Auto Bed Levelling Glitches

[MiniMadRyan]

Thanks Roxy, I really appreciate it! I've been meddling around the code, and frankly, I'm kinda surprised at some of the complexity of it. You're right though, it would seem logical and beneficial if those values were unified, In my haste last night I assumed they were, hence my shock and confusion with the values and what was going on.

I think you're right though, in that I will just need to end up adjusting my Z_PROBE_OFFSET distance. While it seems absurd, it makes sense. Really I was expecting a value similar to what others have had, in the 7-8 range, though with doing that, my numbers will be probably around the 15-16...but I guess what works, works!

AbuMaia, after two G28, my M114 Z reports back at 10. I believe you can see that in my previous outputs I've posted. A further G1 Z0 brings the head down a bit, but I still need to manually adjust the head down a further -7.9 until the nozzle is at a good height for printing. As my RAISE_BEFORE_PROBING is 5, if I modify my code per your suggestion, my offset is actually 12.9. That works great then for any G28 then G1 Z0, however the offset is still way off after the G29.

[AbuMaia]

My fix wasn't for adjusting the Z probe offset, it was for correcting some bad math that's done somewhere in Marlin I can't find, that caused my hotend nozzle to be too high after a G28. It's separate from the offset. That's where we're getting confused. My fix adjusts where Marlin *thinks* the hotend is to be where it really is.

[MiniMadRyan]

Ahh, thank you for that AbuMaia. I guess its a matter of just tweaking and playing with numbers now. I find it rather intriguing, that, most of our units are relatively the same, in terms of electronics, and hardware design, yet the differences between each of our firmwares can be so great, but I'm thankful for yours, Roxy's and everyone else's help and patience for all the silly questions and comments I've had, this has been the most challenging, yet rewarding printer I've owned or built yet!

[AbuMaia]

Here you go OME. http://www.thingiverse.com/thing:335632 This is the one I use on the i3v.

I used this one http://www.thingiverse.com/thing:167440 with the i3.

[MiniMadRyan]

well I'm back to fiddling again after a short holiday and some more pressing projects.

Fiddled with the Z probe offset values quickly, and was just wondering, would I only see the effects of this change after starting a print? What I did was changed that variable, then ran a G28 followed by a G29 and then lastly a G1 Z0...all values were the same prior to modifying it. I understand that my G28 wouldn't be affected, but I always assumed the following commands above would simulate the start of a print.... oh well, will try some prints later this week when I get more time!

[MiniMadRyan]

Well this is weird. So as Roxy suggested, I modified my Z_PROBE_OFFSET_FROM_EXTRUDER and then ran some prints.

My real offset difference was -8.60, so what I did was change this to -10.60 thinking that, since I know the extruder is more than 4mm off of the surface when it prints, this would be fine.

Ran the print, and still too high. Killed the print, and manually measured the difference between the extruder and bed. -9.60mm. I then changed my Z_PROBE_OFFSET_FROM_EXTRUDER to 12.60, theoretically then, on my next print it should bring the head down to -7.60 difference....but to my surprise, the second print, still too high, when cancelled and measured is still almost exactly -9.60 difference.

Right, I haven't a clue where this is all going wrong now.... puzzled!

[MiniMadRyan]

Alright, I removed my Z_RAISE_BEFORE_HOMING thinking maybe that would help. No change in the outcome though.

Most interestingly though, with the Z_RAISE_BEFORE_HOMING set to 0, when the printer moves to the centre of the bed, the X &Y Motion is much much quicker, so much so that I believe I heard the y motor jump a bit, and confirmed it did, when after I killed the print, the bed return crashed slightly.....

[MiniMadRyan]

That's it, I give up. I've tried everything I can think of and everything that's been suggested here. I used AbuMaia's suggestion and have my Z Height half decent after running a G28, G1 Z0.

Everything else though, what the heck!??!!!! I modified my Z_PROBE_OFFSET_FROM_EXTRUDER a FULL 4MM from what it actually is (8.60 to 12.60) looking for any change, and yet my head is still 9mm too high....I don't ******* get it!!!

[AbuMaia]

MMR: in your marlin_main.cpp, find this code:

      #ifdef ENABLE_AUTO_BED_LEVELING
        if((home_all_axis) || (code_seen(axis_codes[Z_AXIS]))) {
          current_position[Z_AXIS] += zprobe_zoffset;  //Add Z_Probe offset (the distance is negative)
        }

Take the distance you found after doing a G28 then G1 Z0, then moving the nozzle to the print bed, and add it to the zprobe_zoffset, like so:

      #ifdef ENABLE_AUTO_BED_LEVELING
        if((home_all_axis) || (code_seen(axis_codes[Z_AXIS]))) {
          current_position[Z_AXIS] += zprobe_zoffset+4;  //Add Z_Probe offset (the distance is negative) *added +4 since my Z0 was too high*
        }

That should help keep the G28 Z0 where it is supposed to be. This should not affect the G29, however.

[printbus]

Has anyone ever looked into tweaking how the Arduino IDE uses the avrdude programming utility? There's no reason why avrdude can't reprogram EEPROM values when reflashing the firmware. I do that all the time with avrdude in my non-Arduino AVR projects. Not always since sometimes I do want to retain the old EEPROM data, but it's a simple command line option with avrdude to go either way.