Confused on settings after executing M502

scyto · January 5, 2022, 9:13pm

I have A350, i upgraded the rails and head, firmwares auto updated, i noted that the numbers in M501 were different, assumed that was right because of upgraded rails and head.

I have been having odd printing issues (severe under extrusion on first layer) and i am struggling to know what changed because i have replace rails, head, hotend, filament batch, updated all my slicers, etc) so trying to go back a little.

I have put the old head back on, i reset from touchscreen and reset using M502. I am confused by output of M502 because its seems to have reverted to the old rails settings…

Before resetting with M502

Recv: echo:Steps per unit:
Recv: echo: M92 X160.00 Y160.00 Z400.00 B888.89 E228.18
Recv: echo:Maximum feedrates (units/s):
Recv: echo:  M203 X300.00 Y300.00 Z12.00 E120.00
Recv: echo:Maximum Acceleration (units/s2):
Recv: echo:  M201 X1000.00 Y1000.00 Z1000.00 E1000.00

after resetting with M502

Recv: echo:Steps per unit:
Recv: echo: M92 X400.00 Y400.00 Z400.00 B888.89 E212.21
Recv: echo:Maximum feedrates (units/s):
Recv: echo:  M203 X120.00 Y120.00 Z40.00 E25.00
Recv: echo:Maximum Acceleration (units/s2):
Recv: echo:  M201 X3000.00 Y3000.00 Z100.00 E10000.00

(i expected E212.21 as thats what a factory reset always did)
the other numbers are confusing me… my suspicion is M502 just did the wrong thing…

–edit–
could someone with upgraded rails see what their setting are as i no longer trust the ‘upgraded’ settings

scyto · January 5, 2022, 9:16pm

i am reapplying firmware in hope this fixes

scyto · January 5, 2022, 9:19pm

no it doesn’t and the huge clonk noise from X and Y during a home is very concerning

FFS snapmaker get it together, your QC and testing just stinks

i opened a guthub issue for you snapmaker, i expect to see this resolved ASAP
Executing M502 on upgraded A350 sets wrong settings, risks damaging new v2 linear rails. · Issue #201 · Snapmaker/Snapmaker2-Controller (github.com)

brent113 · January 5, 2022, 9:51pm

What looks like is supposed to happen is this flag gets reset:

github.com

Snapmaker/Snapmaker2-Controller/blob/c703f97e4f349fedf411813ebb77b941714a2700/Marlin/src/module/configuration_store.cpp#L2120

    
      
          /**
           * M502 - Reset Configuration
           */
          void MarlinSettings::reset() {
            static const float tmp1[] PROGMEM = DEFAULT_AXIS_STEPS_PER_UNIT, tmp2[] PROGMEM = DEFAULT_MAX_FEEDRATE;
            static const uint32_t tmp3[] PROGMEM = DEFAULT_MAX_ACCELERATION;
            uint8_t temp_axis_to_port[X_TO_E] = DEFAULT_AXIS_TO_PORT;
            LOOP_X_TO_EN(i) {
              axis_to_port[i] = temp_axis_to_port[i];
            }
            planner.is_user_set_lead = false;
            LOOP_X_TO_EN(i) {
              planner.settings.axis_steps_per_mm[i]          = pgm_read_float(&tmp1[ALIM(i, tmp1)]);
              planner.settings.max_feedrate_mm_s[i]          = pgm_read_float(&tmp2[ALIM(i, tmp2)]);
              planner.settings.max_acceleration_mm_per_s2[i] = pgm_read_dword(&tmp3[ALIM(i, tmp3)]);
            }
          
          
  planner.settings.min_segment_time_us = DEFAULT_MINSEGMENTTIME;
            planner.settings.acceleration = DEFAULT_ACCELERATION;
            planner.settings.retract_acceleration = DEFAULT_RETRACT_ACCELERATION;
            planner.settings.travel_acceleration = DEFAULT_TRAVEL_ACCELERATION;

And then on next power on the steps/mm is set based on what’s read from the linear module:

github.com

Snapmaker/Snapmaker2-Controller/blob/c703f97e4f349fedf411813ebb77b941714a2700/snapmaker/src/module/linear.cpp#L251

