JKBMS BLE reinstall-local.sh: Building wheel for dbus-fast won't finish #785
Labels
bug
Something isn't working
Comments
|
I tried the same steps on an fresh venusOS installation on my RaspberryPi 4. It seems to work without problems but the step "Building wheel for dbus-fast" took very long. Could any time-out be a problem with the RaspberryPi 2w? Because they are much less powerfull. |
|
Trying a fresh start with the RaspberryPi 2w worked today. The only difference i could find was no USB-devices plugged into the RP 2w. I won't do any further testing for now, because the RP 4 is currently used for the victron-system and the system is not installed at my home (but within my family). I'll also close the issue for now. |
|
I had the same behaviour on my MultiPlus GX, but did not find a solution so I copied over the dbus-fast from my Raspberry Pi 4. |
|
Now I found a solution. See Bluetooth-Devices/dbus-fast#237 This will be implemented in the next version of this driver. For older drivers you have to run export SKIP_CYTHON=falsebefore running the |
mr-manuel
added a commit
to mr-manuel/venus-os_dbus-serialbattery
that referenced
this issue
Sep 5, 2023
Merged
Closed
Louisvdw
pushed a commit
that referenced
this issue
Nov 25, 2023
* Update reinstall-local.sh: Fixed charge current parameter Update reinstall-local.sh: Corrected charge current parameter for initial config.ini * Exclude devices from driver startup This prevents blocking the serial port * implement callback function for update * fix comments to reflect new logic * update changelog * set soc=100% when charge mode changes to float, apply exponential smoothing on current readout * remove scan for devices the scan for devices and check if the BMS to test is present doesn't add value if the device is not within range (or the MAC is wrong), then the subsequent start_scraping call will either fail or fail to produce usable data * JKBMS_BLE driver fixes * added Bluetooth signal strenght, increased debug * Optimized reinstallation procedure - Changed: Optimized restart sequence for the bluetooth installation - Changed: Run serial part first and then bluetooth part. This allows the serial driver to get operative faster - Removed: $DRIVERNAME variable for clearer paths - Removed: Bluetooth system driver restart, since the devices get disconnected by the service before starting the dbus-serialbatterydriver * Improved Jkbms_Ble error handling * optimized disable procedure * small fixes * save custom name and make it restart persistant #100 * changed unique identifier from string to function function can be overridden by BMS battery class * fix typo * fix Sinowealth not loading #702 * fix unique identifier function * enable BMS over config, if disabled by default Now you can also add more then one BMS for BMS_TYPE * show battery port in log * ANT BMS fixes Fixed that other devices are recognized as ANT BMS * Sinowealth BMS fixes Fixed that other devices are recognized as Sinowealth BMS * improved publish_battery error handling switched from error count to seconds * Improve Battery Voltage Handling in Linear Absorption Mode * Refactor change time() to int(time()) for consistency in max_voltage_start_time and tDiff calculation * Refactor battery voltage calculations for efficiency and clarity * Remove penalty_buffer * Reset max_voltage_start_time wenn we going to bulk(dynamic) mode * updated changelog * fix reply processing * Reduce the big inrush current, if the CVL jumps from Bulk/Absorbtion to Float fix #659 * Check returned data lenght for Seplos BMS Be stricter about the return data we accept, might fix the problem of grid meters accidently being recognized as a Seplos * Validate current, voltage, capacity and SoC for all BMS This prevents that a device, which is no BMS, is detected as BMS * removed double check * bump version * fix validation if None * updated changelog * proposal to #659 formatted :) * bugfix proposal to #659 * refactor setting float charge_mode * fix type error, removed bluetooth cronjob * updated changelog * fix rs485 write communication errors by inserting sleeps, add debug print for charge mode and fix crash on write soc failures * fix write problem on set_soc. also changed the switch charge/discharge function, just in case * debug msg * Bluetooth optimizations * Fixes by @peterohman #505 (comment) * fix #712 * fix meaningless time to go values * fix meaningless time to go values * Duration of transition to float depends on number of cells * Float transition - Voltage drop per second * Update hlpdatabms4s.py * Validate setting of