The Calibrate function is not 'Automatic', the user has to understand what the system is doing and perform set of actions to achieve whatever 'setting' he/she is aiming for.  That setting is the lowest value of the three values displayed.  The user has to choose a 'range'.  The system has an acceptable range. It reports pass or fail if the three values are within a system range.

For this next test, I'm aiming for a setting of 200 and a range of 10.  After doing a load of laundry and cleaning up my kitchen while I went through about a dozen calibration cycles, I managed to not pull out my hair and got 195,194,196 with a calculated setting of '194'.   To me, with the model I have, I think this calculation is as simple as finding the smallest number   Fortunately I'm not brain dead, so I can perform this 'calculation' in my head.

Since I'm committed to the stock calibration process, I am going to make a physical modification to the printer: Now that I know the studs are M3 posts, (and I've learned how to pop off the side walls), I'm going to take the print bed off and design a better adjustment nut to make it easier to 'calibrate'. - er - level.

The print at 194 was just about identical to the print at 154, so I'll go to a higher number.

that is due to same old issue. upper layers cooling too fast causing plastic to contract which 'peels' uncured threads loose across gaps like inside hollow curves or in this case between the higher sides looking like hairs of piano/guitar strings.

this is much worse on larger, taller objects like a vase for example. some will stretch until cured then break like they did inside my vacuum Cyclone separator inside making for very difficult cleanup. this is not just a type/brand of printer issue. any extruder type 3D printer can have this issue with some materials like ABS that are fussy about shrinkage with more than slight temp differences in layers.

Hi n2ri, 

I'm well aware of the 'oldness' of the shrink issue and the reasons for it.  This thread is really about me examining the print quality of the Da Vinci, which is dependent on  basic issues applicable to all machines, and specific issues that are due to the Da Vinci . . . and perhaps specific to the 2.0 Duo.

In addition, the thread is about examining what can be done within the context of the 'stock' xyz slicing software, and the 'stock' bed leveling and calibration. 

I'm well aware that many DIY's (for lack of a better description) use other slicers and mods to the physical Da Vinci Printers.  But that's not the idea behind this thread.  At some point If I can figure out how on this forum, I might update the thread title to 'Printing Quality of the 2.0 Duo, using stock software and AutoCal'  . . . but I might have to ask a moderator to do that.

I welcome all input.

just finished printing the same object on my 2.0 with expected results... i.e. print quality comparable to the item printed on the 1.0
so i would suggest that the results you are looking for may be centered on the amount of glue you use, and getting the right amount of glue may be the tricky part.
I can post pics if you want them.
I agree that your print seems to have more bands than mine... maybe a 'feature' of the mac XYZware is that it doesn't really do .1 mm. Do you use the windows XYZware?

Thanks, I think pictures are unnecessary, it's pretty clear from your previous pictures that my setup and yours has some unexplained differences.  I'm working on a modeling project tonight, so I'm going to do another attempt and use 4 layers of glue and a 'stock' gap (I don't know what else to call it) of around 150, with the bed leveling as flat as I can make it:

156, 154,159 'stock' gap: 154 which I think means the closest gap between the nozzle(?) and the bed(?), which is this case is the right front corner for a 2.0 Duo.

------------------------------

It printed well and I used my 'ball peen' hammer and plastic scraper' approach to removing it (it was stuck hard!)   

Results are . . . Good?  My ultimate goal is to figure out how to use the stock software and firmware and cal procedure to get results that are good enough to include as components in fixture systems that I design and build for my clients.  My clients are high tech companies in Silicon Valley (mostly), so they expect professional grade fixtures with professional grade components.

'Good' is ok for R&D work, so I'm going to continue - the '4 layer' glue technique, coupled with a 'stock' gap of 154 for the 2.0 Duo got me to good.  So my next print is to try the same glue technique, with a 'stock gap' of 200.

n2ri.... as has been pointed out to you in other threads, this is a daVinci forum. Not a solidoodle forum... your comments are not really helpful here.

yes and as I stated in my post. this problem is not just a davinci problem. its an extruding 3D printer problem using ABS especially.

and this is the same problem by same person I believe. so my posts are relevant and helpful in that I am sharing experiences by myself and others with this problem and a direction to search for a solution to it. 'take of the blinders' as a former boss used to tell people being bullheaded about not wanting to look at all possibilities. this forum is for all to share and help each other with solving problems. not competing on whos printer is better etc. just because a member uses 1 type/brand printer dont mean they are only exclusively familiar with the 1. many members have or had several brands/types as 3D printers are very similar in most aspects.

this is much worse on larger, taller objects like a vase for example. some will stretch until cured then break like they did inside my vacuum Cyclone separator inside making for very difficult cleanup. this is not just a type/brand of printer issue. any extruder type 3D printer can have this issue with some materials like ABS that are fussy about shrinkage with more than slight temp differences in layers.

