* M600 used 1°C threshold, which may increase the waiting time a bit
* Wizard used 3°C
Sync both to use TEMP_HYSTERESIS for consistancy
No change in memory footprint
Use 5°C threshold to be consistant with other parts
of the firmware.
Relying on 95% of the target temperature creates
a dependency on the temperature:
PLA: Target = 215°C, threshold = 10.75°C
PETG: Target = 230°C, threshold = 11.5°C
ABS: Target = 255°C, threshold = 12.75°C
ASA: Target =260°C, threshold = 13.0°C
PC: Target = 275°C, threshold = 13.75°C
My proposal is we instead use a constant
TEMP_HYSTERESIS = 5, which is consistent with
M109, and behavior when restoring print from RAM
and some of the MMU code (like unload function)
Change in memory:
Flash: +2 bytes
SRAM: 0 bytes
This commit adds the ability for the firmware to dim and wake the
backlight when LCD updates are disabled. Such as in the MMU error screen
or when rendering full screen messages which typically
disable the LCD updates to prevent the status screen from rendering.
Fixes#2777
Change in memory:
Flash: -26 bytes
SRAM: +1 byte
It may be useful to view the Sensors menu
while the toolchange loading test is taking
place. For example to see if the reading is flickering
The firmware needs to call lcd_update(0) to update the screen rendering.
Change in memory:
Flash: -2 bytes
SRAM: 0 bytes
Model UV as power-invariant, so that scaling P doesn't change the
intercept location (that is, the zero point remains at the same
temperature even for more powerful heaters).
NOTE: It's not clear to me whether this is generally true (couldn't
find a datasheet from the same heater in diffent power variants
showing the efficiency loss)
- Expose TEMP_MODEL_fS and TEMP_MODEL_LAG as D and L respectively,
initializing the default values based on the previous hard-coded
values.
- Always round L to the effective sample lag to future-proof model
upgrades or changes (the stored value _is_ the effective value).
- Introduce UV as a new linear term for PTC heaters, defaulting
to an identity for model backward-compatibility.
On printers without the MMU no error screen should occur and no attempts of communication with the MMU should be performed -> EEPROM_MMU_ENABLED should default to 0.
PFW-1418
Use "echo:" for thermal model error reporting to avoid octoprint
automatically sending a M112 kill.
Keep using "error:" instead for other thermal errors (MAXTEMP/etc).
This should allow resuming a thermal mode pause with the default
octoprint settings.
This is an 80% solution to PFW-1488 which should be somewhat compatible with the existing languages/translations infrastructure.
I don't see a point in extending the infrastructure a great deal to support some compile-time replacement in order to patch just MSG_DESC_FW_UPDATE_NEEDED.
Related PR: https://github.com/prusa3d/Prusa-Firmware/pull/3993
Main changes:
* setAllTargetHotends() is removed
* setTargetHotendSafe() is removed
* Extruder parameter on setTargetHotend() is dropped
Change in memory:
Flash: -192 bytes
SRAM: 0 bytes
The general idea is to keep platform specific implementation away from the MMU state machines as much as we can.
That would enable unit testing the top level MMU state machine and integration into other project as well (if needed).
The function adds a bit of overhead compared to code_value()
I suspect this function is a relic of the past.
Change in memory:
Flash: -82 bytes
SRAM: 0 bytes
During calibration model warnings are possible/normal. This is
apparently leading some into thinking the hotend might be faulty
and abort the calibration with a hard reset.
Disable warnings/beeps during calibration.
Rely on the error threshold only.
It may seem counterintuitive to disable the E-motor, but it gets enabled in the planner whenever the E-motor is to move.
The idea behind disbling the E-motor when it won't be needed for some time is to save power and lower the heat of the motor.
Therefore it is especially important to disable the motor before the printer pauses a print due to some MMU error screen.
Used lcd_print_pad_P is some obvious places. We can probably use
this function in more places to further save flash memory
Change in memory:
Flash: -268 bytes
SRAM: 0 bytes
Because this error is not an MMU error but a printer one, the existing infrastructure has been bent to support such a scenario.
During testing it turned out, that some machines fail to draw the error screen due to previous internal states.
This PR resets the internal states so that the conditions for drawing the error screen are met.
During thermal errors all fans are set to full speed.
When the print is resumed or stopped *and* the thermal error is gone,
also restore the autofan state.
Fixes#3893
Running `M310 A` should set the TM calibration bit status if the
autotune procedure was successful.
Partially addresses #3891 and #3890, since running `M310 A F0` should
*still* count as valid calibration data and is not something that can be
replicated by using the LCD (which enforces self-test).
This fixes#3891 without having to set the calibration bit, meaning
that if the model is later re-enabled without a real calibration, a
prompt is shown as expected.
Alert messages are intended to persist above "info" messages, thus
preventing further calibration status updates to be shown (such as
thermal model calibration).
Just set the message as a regular status message.
Fixes#3892
It was only enabled when multiple extruders are enabled
The firmware doesn't support it
and even if the M218 gcode could be used, the code does not compile
Disabling the model during a warn/error condition will also stop
updating the warning/error flag, keeping the printer in an error state.
Clear all flags as well when changing model settings.
Follow Marlin's behavior and simply insert a delay for the requested
duration when using M300 S0.
When S is not specified, use the default tone instead.
Fixes#3856
This PR tries to address the 3 different startup scenarios for 1st layer calibration with the MMU:
- 1st lay cal started with correct filament already loaded in the nozzle - we should continue, but skip the first 58mm (first 2 g-codes in the hard coded sequence) of purge line extrusion
- 1st lay cal started with other filament already loaded in the nozzle - we should unload and then issue a toolchange with no extra unload
- 1st lay cal started without loaded filament - we should just do a toolchange with no extra unload
PFW-1457
This is a prototype implementation of having the ability to show an MMU error screen even for printer's errors (during an MMU operation).
Also, the retry count of unloads after failed load the extruder tube is now limited to ~3 attempts.
Technically, since this very error is not an MMU's one (MMU is just fine at this stage) but a printer's one I tried to hack the existing error-reporting infrastructure to handle such a case.
The original idea of this approach was suggested by @vintagePC
Instead of checking for explicit versions, resume the wizard if some
(new) wizard is missing.
This handles both the old SELFTEST check, the new thermal model
and any future check in the same fashion.
Update the eeprom FW version as soon as migration is complete, to avoid
resetting the CALIBRATION_V2 variable at each reset.
Do not implicitly reset the calibration steps for WizState::Run: do this
only for the menu action.
- Handle the conversion from the V1 calibration status to the V2 bitmap
for all past FW versions.
- Allow to run the wizard in "Restore" mode (for example during upgrade
or service prep) to complete any missing steps.
- Fix service prep to use the above feature by clearing the appropriate
calibration bits.
- Remove exceptions for FW 3.2.0.4 which can now be handled cleanly.
Simplify status tracking:
- S::Restore to continue to the next logical wizard item
- S::Finish for a successful failure
- S::Failed to exit while showing a failure
- Move calibration status/declarations to util.h
- Allows to query the status of each step independently without assuming
any order
- Allows to extend the calibration with new steps transparently
- Generalize force_selftest_if_fw_version() so we can check when upgrading
between arbitrary FW versions.
- Do not call update_current_firmware_version_to_eeprom() prematurely
when forcing self-test (if a reset happens before self-test completes,
the check would be incorrectly skipped on the next run).
Expect these to be always at the start of the command (which is now
stripped of the extra line number), reducing the number of corner cases
due to these not following the g-code spec.
Instead of reading and checking if a character is a whitespace
just print it out onto the LCD. This makes the code a bit less
complex. Whether there is a whitespace character, we always end
up reading from program space, so I doubt this has much performance
impact.
Change in memory:
Flash: -38 bytes
SRAM: 0 bytes
The Extra Loading Distance is configurable by the user.
We need to compensate the hardcoded sequence
such that it does not extrude too much or
too little. Currently the firmware
extrudes too little.
Fixes issue where First Layer Cal. triggers an unload before any
filament is loaded
The purpose of this if() statement is to handle the case where
a user or developer is sending T-codes to the printer directly
via Serial. Such as when one first sends T0 and then T4.
And unload must be triggered in-between, we can do this automatically
if FINDA detect filament.
Change in memory:
Flash: -10 bytes
SRAM: 0 bytes
If there is no running print, and the printer is Stopped, add a new
"Acknowledge error" menu entry to unlock the printer.
This simply calls lcd_print_stop(), which is identical in behavior to a
thermal error with a running print.
Set a menu block for fatal thermal errors instead of abusing
lcd_return_to_status() to kick the user out of the menus.
This now allows a thermal model error to be recoverable through menu
access.
We allow resuming from the LCD via start print and resume print,
it makes sense to clear the error on stop too.
For this reason distinguish whether the action is performed
automatically or manually (ie: interactively).
The error is only cleared when the command is run interactively.
The printer shouldn't continue endlessly asking for a version if an old MMU is detected.
Also if the protocol is supposed to be turned off, it
shouldn't be sending S0 queries.
In M600 and M702, the Z-lift is taken care of outside this function
so it makes no sense to have a Z-lift within unload_filament()
Only exception to this is in the wizard, there I moved the Z-lift to
happen before we start heating the nozzle.
Change in memory:
Flash: -12 bytes
SRAM: 0 bytes
Ensure that fan checks are reset and use the new speed at each step of
the calibration.
This also gives extra time to the fan to ramp-up from a cold start,
when a fancheck could previously start right *after* the speed change.
Should fix#3791
Since M310 cannot change the heaters, allowing M310 S0 (and changing
parameters) allows to recover a usb-controlled printer which has been
locked-out due to bad model settings.
This prevents the ability to run gcode from the SD card.
In a thermal error with faulty values, and without serial access, this
is the only way to process an M310 instruction and recover.
- Correctly sequence the enqueued commands in order to check for the
real calibration result.
- Perform autocalibration with self-check enabled.
- Only save/continue the wizard if autocalibration succeeds.
- Ensures repeated autotune attempts with self-check can't succeed due
to different starting conditions.
- Allows for a simpler workflow during selftest and wizard if autotune
fails.
Provide average calibrated values for all fan levels for MK3/MK3S
variants and enable the model by default.
This ensures the calibration step is not skipped unless explicitly
disabled.
Force-clear fan values during autotune to restore the previous
calibration behavior, as fan resistance levels can be too big and
prevent the autotune-with-selfcheck to fail.
3d-gussner
Guðni Már Gilbert
D.R.racer
Yuri D'Elia
vintagepc
Giles Bathgate
RoboMagus
Alex Voinea
Yuri D'Elia
Alex Voinea