FLOAT_CELL_VOLTAGE and avoid misconfiguration * consider utils.LINEAR_RECALCULATION_EVERY to refresh CVL * cleanup * consider utils.LINEAR_RECALCULATION_EVERY to refresh CVL * small refactor, introduced set_cvl_linear function to set CVL only once every LINEAR_RECALCULATION_EVERY seconds * fix typo * updated changelog * remove debug msg * remove debug msg * undo debug change * Daly BMS make auto reset soc configurable * added debug and error information for CVL * fix proposal for #733 (#735) * Added: Tollerance to enter float voltage once the timer is triggered * Add bulk voltage Load to bulk voltage every x days to reset the SoC to 100% for some BMS * JKBMS disable high voltage warning on bulk reenable after bulk was completed * fixed error * disable high voltage warning for all BMS when charging to bulk voltage * fix error and change default value measurementToleranceVariation from 0.025 to 0.5 else in OffGrid mode max voltage is always kept * Added temperature names to dbus/mqtt * Use current avg of last 300 cycles for TTG & TTS * Calculate only positive Time-to-SoC points * added current average of last 5 minutes * make CCL and DCL more clear * fix small error * bugfix: LLTJBD BMS SOC different in Xiaoxiang app and dbus-serialbattery * black formatting * JDB BMS - Control FETs for charge, discharge and disable / enable balancer (#761) * feature: Allow to control charge / discharge FET * feature: Allow to enable / disable balancer * bugfix: Cycle Capacity is in 10 mAh Fixes SoC with factor 100 * 100% percentage * JBD BMS show balancer state in GUI page IO (#763) * Bump version * Fix typos * Smaller fixes - fixes #792 (comment) * Removed comments from utils.py This should make more clear that there are no values to change * Updated changelog * possible fix for LLT/JBS connection problems #769 #777 * bugfix: LLT/JBD BMS general packet data size check * improved reinstall and disable script * LLT/JBD BMS - Improved error handling and automatical driver restart in case of error. Should fix: - #730 - #769 - #777 * Fixed Building wheel for dbus-fast won't finish on weak systems Fixes #785 * Support for Daly CAN Bus (#169) * support for Daly CAN Bus * fix constructor args * revert port, needs fix * add can filters * comment logger Some changes are still needed to work with the latest version. They will follow in a next PR. --------- Co-authored-by: Samuel Brucksch <samuel@iternio.com> Co-authored-by: Manuel <mr-manuel@outlook.it> * JKBMS BLE - Introduction of automatic SOC reset (HW Version 11) (#736) * Introduction of automatic SOC reset for JK BMS (HW Version 11) * Fixed value mapping * Rework of the code to make it simpler to use without additional configuration. Moved execution of SOC reset. It's now executed while changing from "Float" to "Float Transition". * Implementation of suggested changes Persist initial BMS OVP and OVPR settings Make use of max_cell_voltage to calculate trigger value for OVP alert * Added: Daly CAN and JKBMS CAN * added CAN bms to installation script optimized CAN drivers * smaller fixes * Trigger JK BLE SOC reset when using Step Mode * Moved trigger_soc_reset() * fixes LLT/JBD SOC > 100% #769 * changed VOLTAGE_DROP behaviour * Fix JKBMS not starting if BMS manuf. date is empty * corrected bulk, absorption and soc reset terms * fix typo * add JKBMS_BLE debugging data * fix small error * added logging to config * add sleep before starting driver prevents lot of timeouts after reinstalling the driver, since the restart is now much faster than before * changed post install info * fix error * Daly BMS fixed embedded null byte #837 * added info for SoC reset to default config file * fix for #716 #716 * fix for #716 and JKBMS model recognition #716 * optimized logging * fix JKBMS recognition * added debugging * fixes #716 #716 --------- Co-authored-by: Holger Schultheiß <hoschult@web.de> Co-authored-by: Stefan Seidel <lkml@stefanseidel.info> Co-authored-by: Bernd Stahlbock <6627385+transistorgit@users.noreply.github.com> Co-authored-by: seidler2547 <github@stefanseidel.info> Co-authored-by: ogurevich <50322596+ogurevich@users.noreply.github.com> Co-authored-by: wollew <wollew@users.noreply.github.com> Co-authored-by: Oleg Gurevich <oleg@gurevich.de> Co-authored-by: peterohman <psub@fieber.se> Co-authored-by: Strawder, Paul <paul@coach-ai.com> Co-authored-by: Paul Strawder <paulidstein@gmail.com> Co-authored-by: Samuel Brucksch <SamuelBrucksch@users.noreply.github.com> Co-authored-by: Samuel Brucksch <samuel@iternio.com> Co-authored-by: ArendsM <136503378+ArendsM@users.noreply.github.com> Co-authored-by: Meik Arends <github@arends.info>
