Implement fault handling of safety controller and document it #5

New Issue

mhu · 2020-08-16T01:30:07+02:00

mhu commented

2020-08-16 01:30:07 +02:00

Safety Controller

must stop PID in case of severe error
must force oven driver output to 0 in case of a severe error
Must enter panic mode in case of non recoverable hard error.

Severe errors

These errors stop the PID and force the oven off but do not force user to restart the whole device.

Analog Watchdog flag set: ERR_FLAG_MEAS_ADC_WATCHDOG
Timing monitor violation of PID: ERR_FLAG_TIMING_PID
Timing monitor violation of measurement ADC: ERR_FLAG_TIMING_MEAS_ADC
Measurement ADC off: ERR_FLAG_MEAS_ADC_OFF
Reference voltage analog monitor: ERR_FLAG_AMON_VREF
measurement ADC overflow: ERR_FLAG_MEAS_ADC_OVERFLOW
All hard errors, listed below

Hard Errors

These errors force the reflow controller into panic mode. It will not serve a shell or GUI anymore and can only be restarted. Panic mode is still to be defined in detail. Panic can't be terminated without power cycling the device.

The following errors trigger the panic mode:

Hanging main loop: ERR_FLAG_TIMING_MAIN_LOOP
Fired watchdog: ERR_FLAG_WTCHDG_FIRED
Over temperature of uC: ERR_FLAG_AMON_UC_TEMP
Hanging systick: ERR_FLAG_SYSTICK
Stack overflow / stack-heap-collision protection: ERR_FLAG_STACK
Safety ADC error: ERR_FLAG_SAFETY_ADC
Check issue #15. If a broken structure of the safety stuff is detected, the controller shall also go into panic mode.

Safety Controller * must stop PID in case of severe error * must force oven driver output to 0 in case of a severe error * Must enter panic mode in case of non recoverable hard error. ## Severe errors These errors stop the PID and force the oven off but do not force user to restart the whole device. 1. Analog Watchdog flag set: `ERR_FLAG_MEAS_ADC_WATCHDOG` 2. Timing monitor violation of PID: `ERR_FLAG_TIMING_PID` 3. Timing monitor violation of measurement ADC: `ERR_FLAG_TIMING_MEAS_ADC` 4. Measurement ADC off: `ERR_FLAG_MEAS_ADC_OFF` 5. Reference voltage analog monitor: `ERR_FLAG_AMON_VREF` 6. measurement ADC overflow: `ERR_FLAG_MEAS_ADC_OVERFLOW` 7. **All hard errors, listed below** ## Hard Errors These errors force the reflow controller into panic mode. It will not serve a shell or GUI anymore and can only be restarted. Panic mode is still to be defined in detail. Panic can't be terminated without power cycling the device. The following errors trigger the panic mode: 1. Hanging main loop: `ERR_FLAG_TIMING_MAIN_LOOP` 2. Fired watchdog: `ERR_FLAG_WTCHDG_FIRED` 3. Over temperature of uC: `ERR_FLAG_AMON_UC_TEMP` 4. Hanging systick: `ERR_FLAG_SYSTICK` 5. Stack overflow / stack-heap-collision protection: `ERR_FLAG_STACK` 6. Safety ADC error: `ERR_FLAG_SAFETY_ADC` 7. Check issue #15. If a broken structure of the safety stuff is detected, the controller shall also go into panic mode.

mhu commented

2020-08-21 00:48:15 +02:00

Add this information to the sphinx docu!

mhu referenced this issue from a commit

2020-08-22 23:57:30 +02:00

Issue #5: Write documentation for error flags and their handling

mhu referenced this issue from a commit

2020-09-27 22:54:11 +02:00

issue #5: Implement safety weight checking in control loop

mhu referenced this issue from a commit

2021-05-22 00:21:47 +02:00

Issue #15: Implement safety weight table...

mhu commented

2021-05-22 11:51:50 +02:00

Change behavior:

After a panic occured, the controller will restart normally, but report that a panic occured and you have to check the safety memory for details.

Change behavior: After a panic occured, the controller will restart normally, but report that a panic occured and you have to check the safety memory for details.

mhu referenced this issue from a commit

2021-05-22 12:02:05 +02:00

Issue #5: Implement startup screen in case apanic happened and prevent the PID controller from starting when a watchdog reset has occured

mhu added a new dependency 2021-12-29 01:18:16 +01:00

#33 Write documentation for v1

Sign in to join this conversation.