not entirely happy with this yet, so it probably needs additional work
later in order to adjust the weird ramp shape (these 7135 chips have a weird
response curve), add dynamic underclocking, cut down the ROM size if possible,
re-add tint ramping brightness correction, etc. Multi-channel stuff in
particular added a lot to the size.
This is a pretty big change from the previous working build, so some users
may want to stick with the last pre-multi-channel version. Non-trivial
sacrifices were made to bring in more recent features.
|
| | |
|
| |
|
|
|
| |
when smooth_steps was enabled
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
which gives much better resolution, especially for the 8-bit channel.
Also...
- set_channel_mode() aborts when going from/to the same channel,
to avoid unnecessary flicker
- hsv2rgb() uses 16-bit R/G/B and V now
- changed default channel to All
- reduced default channel modes to just A, B, C, and All
- smooth ramp floor defaults to 1/150
- raised level when aux LEDs turn on high during use
(for better compatibility with red main LEDs)
|
| | |
|
| |
|
|
|
| |
(bug introduced in r772, when adding AA/NiMH RGB voltage support)
|
| |
|
|
|
| |
removed SOS mode to make room for other stuff
|
| | |
|
| | |
|
| |
|
|
|
|
|
| |
(my FW3A dev host is dead though, so the DD FET channel
isn't 100% confirmed to work ... will have to solder together
a new dev host at some point)
|
| | |
|
| |
|
|
|
| |
(150 was a misunderstanding, and is too high)
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- Off -> 1H
- Off -> 2H and release
- Off -> 2C
- Off -> 3C or more
- Ramp -> 2C (in smooth ramp mode)
... also reduced ROM size a bit.
Now it does smooth animations for 2C turbo, regardless of the current ramp
mode. Before, in smooth ramp mode, 2C turbo did a hard step. Hard steps were
also eliminated for 1H from Off, momentary turbo from Off, regular turbo from
Off, and anything longer than 2C from Off.
|
| | |
|
| | |
|
| |
|
|
|
| |
(works well in a FW3A host)
|
| |
|
|
|
| |
slightly simplified gradual_tick()
|
| |
|
|
|
|
| |
+ added smooth steps
- removed SOS mode
|
| |
|
|
|
| |
emisar-d4k-3ch: removed redundant #define
|
| |
|
|
|
| |
with a new ramp and some new hwdef algorithms
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
- emisar-d4
- emisar-d4v2
- emisar-d4v2-nofet
- emisar-d4sv2
- emisar-2ch
- emisar-2ch-fet
- noctigon-dm11-boost
- noctigon-k1
- noctigon-kr4
- noctigon-kr4-nofet
- sofirn-lt1s-pro
... and removed old build targets for d4sv2-tintramp, because it was
replaced by emisar-2ch a while ago.
|
| |
|
|
|
|
|
|
| |
when bottom step was too low (stepped ramp floor 1/150 or 2/150),
instead of just stepping down
(this also seemed to affect 1C from Off, when at stepped ramp floor)
(but it didn't affect the smooth ramp)
|
| |
|
|
|
| |
(it would jump up the last few levels before, instead of being smooth)
|
| |
|
|
|
|
|
|
| |
(for consistency, and in case they ever make a full RGB version)
(also, raised the post-off voltage brightness threshold, because
these aux LEDs are pretty bright)
(and updated the TS10 hwdef to the current API)
|
| |
|
|
|
| |
by using different curves for ramp-up and ramp-down
|
| |
|
|
|
| |
(also made ramp extras config menu count its own steps)
|
| |
|
|
|
|
| |
and made it longer on emisar-d4k-3ch
because some channels don't flash fast enough
|
| |
|
|
| |
aux setting was high
|
| |
|
|
|
|
|
|
|
|
|
| |
made USE_AUX_RGB_LEDS_WHILE_ON work more like the old indicator LEDs,
where it gets set automatically with set_level()
Using set_level_zero() reduces space used by channel modes, and simplifies
code for each mode's set_level_*() functions. I measured about 220 bytes
less in the emisar-d4k-3ch build this way, while also reducing the chance
of bugs.
|
| |
|
|
|
|
| |
fixed some minor issues with LEDs not getting turned off in corner cases
(strobe on channel 1 -> strobe on channel 2, 1 might not get turned off)
|
| | |
|
| |
|
|
|
| |
and fixed issue in tactical mode where strobes wouldn't stop on button release
|
| | |
|