Louisvdw
pushed a commit
that referenced
this issue
Feb 28, 2024
* fix Sinowealth not loading #702 * fix unique identifier function * enable BMS over config, if disabled by default Now you can also add more then one BMS for BMS_TYPE * show battery port in log * ANT BMS fixes Fixed that other devices are recognized as ANT BMS * Sinowealth BMS fixes Fixed that other devices are recognized as Sinowealth BMS * improved publish_battery error handling switched from error count to seconds * Improve Battery Voltage Handling in Linear Absorption Mode * Refactor change time() to int(time()) for consistency in max_voltage_start_time and tDiff calculation * Refactor battery voltage calculations for efficiency and clarity * Remove penalty_buffer * Reset max_voltage_start_time wenn we going to bulk(dynamic) mode * updated changelog * fix reply processing * Reduce the big inrush current, if the CVL jumps from Bulk/Absorbtion to Float fix #659 * Check returned data lenght for Seplos BMS Be stricter about the return data we accept, might fix the problem of grid meters accidently being recognized as a Seplos * Validate current, voltage, capacity and SoC for all BMS This prevents that a device, which is no BMS, is detected as BMS * removed double check * bump version * fix validation if None * updated changelog * proposal to #659 formatted :) * bugfix proposal to #659 * refactor setting float charge_mode * fix type error, removed bluetooth cronjob * updated changelog * fix rs485 write communication errors by inserting sleeps, add debug print for charge mode and fix crash on write soc failures * fix write problem on set_soc. also changed the switch charge/discharge function, just in case * debug msg * Bluetooth optimizations * Fixes by @peterohman #505 (comment) * fix #712 * fix meaningless time to go values * fix meaningless time to go values * Duration of transition to float depends on number of cells * Float transition - Voltage drop per second * Update hlpdatabms4s.py * Validate setting of FLOAT_CELL_VOLTAGE and avoid misconfiguration * consider utils.LINEAR_RECALCULATION_EVERY to refresh CVL * cleanup * consider utils.LINEAR_RECALCULATION_EVERY to refresh CVL * small refactor, introduced set_cvl_linear function to set CVL only once every LINEAR_RECALCULATION_EVERY seconds * fix typo * updated changelog * remove debug msg * remove debug msg * undo debug change * Daly BMS make auto reset soc configurable * added debug and error information for CVL * fix proposal for #733 (#735) * Added: Tollerance to enter float voltage once the timer is triggered * Add bulk voltage Load to bulk voltage every x days to reset the SoC to 100% for some BMS * JKBMS disable high voltage warning on bulk reenable after bulk was completed * fixed error * disable high voltage warning for all BMS when charging to bulk voltage * fix error and change default value measurementToleranceVariation from 0.025 to 0.5 else in OffGrid mode max voltage is always kept * Added temperature names to dbus/mqtt * Use current avg of last 300 cycles for TTG & TTS * Calculate only positive Time-to-SoC points * added current average of last 5 minutes * make CCL and DCL more clear * fix small error * bugfix: LLTJBD BMS SOC different in Xiaoxiang app and dbus-serialbattery * black formatting * JDB BMS - Control FETs for charge, discharge and disable / enable balancer (#761) * feature: Allow to control charge / discharge FET * feature: Allow to enable / disable balancer * bugfix: Cycle Capacity is in 10 mAh Fixes SoC with factor 100 * 100% percentage * JBD BMS show balancer state in GUI page IO (#763) * Bump version * Fix typos * Smaller fixes - fixes #792 (comment) * Removed comments from utils.py This should make more clear that there are no values to change * Updated changelog * possible fix for LLT/JBS connection problems #769 #777 * bugfix: LLT/JBD BMS general packet data size check * improved reinstall and disable script * LLT/JBD BMS - Improved error handling and automatical driver restart in case of error. Should fix: - #730 - #769 - #777 * Fixed Building wheel for dbus-fast won't finish on weak systems Fixes #785 * Support for Daly CAN Bus (#169) * support for Daly CAN Bus * fix constructor args * revert port, needs fix * add can filters * comment logger Some changes are still needed to work with the latest version. They will follow in a next PR. --------- Co-authored-by: Samuel Brucksch <samuel@iternio.com> Co-authored-by: Manuel <mr-manuel@outlook.it> * JKBMS BLE - Introduction of automatic SOC reset (HW Version 