Auto_Bed_Leveling: Enhanced G29 command

[RobH2]

I appreciate your time. I'm not a noob. I just can't seem to get things settled in.

Here's a screen shot:
1stLayer.jpg

The Slic3r default recommended is 0.35. Maybe I'm misunderstanding this setting for 'First layer height:' and need to set it to be 0.20 also. Is that what you are trying to get across to me? Now that I think of it, if the 'First layer height:' is set to 0.35mm, then the hot end needs to be 0.35mm away from the bed to do that. Hmmm, this is probably it now that I wrap my head around it.

Off to do a new test. Thanks for checking in and getting me to do better analysis of parameters.

[brucehvn]

I appreciate your time. I'm not a noob. I just can't seem to get things settled in.

Here's a screen shot:
1stLayer.jpg

The Slic3r default recommended is 0.35. Maybe I'm misunderstanding this setting for 'First layer height:' and need to set it to be 0.20 also. Is that what you are trying to get across to me? Now that I think of it, if the 'First layer height:' is set to 0.35mm, then the hot end needs to be 0.35mm away from the bed to do that. Hmmm, this is probably it now that I wrap my head around it.

Off to do a new test. Thanks for checking in and getting me to do better analysis of parameters.

.35 is not a bad first layer height. Can you paste about the first 25 lines of the generated gcode here, so we can see what Slic3r is outputting? And yes, if your first layer height is set to .35, then the first printing move should place the nozzle .35 off the bed.

Bruce

[RobH2]

Understood. I'm getting predictable performance now. It was all about that 'First layer height' that I was misinterpreting. I have my paper gap (just measured that paper and it's 0.11mm) so I''ll now respect that 'First layer height' properly. Thanks for all of your guidance. We all know a lot of stuff but none of us knows everything. That's what makes forums so powerful. The expertise that is shared in collaboration is just priceless. Sometimes we get so wrapped up we can't see the forest for the trees.

Thanks again. I'm really pleased with the consistency in gaps that I'm getting now.

[RobH2]

Zeroing in now. So if .35 is not a bad first layer height why is everyone struggling so hard to get that perfect .2mm gap? I guess in reality, as long as the filament is getting pushed into the bed, you'll have good adhesion. And if you have a .5mm nozzle, the .35 is still being pushed in 30% of that diameter. So it you are at least pushing it in and not having it "fall" onto the bed, then the most important thing is "level" and, that's what this whole autoleveling thing is all about. Getting that first layer as consistent as possible. Whether it's .2mm or .35mm, it's about level. Correct?

I'm convinced that the 'First layer height:' setting was my culprit. I kept having to adjust the head down manually and about .15 was what I was finding it off. Now I know. I'll go check my other slicers for similar 'First layer height' parameters. I bet I'll be surprised and will solve this.

[brucehvn]

Zeroing in now. So if .35 is not a bad first layer height why is everyone struggling so hard to get that perfect .2mm gap? I guess in reality, as long as the filament is getting pushed into the bed, you'll have good adhesion. And if you have a .5mm nozzle, the .35 is still being pushed in 30% of that diameter. So it you are at least pushing it in and not having it "fall" onto the bed, then the most important thing is "level" and, that's what this whole autoleveling thing is all about. Getting that first layer as consistent as possible. Whether it's .2mm or .35mm, it's about level. Correct?

I'm convinced that the 'First layer height:' setting was my culprit. I kept having to adjust the head down manually and about .15 was what I was finding it off. Now I know. I'll go check my other slicers for similar 'First layer height' parameters. I bet I'll be surprised and will solve this.

In reality Z=0 should mean the nozzle is touching the bed. However, we usually calibrate the bed with the nozzle cold, so we allow for heat expansion by leaving just a paper width gap (or about .1mm). So theoretically, when the nozzle is at temperature, Z=0 will have no gap. During printing, the nozzle will always be the height corresponding to the layer height. So in your case, the first layer will put z=0.35, the second layer will have z=0.55, etc. The adhesion comes from the volume of filament being extruded. In Slic3r, the default is to use a 200% extrusion width on the first layer. So while your nozzle is still 0.35 above the bed, the volume of filament is causing it to get smooshed onto the bed and hopefully causing good adhesion. Cura just recently in the latest version allows you to set the extrusion width for the first layer. It does have the first layer height setting and I think it limits it to less than Slic3r's default. Triffid Hunter's calibration guide says that you should really never print at a layer height > 80% of your nozzle width and I usually try to respect that when setting the first layer height. I have a .4mm nozzle, so I don't set the first layer height > .32. I think Cura won't allow me to set anything above .30.

Bruce

[RobH2]

Still getting an error when I compile. From Post #75 I'm not real sure where to put this:

In ConfigurationStore.h:
#ifdef EEPROM_SETTINGS
void Invalidate_EEPROM_Settings(); // Roxy routine to guarantee we are using default firmware values
void Config_StoreSettings();
void Config_RetrieveSettings();
#else
FORCE_INLINE void Config_StoreSettings() {}
FORCE_INLINE void Config_RetrieveSettings() { Config_ResetDefault(); Config_PrintSettings(); }
#endif

I put it after:

29: #define EEPROM_READ_VAR(pos, value) _EEPROM_readData(pos, (uint8_t*)&value, sizeof(value))

Here's the error:

Arduino: 1.5.6-r2 (Windows 7), Board: "Arduino Mega or Mega 2560, ATmega2560 (Mega 2560)"

ConfigurationStore.cpp: In function 'void Invalidate_EEPROM_Settings()':
ConfigurationStore.cpp:35: error: 'EEPROM_OFFSET' was not declared in this scope

Edit: I put the snippet for ConfigurationStore.cpp at the end after everything else and it seems to compile. So for now disregard this post.

[Tachout]

I just watched a Solidoodle video and the Autobed Leveling on that checked the bed in 9 places. I watched a video on You Tube stating to check in 3 places. I am going to be putting this in on my Marker Farms i3V and wondered if there was an advantage to taking the numbers in multiple places? I would think it would be more accurate. Opinions?

[Roxy]

3 points is really not enough. The problem is the corner of the bed you did not probe is going to magnify any error in the three points you did measure. Probing 3 x 3 works well, but I do get slightly better results doing a 4 x 4 matrix. I think part of the reason is any error gets more of a chance to be averaged out.

I would suggest printing one of those calibration prints that puts a few large squares down close to the edges of your print bed. You will be able to see if the amount of measurement error you have makes a difference between doing a 3 x 3 or a 4 x 4 probe matrix. If the first layers go down fine either way, it may not be worthwhile to do a 4 x 4 probe on most stuff.

[Roxy]

Rob,

I apologize once again... It shouldn't be this difficult for you! Obviously, the directions aren't good enough!!!

Here are my ConfigurationStore files. You should be able to just lay these down on top of your files and have the same code as me. Let me know if any hiccups happen.

And actually.... The ConfigurationStore.cpp file (that is attached) has an extra line of code that really should be in there. When you print the values using an M503 command, somehow, nobody figured the Z-Probe_Offset needed to be in the displayed data. It really should be there because without that display of data you have no idea what the printer is using. The attached file has that extra variable displayed.

[RobH2]

Hey Roxy. Thanks for this. I'll work on it tonight. No need to apologize. The good will and work you do here is amazing. Don't worry about it all.