Devt #785
Merged
Devt #785
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
* Fixed various small bugs * Fixed potential cast bug * Fixed double reporting of errors Co-authored-by: Stefan de Bruijn <stefan@nubilosoft.com>
* Devt (#571) * Handles Tranimic drivers errors better - If an unsupported driver is specified, it will give a message and not crash. * Cleaned up unused files Got rid of old unipolar files Got rid of servo axis feature - it is a motor class now Got rid of solenoid pen feature - never really used and it should be a motor class if it is. * Fix ENABLE_AUTHENTICATION (#569) * Fixed authentication code. * Removed another const cast Co-authored-by: Stefan de Bruijn <stefan@nubilosoft.com> * Fix step leakage with inverted steps (#570) * Fix step leakage with inverted steps * Update build date for merge Co-authored-by: Bart Dring <bdring@buildlog.net> Co-authored-by: Stefan de Bruijn <atlaste@users.noreply.github.com> Co-authored-by: Stefan de Bruijn <stefan@nubilosoft.com> Co-authored-by: Mitch Bradley <wmb@firmworks.com> Co-authored-by: Bart Dring <bdring@buildlog.net> * Update platformio.ini Per PR 583 * Created an enum for mode * Removing some unused machine defs * Added test machine definition * Clean up for PR * Remove test machine def. Co-authored-by: Stefan de Bruijn <atlaste@users.noreply.github.com> Co-authored-by: Stefan de Bruijn <stefan@nubilosoft.com> Co-authored-by: Mitch Bradley <wmb@firmworks.com> Co-authored-by: Bart Dring <bdring@buildlog.net>
Pause resume fix
* Initial Tests Complete * Update Grbl.h * Update variables Co-authored-by: Mitch Bradley <wmb@firmworks.com>
* fixed dual switches when inverted * Removed debug message
Removed unused #defines.
* Store coordinate offsets in NVS * Handle both old Eeprom formats
- Stop creating additional tasks when limit_init() gets called again from homing and resets - Explicitly delete an object that was causing a memory loss.
Tweak memory fix and add $H check for $Homing/Cycles
Fix G28.1 and G30.1
* Changed to add homing cycle defaults There needs to be a way to set the homing cycle defaults in a machine definition. There will likely be a better way to do this in the future. * Update 10vSpindle.cpp Had wrong error message
* Cleanup probing code * Update Grbl.h * Update after review
…626) This is a patch necessary for F360 personal users, because they decided to add a very lengthy comment... Co-authored-by: Stefan de Bruijn <stefan@nubilosoft.com>
* $sd/show and handle settings in SD files * Added $LocalFs/Show and fixed $LocalFs/Run output * Infer / at beginning of SD path name The LocalFS path processing code already inserts a / at the beginning of the path is one isn't present. This patch does the same for SD files. * Show $ command responses in WebUI console * Added $Settings/ListChanged AKA $SC This is useful for saving settings in a compact form that leaves defaults unchanged. * $sd/show works in idle or alarm state * Apply idle/alarm checks to SPIFFS files too
sd_close was a temporary function to check for memory usage
-- Changes that affect behavior Fixed the bugs with report_build_info() Build info is no longer stored in the fixed "EEPROM" section; instead it is a proper Setting named $Firmware/Build . You can change it in the usual way with $Firmware/Build=<whatever> $I without the = still works. -- Changes that affect configurability for developers Converted a couple more #defines into enums - SETTINGS_RESTORE_* and BITFLAG_RT_STATUS_* . A side effect of this is that it is no longer possible to configure the behavior of $RST=* by defining SETTINGS_RESTORE_ALL to include only a subset. I think it is a bad idea from a customer support perspective to have the meaning of this command be different for different builds. Changed some of the #define ENABLE_ names to eliminate "EEPROM" -- Changes that are purely cosmetic Clarified descriptions in Config.h, to eliminate spurious/incorrect mentions of "EEPROM" Eliminated all mentions of the name "EEPROM" except the ones that truly mean the EEPROM section, as opposed to generalized non-volatile storage. The contents of SettingsStorage.h and SettingsStorage.cpp, which were really related to coordinate storage in Eeprom, not proper settings, were moved to Eeprom.h and Eeprom.cpp. The SettingsStorage files are gone. Got rid of get_step_pin_mask() and get_direction_pin_mask() because they were just aliases for bit(). That eliminated some junk from the SettingsStorage/Eeprom files. Those files now tightly contain only the residual stuff related to the storage of coordinate data in EEPROM.
* Many more #defines bite the dust * Fixed botch in rt accessory logic * Update Probe.cpp * Update System.cpp * Typo
While eliminating a redundant definition of is_realtime_command(), I inadvertently introduced a recursion due to the similarity of the names "is_realtime_command()" and "is_realtime_cmd()". The solution is to eliminate the latter entirely.