11) (#736) * Introduction of automatic SOC reset for JK BMS (HW Version 11) * Fixed value mapping * Rework of the code to make it simpler to use without additional configuration. Moved execution of SOC reset. It's now executed while changing from "Float" to "Float Transition". * Implementation of suggested changes Persist initial BMS OVP and OVPR settings Make use of max_cell_voltage to calculate trigger value for OVP alert * Added: Daly CAN and JKBMS CAN * added CAN bms to installation script optimized CAN drivers * smaller fixes * Trigger JK BLE SOC reset when using Step Mode * Moved trigger_soc_reset() * fixes LLT/JBD SOC > 100% #769 * changed VOLTAGE_DROP behaviour * Fix JKBMS not starting if BMS manuf. date is empty * corrected bulk, absorption and soc reset terms * fix typo * add JKBMS_BLE debugging data * fix small error * Some changes for lost bluetooth connection / hci_uart stack restart * added logging to config * add sleep before starting driver prevents lot of timeouts after reinstalling the driver, since the restart is now much faster than before * changed post install info * fix error * Daly BMS fixed embedded null byte #837 * added info for SoC reset to default config file * fix for #716 #716 * fix for #716 and JKBMS model recognition #716 * optimized logging * fix JKBMS recognition * added debugging * fixes #716 #716 * Bind device instance to unique_identifier #718 * added data types to battery class disabled unused variables * save current charge state #840 * correct file permissions * updated changelog * added periodic saveChargeDetails * fix some small errors * fix issue with ruuvi tags When there are hundreds of unused ruuvi tags in the settings list that where added because thei where nearby the driver does not start correctly. These stale entries are disabled on the driver startup. The issue was already filed to Victron developers * CVL with i-controller instead of penaltysum * cvl_controller: switch to choose PenaltySum or ICOntroller + documentation * docu enhancement * Add setting and install logic for usb bluetooth module * round temperatures * changed battery disconnect behaviour * Fixes #891 #891 * updated changelog * Add bluetooth device note to config.default.ini * Fix typo in bluetooth note in config.default.ini * fixed error in new cvl_controller * fixed float division by zero and code optimization * Restart MAX_VOLTAGE_TIME_SEC if cell diff > CELL_VOLTAGE_DIFF_KEEP_MAX_VOLTAGE_TIME_RESTART * Calculation of the SOC based on coloumb-counting (#868) * Calculation of the SOC in the driver based on coloumb-counting * soc_calc: add current correction before integration * soc_calc: correction map for current * Soc_calc: CorrectionMap, switch to turn on/off correction, selectable initial value * soc_calc: Bugfix * soc_calc: Bugfix * store soc in dbus for restart * store soc in dbus for restart (formatted) * store soc in dbus for restart (bugfix) * save soc_calc only after change > 1.0 * store soc in dbus for restart (bugfix) * logger does not work this way. do not know why * writing and reading to dbus works * Removed options: SOC_CALC_CURRENT_CORRECTION, SOC_CALC_RESET_VALUE_ON_RESTART, SOC_CALC_INIT_VALUE sort soc_calc alphabetically * fixed comments * Updated changelog, small fixes * Changed: PUBLISH_CONFIG_VALUES from 0/1 to True/False * Changed: Code optimizations - Changed some variables to be more clear - Added comments for easier code understanding * Calculated SOC: Added two decimals, added BMS SOC for MQTT & Node-RED * Updated changelog, small fixes * Changed: PUBLISH_CONFIG_VALUES from 0/1 to True/False * Changed: Code optimizations - Changed some variables to be more clear - Added comments for easier code understanding * Calculated SOC: Added two decimals, added BMS SOC for MQTT & Node-RED * Fix #898 #898 * Changed: Fix issue loading settings from dbus * Added nightly install option makes it easier for users to pretest fixes * Changed: more detailed error output when an exception happens * Possible fix for #912 #912 * Fixes #919 #919 * Changed: Exit script with error, if port excluded else the serialstarter stops at the dbus-serialbattery * Fixed some smaller errors * Updated pre-release workflow * Fix JK BMS connection restart when bluetooth fails. This fix installs a new thread to monitor the state of the original scraping thread. If scraping thread dies, it verifies that it did not because the scraping was intentionally stopped by calling stop_scrapping. When restarting the scrapper, it first calls the bluetooth reset lambda function that was passed in the class