[bugfix-2.0.x] Error:EEPROM datasize error when enabled 2 servos (dondolo + bltouch)

[alexxy]

Hi!

I get this error with recent bugfix-2.0.x branch. Configs attached
config.zip

Compiled: Oct  5 2018
Free Memory: 2724  PlannerBufferBytes: 1488
Error:EEPROM datasize error.
ttyUSB0 PC> Error:EEPROM datasize error.

Index: 607 Size: 615
Hardcoded Default Settings Loaded

[gloomyandy]

Have you tried doing an m502 followed by m500 to reset the default values and store them in eeprom. You will often usually to do this when installing a new version of the firmware.

[alexxy]

yes. doesnt work for me. actualy it wasfirst thing that i tryed.

[blairq]

im getting it too. also the machine started skipping steps and so... already did a m502 and m500. changed z_offset (i use a touch sensor) , saved the changes but they didnt go to the eeprom.
flashed 2 times. same error.

[GMagician]

@alexxy @blairq have you tried with DEBUG_EEPROM_READWRITE and check log

[blairq]

no, its related to c40ea1f

i changed back #define EEPROM_OFFSET 0 to the original #define EEPROM_OFFSET 100 and everything works fine again.

had to do a m502 - m500 again

[mrambossek]

i can confirm that there seems to be some kind of issue. after upgrading last night, i keep getting

Error:EEPROM CRC mismatch - (stored) XXXXX != YYYYY (calculated)!
echo:Hardcoded Default Settings Loaded

EDIT: i did M502/M500 .. many times. i am aware of what it does. i still kept getting the crc errors.

i changed EEPROM_OFFSET back to 100 as suggested by #12015 (comment) and now it seems to work again. so appearently those 100 bytes ARE somehow used? :)

[alexxy]

I tryed with DEBUG_EEPROM_READWRITE

SENDING:M500
Error:Field lcd_preheat_hotend_temp mismatch.
Error:Field lcd_preheat_hotend_temp mismatch.

[alexxy]

Actualy error is the same as for #11883 and #11757

[GMagician]

@thinkyhead in servo.h there is something not correct I think extern uint16_t servo_angles[NUM_SERVOS][2];. Not related with this but may lead to memory corruption since I think it should use NUM_SERVO_PLUGS

Edit: Maybe not corruption since NUM_SERVOS should always be bigger but I think not correct anyway
Also in servo.ccp there is uint16_t servo_angles[NUM_SERVOS][2];

Edit2: Ok I don't understand... in eeprom data struct is used NUM_SERVO_PLUGS hence this should be used all over the code but this is not true... is it not too weird all this?

[GMagician]

@thinkyhead sorry to bother but shouldn't I see eeprom_index changed every read/write? I see it is incremented only in EEPROM_SKIP

[alexxy]

I have NUM_SERVOS = 2 in Configuration.h
But may be problem is that i use BLTOUCH? and enabled editable servo angles?

[alexxy]

Disableing EDITABLE_SERVO_ANGLES solves issue (so seems like its a root of problem)

[GMagician]

@alexxy not sure about problem but I think that this mixing of NUM_SERVOS and NUM_SERVO_PLUGS on servo_angles may lead to some problem...

[GMagician]

@alexxy that confirm me that such mix is the issue...

[alexxy]

NUM_SERVO_PLUGS not defined in Marlin/src/module/servo.h

[GMagician]

@alexxy try replacing NUM_SERVOS with NUM_SERVO_PLUGS in servo.cpp and servo.h where servo_angles are referred

[GMagician]

If it will fix I'll post a PR

[GMagician]

OT @alexxy your dondolo works fine? I decided to remove it because of too much problem with not in use filament retraction

[alexxy]

@GMagician it works. but its hard to calibrate offsets

[GMagician]

@GMagician it works. but its hard to calibrate offsets

Dondolo? or fix? If dondolo is no a problem of offsets. The problem is that the unused filament is not kept in position when head moves. When you come back to such filament you miss material (you have to create a tower but then no benefit from having two nozzles)

[alexxy]

Dondolo works.

[cjsoong]

@alexxy
If your problem has not been resolved, you can try my modified version.
No EEprom data size error & M281 can set the servo angle.
https://drive.google.com/file/d/1dixj-e1lBUvuxazjgOgqVz0tbXnl2zKI/view?usp=sharing

[monterman]

@alexxy could you help me with my issue. see here please teemuatlut#27

thank you.

[github-actions]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.