I matched your printing specs re density. layer height and speed.
for calibration i just use the XYZ firmware's calibration function. I've never particularly concerned myself with the numbers and i'll go with anything the firmware accepts.
although I did discover that the 1.0 is happiest with numbers in the 200 range, but that the 2.0 has to be closer to 100 or the extruder bumps into the glass of the bed.

Interesting, your side photo seems to show fewer, thicker layers.  You also lay down 4 layers of glue, which is something I've done in the past but with that much glue, my parts seem to create a glue cavity, almost as if the heat of the ABS is softening the glue.

On my second run, which is halfway done, the skirt stayed glued down, but the left front corner has lifted up.  As I was writing that, the whole part caught on the nozzle and was ripped off the print bed.  LOL

To be continued.

How do you calibrate and do you know what the 'gap' is?

All in all I'm pleased with the print, and the only preparation I made was making sure I had a good coating of glue on the bed. After thoroughly cleaning the bed I applied the glue left to right using moderate pressure, then front to back, then angled from left front to right back, the from right front to left back then front to back again and finally left to right again.
The only problem I can see is the term "moderate pressure" because its a subjective term. What I aim for is a coating of glue on each pass that is as wide as the glue stick (I used the larger diameter uhu) and sufficient enough to completely coat the path, but not so much as to have it clump or deform the glue in the holder so that it starts bulging at the side.

Here is a copy of what I have in my spreadsheet (it doesn't copy to well) to capture what settings I used.  When I have some time, I'll update it with my notes about this first print.

Cal Test:         106    114    113    (106) 

The first three numbers are what are shown on the screen the forth one is what the printer (I think) uses for it's internal calibration.  It's obvious to me that this printer (the 2.0 duo    ) uses the lowest number, which I think is actually the smallest distance between something (the sensor?) and something on the print bed (the metal brackets?)

Or is it more sophisticated? does it represent the calculated gap between the nozzle and the glass?

If that is the case, then, for this print, it puts the nozzle at the same height as the layer.  - But I have no idea if this is what it means.

Print 1:   
        Quality        Excelent        Slicing    48 sec
    3D Density    Solid(90%)    Material    2.3 m
    Shells        Normal        Time            1:14
    Layer Height    0.1                Transmit    8 sec
    Speed        Standard        Start            8:48
    Raft                No                 End          10:02
    Support        No                 End

Rather than starting a new post, I'm going to add my next attempt to the attached file list of this post.  It's the same as before, but with a .1inch skirt that is .1inch away from the first part.

It's a thin triangle with a hole in the middle and some different surface geometries.  I ran it with a .1 layer height with the Cal value at 106,113,114 (106).  This is the short hand I'm using for cal values.  The print took about an hour and a half.  It's 2.54 thick (.1 inch) as a model and measures 2.49mm in the front middle as printed.

I set the model design up so that the corners of the triangle are approximately where the leveling posts are on the Da Vinci 2.0 Duo, which has two posts in front and the third post buried a little towards the back of the middle of the rectangular plate (two extruders mean a narrower build plate).

It obviously warped at each corner, with the worst warp at the two front corners.  Using my microscope, I took images of the corner edge on and the center of the edge that faces the camera.   

The following is a shot through the microscope looking at the middle of the edge facing front: 23 Layers, with the first layer thicker than the others, quality seems really good.

And here is a shot through the microscope looking at the left front corner:

BTW, To upload these pictures, I joined 'postimage.org' with a free account.
photo hosting

Here is a close up view at a slightly different angle

I'm out of time this morning, but wanted to get this up here, I'll return later and take a close look at the top surface quality and plan my next trial.

I don't have feeler gauges, but I am planning to print a part with dimensions that I might be able to use to confirm.

I plan to create a triangular part that is 2.5 mm high, whose corners are above the leveling studs.  I'll print this using a variety of settings and I'll record the values (including the 4th 'cal' value reported by the printer).  I've got a microscope, so I can count the layers. I'm hoping that this exercise will give me the insight about the printer I'm looking for. 

I know that the members of the open source community have all this figured out for the printers they've used, but I'm a beginner and this printer is . . . what it is.