contructor, such that bluetooth is ready to make a proper connection. * Fixes #916 #916 * Added Venus OS version to logfile * Fix #840 #840 * Small code formatting fixes * Optimized reinstall script. Restart GUI only on changes. * Display debugging data in GUI when DEBUG enabled * Install script now shows repositories and version numbers * Update daly_can.py Fixing #950 for DalyBMS * Update jkbms_can.py Fixing #950 for Jk BMS * Fix black lint check * Fixes #970 #970 * Fixed some errors in restoring values from dbus settings * Moved sleep on start for all BMS * Update config description * Reworked a part of the default config * fix typo in stopping services when reinstalling * Fix Time-to-SoC and Time-to-Go calculation * Add changelog info * Round sum and diff voltage * Temperature limitation variables where changed * SoC limitation variables where changed * Added error messages * Remove unneeded code * Reset SoC to 0% if empty * Add GUIv2 for dbus-serialbattery * Check free space before installing * Added new GUIv2 version * Removed Python 2 compatibility * Changelog update * Code cleanup - Removed: get_temperatures() - Removed: update_last_seen() * Bluetooth code optimizations * Fixed some JKBMS BLE not starting #819 * Check if packages are already installed before install * Fixed some SOC calculation errors * Fixed None SOC on driver start * Do not show and allow button change when callback is missing for: - ForceChargingOff - ForceDischargingOff - TurnBalancingOff * Check if a device instance is already used by creating a PID file * Log and execute SOC reset to 100% or 0% only once * Update GitHub workflow and issue templates * Fixed LLT/JBD BMS with only on temperature sensor #791 #971 * Fix warning on reinstall * Fix missing IO control for JBDBMS #992 #992 * Prepare for removing dev branch --------- Co-authored-by: ogurevich <50322596+ogurevich@users.noreply.github.com> Co-authored-by: Bernd Stahlbock <6627385+transistorgit@users.noreply.github.com> Co-authored-by: wollew <wollew@users.noreply.github.com> Co-authored-by: Oleg Gurevich <oleg@gurevich.de> Co-authored-by: peterohman <psub@fieber.se> Co-authored-by: Strawder, Paul <paul@coach-ai.com> Co-authored-by: Paul Strawder <paulidstein@gmail.com> Co-authored-by: Samuel Brucksch <SamuelBrucksch@users.noreply.github.com> Co-authored-by: Samuel Brucksch <samuel@iternio.com> Co-authored-by: ArendsM <136503378+ArendsM@users.noreply.github.com> Co-authored-by: Meik Arends <github@arends.info> Co-authored-by: Marvo2011 <Marvin.Edeler@gmail.com> Co-authored-by: cflenker <christian.flenker@gmx.de> Co-authored-by: cflenker <125555670+cflenker@users.noreply.github.com> Co-authored-by: Cupertino Miranda <cupertinomiranda@gmail.com> Co-authored-by: Martin Polehla <p0l0us@users.noreply.github.com>
|
This problem still exists. |
|
@fableman73 did you also use a RaspberryPi 2 w? Whats the driver version you tried and did you try the solutions from above? I haven't experienced the problem since then and had to install the latest stable release from mr-manuel today because of the appearance of another bug that was reported in nov. 2023 and seems to be fix in the current stable. |
|
If I use export SKIP_CYTHON=false Looking in indexes: https://pypi.org/simple, https://www.piwheels.org/simple The installation will not hang But I get this error when I try to use mpp-solar: `root@husb:~# mpp-solar -p C8:47:81:17:3E:36 -c getInfo Exception ignored in: <function Client.del at 0x7f9c7e24c0> |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Describe the bug
I tried to connect two JKBMS via bluetooth on an freshly installed venosOS, but the RP 2 w i'm using keeps getting stuck while "building wheels for dbus-fast .." and seems to crash after a while. Even manuell installing of bleak via "pip3 install bleak -no-cache-dir" fails while "building wheels for dbus-fast" and seems to show the same behavior. I even tried to reinstall venusOS, but i keep getting the same behavior. Any ideas where to find the root of the problem?
How to reproduce
Steps to reproduce the behavior:
Expected behavior
installing bleak
Driver version
v1.0.20230531
Venus OS device type
Raspberry Pi
Venus OS version
v3.01
BMS type
JKBMS / Heltec BMS
Cell count
16
Connection type
Bluetooth
Config file
Relevant log output
Any other information that may be helpful
There are two MPPTs connected via USB and an MK3 with 3 Multiplus-II 3000. The Batteries are close to the RP but the wifi signal is weak (only two bars are shown on the remote-console.)
The text was updated successfully, but these errors were encountered: