Partial Answer I am trying to get the sizing correct, suggestions please.

Been playing around for a couple of weeks now and I am having problems with prints sizes. First a bit of background on what I have already done. Got the Robo, printed a repair Y-Axis belt tentioner, repaired Robo. Then I could start printing some real stuff. My first problem that I tried solving was that the Robo R1 default firmware would way under the amount of filament that Thomas Sanladerer explains in this video of extruder calibration. So I set about measuring and running multiple tests of extruding the proper amount of filament. The default E-step Robo uses is 723.38 (seems like an odd number since it simply cannot be that precise), anyway by measurement I came up with 804 E-steps. I tested this number a couple of different ways (multiple short, one long, a combination of long and short extrusions) and was satisfied the number was correct. Now print a part and way over-extruded. So I went back to the default setting and started working on other aspects, thin walls too thick, etc. Played with temperatures, extrusion widths, layer heights, filament diameter, every aspect I could, but all within reason. By the way, I am using Repetier-Host with Slic3r. I tried MatterControl (behaves like a beta product) and Cura and was surprised that the interface Repetier-Host provided just made more sense to me that the other two, I get better results from Slic3r than I do from Cura, I have not tried Skeinforge.

With the current setup all my parts look good, no gaps, good adhesion between layers and if size doesn't matter, I would call them successful prints. I am printing primarily in PLA (not the starter PLA) and PET-G. I have not tried any other types of filaments. So here is the issue:

Every item I print is about 4-5% undersized in X/Y and 1% in Z.

In every axis, no matter the material. A cube is a cube but a 20 mm cube measures measures just over 19 mm on every side, including height, although height is closer to actual than the other measurements. So I know the number of steps is correct in the firmware. I know how to update and modify it if needed. I just can't get anything to print out full sized. So I am doing more experiments. Tried scaling the parts in Repetier-Host, that didn't work as now I couldn't get a good looking print. Went back and redid the extruder calibration and measured the exact same 804 I tried before. So I modified the firmware and now my code looks like this:

Code:

// default settings

#define DEFAULT_AXIS_STEPS_PER_UNIT   {80,80,2560,804}  // 80,80,2560,723.38 default steps per unit for RoBo 3D R1
#define DEFAULT_MAX_FEEDRATE          {500, 500, 5, 25}    // (mm/sec)
#define DEFAULT_MAX_ACCELERATION      {9000,9000,100,10000}    // X, Y, Z, E maximum start speed for accelerated moves. E default values are good for Skeinforge 40+, for older versions raise them a lot.

#define DEFAULT_ACCELERATION          1300    // X, Y, Z and E max acceleration in mm/s^2 for printing moves  /robo
#define DEFAULT_RETRACT_ACCELERATION  3000   // X, Y, Z and E max acceleration in mm/s^2 for retracts

Other than still needing to calibrate Solid Infill (it way over-extrudes when an object is solid plastic). I have very good looking prints and no gaps in top or bottom layers, but they still are too small. I am willing to try any suggestions. I am almost to the point of replacing Slic3r 1.1.7 Stable with 1.2.7 Experimental since it added a New XY Size Compensation option for handling dimensional tolerance feature.

So it would be wonderful if the experts in this community could put on their thinking caps and offer a few suggestion, even if they are out-of-the-box, to help me solve this issue.

 

@Mike Kelly , I have seen this and did not do any of those old fixes. Did you see the code I posted, nothing is changed from the defaults except E-Steps, which Thomas advocates in his calibration video. He also suggests the same thing I am suggesting moving to Slic3r Experimental 1.2.7, but in it's current implementation it is a fixed amount of compensation rather than a percentage.

 

So as a workaround, I am kludging something together, Slic3r 1.2.7 with XY compensation set to -0.15 mm and scaling the object in the X and Y only @ 102%. This makes holes the correct size and the print the correct overall size, or at least within a few hundredths of a millimeter.

 

R&R ???

1. Rest and Recuperation?
2. Rescue and Resuscitation?
3. Revise and Resubmit?
4. Read and Review?
5. Repeatability and Reproducibility?
6. Roles and Responsibilities?
7. Redbull and Ritalin?
   
8. Rock and Roll?

 

I like #8

#7 is probably more appropriate for getting started with the printer.

 

Number 5 is alive

It is consistently off by 2.5% after repeated tests with multiple types of filament.

Number 7 is a College Cocktail.