* Fix I2S stepper hung just after the completion of motor moving * Fix recompile issue Fixed a problem with the recompile not being recompiled even if the files under the Custom folder are changed. * More comment for macOS in debug.ini * Fix the timing of calling I2S out's exclusion function and reset sequence The reset sequence did not seem to be correct, so I changed it. According to the ESP-IDF PR, the correct sequence is as follows: 1)TX module 2)DMA 3)FIFO espressif/esp-idf@c7f3352#diff-27688c6b3c29373d2a2b142b8471981c * Changed the message level for I2S swtiching from warning to debug * Add some comments
When loading Grbl_Esp32 into a fresh ESP32, the WCOs would often have strange, very large, values. The problem was the code that tries to propagate data from the old "Eeprom" storage format into the new NVS scheme. The old format had a broken checksum computation that made the checksum so weak that if succeeds about half the time on random data. The solution is to get rid of all that old code. The downside is that migration from a build that uses the old format will lose the WCO values. The user will have to reestablish them. Subsequent updates between different versions that both use the new NVS format will propagate WCO values correctly.
* Fixes to homing * Update Grbl.h * Clean up after code review.
* WIP * WIP * Updates * Update Grbl.h * Removing some test machine definitions * TMC5160 Drivers were not in tests
If a command is critical and fails to receive a response, it should trigger an Alarm. However, because the critical check was only evaluated if the spindle was not already unresponsive, it meant that a critical command failure would be silently ignored if a non-critical command failed before it (putting the VFDSpindle in unresponsive state). Therefore, I've moved the critical check to occur regardless of whether the spindle was already unresponsive. Second, I believe that setting `sys_rt_exec_alarm` is not sufficient to stop the machine and put it into alarm state. Other alarm conditions (such as hard limits) also run an `mc_reset()` to stop motion first. It appears that without this, motion will not be stopped, and in fact, the alarm appears to get cleared if it occurs during motion!
* Settings filtering via regular expressions Implements only the most basic - and the most useful - special characters - ^$.* If the search string does not contain a special character, it is interpreted as before. Otherwise the match is either more strict if anchored by ^ or $, or less strict if it contains a . wildcard or a * repetition. * Commentary * Eliminated . metacharacter
* Fix SD/List repetition error The one line change that actually fixes it is Serial.cpp line 162, where the SD state is compared to "not busy" instead of "is idle", thus also handling the "not present" case. In the process, I converted the "const int SDCARD_ ..." to an SDState enum class, thus proving type safety and eliminating yet another untyped uint8_t . * Updates after testing Co-authored-by: bdring <barton.dring@gmail.com>
* PWM fix and simplification This is an alternative solution to the PWM/ISR/float problem. 1. The set_level() argument is the exact value that is passed to the LEDC Write function. It can be interpreted as the numerator of a rational fraction whose denominator is the max PWM value, i.e. the precision, == 1 << _resolution_bits 2. There is a new denominator() method that returns the precision. 3. The sys_pwm_control(mask, duty, synchronize) function is replaced by two functions sys_analog_all_off() and sys_set_analog(io_num, duty). This closely matches the actual usage. The old routine was called from two places, one where the mask was alway 0xFF, the duty was always 0, and synchronize was always false. That is the one that was troublesome from ISR context. The other call always affected a single pin, so the mask formulation with its loop was useless extra baggage. By splitting into two functions, each one becomes much simpler, thus faster and easier to understand. The "synchronize" argument and associated logic moved out to the caller (GCode.cpp), which more closely reflects the behavioral logic. 4. For symmetry, sys_io_control() was similarly split. 5. sys_calc_pwm_precision() was moved into UserOutput.cpp since is it driver specific, not a general system routine. * Update Grbl.h * Delete template.h Co-authored-by: bdring <barton.dring@gmail.com>
* TMC2209 Stallguard - Added StallGuard homing support to TMC2209 (UART) - Killed off TMC2208 for now. Too many conflicts with TMC2209. Will return with Diamond motor class hierarchy - Increase StallGuard setting range for TMC2209. Constrianed in each class to actual limits - Added a machine def to test TMC2209 * Update build date
* Update System.cpp * WebCommand with configurable modes * Added a few more ESP commands to work in anu state * Update Grbl.h Co-authored-by: Mitch Bradley <wmb@firmworks.com>
- $Message/Level - ISR safe ledcWrite
* Updates for CoreXY * Delete fystec_ant.h
* Changed delay type - mc_dwell was causing a recursive loop the overflowed the stack - See https://discord.com/channels/780079161460916227/786061602754396160/809288050387189782 * Changed spindle delays from floats to ints in spindle classes - Used local copies, because I did not want to change/break the existing setting. * Cleaning up parking - Added a coolant delay setting - Made an enum class for the dwell types - Got rid of the safety door sepcific delays * Update Grbl.h
* - moved invert option in front of per motor enables. * Added code to prevent motors_set_disable() from setting values that already exist. * Added the enable delay from PR 720 * Adding a defined default for step enable delay * Fixing feed rates with kinematics and arcs. - Kinematics changes the feed rate and the loop that creates arc was re-using the altered rate. This caused a runaway situation.
So machine definitions can change the SPI pins before we talk to any Trinamic drivers.
* Fix compilations error due to new enum in esp32 core release 1.0.5 * Update Grbl.h Co-authored-by: Luc <luc.@tech-hunters.com>
MitchBradley
approved these changes
Feb 26, 2021
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.