    
      
          if ((mac_index_[LINEAR_AXIS_Y1] != 0xff) || (mac_index_[LINEAR_AXIS_Y2] != 0xff)) {
            axis_steps_per_unit[Y_AXIS] = mac_index_[LINEAR_AXIS_Y1] != 0xff ? lead_[LINEAR_AXIS_Y1] : lead_[LINEAR_AXIS_Y2];
          }
          
          
if ((mac_index_[LINEAR_AXIS_Z1] != 0xff) || (mac_index_[LINEAR_AXIS_Z2] != 0xff) || (mac_index_[LINEAR_AXIS_Z3] != 0xff)) {
            axis_steps_per_unit[Z_AXIS] = mac_index_[LINEAR_AXIS_Z1] != 0xff ? lead_[LINEAR_AXIS_Z1] : mac_index_[LINEAR_AXIS_Z2] != 0xff ? lead_[LINEAR_AXIS_Z2] : lead_[LINEAR_AXIS_Z3];
          }
          
          
LOOP_XYZ(i) {
            if (!planner.is_user_set_lead) {
              planner.settings.axis_steps_per_mm[i] = axis_steps_per_unit[i];
            }
            SERIAL_ECHOLNPAIR("axis index:", i, "  pitch:", planner.settings.axis_steps_per_mm[i]);
          }
          
          
planner.refresh_positioning();
          
          
// check if all linear modules are the same generation
          for (i = LINEAR_AXIS_X1; i < LINEAR_AXIS_MAX; i++) {
            if (mac_index_[i] == 0xff) continue;

It requires a power cycle - did you try that by chance?

If that works you might clarify in your github issue that maybe M502 should force a reboot afterwards, or otherwise read the values from the linear module again.

brent113 · January 5, 2022, 10:22pm

Just found another edge case they missed:

github.com

Snapmaker/Snapmaker2-Controller/blob/c703f97e4f349fedf411813ebb77b941714a2700/Marlin/src/gcode/config/M92.cpp#L89

    
      
                    planner.max_jerk[E_AXIS] *= factor;
                  #endif
                  planner.settings.max_feedrate_mm_s[E_AXIS_N(target_extruder)] *= factor;
                  planner.max_acceleration_steps_per_s2[E_AXIS_N(target_extruder)] *= factor;
                }
                planner.settings.axis_steps_per_mm[E_AXIS_N(target_extruder)] = value;
              }
              else {
                planner.settings.axis_steps_per_mm[i] = parser.value_per_axis_units((AxisEnum)i);
              }
              planner.is_user_set_lead = true;
            }
          }
          planner.refresh_positioning();
          
          
#if ENABLED(MAGIC_NUMBERS_GCODE)
            #ifndef Z_MICROSTEPS
              #define Z_MICROSTEPS 16
            #endif
            const float wanted = parser.floatval('L');
            if (parser.seen('H') || wanted) {

The is_user_set_lead flag is set for any axis, including the E axis.

It’s possible that someone calibrates e-steps, then upgrades to the new linear modules - that will not be correctly handled also leading to this problem.

scyto · January 5, 2022, 10:44pm

I think it did that when i tried to fix reapplying the the firmware?

brent113 · January 5, 2022, 10:46pm

Try M502, then M500 to save, then power cycle? M502 used to immediately write to EEPROM but I’m not seeing that at the moment.

scyto · January 6, 2022, 12:13am

This is indeed correct.
Doing this, then hard power reboot, M503 then reports right settings.

Topic		Replies	Views
Question about New Linear Rails upgrade? Snapmaker 2.0 (3-in-1)	5	909	October 17, 2022
Horrible grinding noise on Y Axis after new LM Install Snapmaker 2.0 (3-in-1)	19	2396	July 3, 2023
M500 Command Problems Snapmaker 2.0 (3-in-1)	0	439	December 12, 2021
Things that change with doing a firmware-update Snapmaker 2.0 (3-in-1)	8	1026	May 15, 2021
Calibration Issues After Update Snapmaker 2.0 (3-in-1)	7	1511	December 12, 2021

Confused on settings after executing M502

Related topics