Solved Z height incorrect on calibration cube

I just received my R1 and after loading the latest firmware and getting the latest matter control software, I attempted printing the calibration cube. Using the Robo blue PLA, with hair spray on the bed, the cube printed reasonably well with just a minor amount of separation on the bottom layers. I used the standard setting for PLA in matter control and used high quality. My cube measurements at the most favorable points are 19.90 x 19.92 x 10.48. The x and y axis seem reasonable (off by .5%) and are consistent along the whole cube. However the z axis is off by 4.8% at its best and 6.7% (10.67mm) at its worst. I checked the configuration file and the z steps per unit is 2560 which I believe is the correct number for the 8mm z rod. Probably a third to half the error is a results of the separation. I don't know what caused the rest. Two questions; what causes the minor separation of layers at the bottom of the print? Is it cooling to fast before the next layer is added? And is there an adjustment to correct the z height. When I start making the items for which I purchased the R1 for, I will need accuracy of 1/2 % or better in all three axis.

 

Are you sure you have the 8mm firmware installed?

I had the same problem although 10.48 is crazy off.
Mine was like 19.9 ish - 20.,05 , 19.9ish - 20.05, and 17 or 18 on Z. I put the 8mm firmware on and it was perfect.

Mine had the wrong firmware.. I didn't look at the z-steps number.. I just got mine a couple weeks ago from amazon.

 

Yes it was the 8mm firmware. The file name was Auto_Level_7_17_8mm.zip I thought matter control indicated my cube should be 20 x 20 x 10. I am not sure why some are 20 x 20 x 20. Some on the forum have also indicated that their cubes were 20 x 20 x 10, not sure why there is a difference. Maybe because I downloaded off the website instead of using the software off the memory card. I also downloaded the firmware off the website and uploaded into the R1 to make sure I had the latest stuff.

 

Did you confirm your threaded rods are actually 8mm? They could of by some freaky thing put 5/16. just a thought.

 

I will check with my calipers when I get home tonight. I assume the measurement is the diameter of the rod.

 

Have you lubricated your Z threaded rods to make sure they're not sticking?

 

The thread count is 20-21 so it must be the 8mm. No I have not lubricated, I wanted to keep the machine as clean as possible in case something didn't work and I needed to return it. Since it is mostly working I will try lubrication

 

Oh i didn't realize you were printing a 20x20x10 model.

I made my own 20x20x20 cube model to print.

.5 mm off isn't horrible but it should be better than that and may be worse if you go larger What layer height did you use?

PS I think the machine comes with like no lube. just put a little dab with your finger on the Z screws. it will quiet it down and maybe help. I use SuperLube from amazon, but anything similar should work.

 

layer height was .1mm I just got done lubricating but too late to start a print. Will try again tomorrow night.

 

If you want more accuracy it's a good idea to do estep calibration:

 

I'm having a similar issue. The cube is coming out 20x20x9.2mm. This is using standard settings out of the box with sample PLA roll measuring 1.9mm in diameter. (I have set this value in the software but has little effect) I have 8mm rods (I think, 0.0025" difference between it and 5/16" - I count 20 to 21 threads per inch though) and have tried slowing down the speed and increasing the extrusion multiplier up to 1.1. The layer height is set to 0.2mm. Overall quality is great with one very tiny hole on top. The 100mm extrusion test pushed out 95 to 96mm of PLA. Please help me resolve this issue.

 

I have resolved my problem. The first thing I did was calibrate the extruder, as per Mike Kelly's suggestion. I then started printing 10x10x10 cubes and adjusting the z factor in the the configuration.h file. I don't know if this is the correct thing to do, but it works. The number I am now using is 2475 rather than the original 2560. I now have a z axis error of no more than .04mm on a 29.5 mm side. The x and y axis is off by more, but still reasonable. Does anyone know if the x and y axis factors in the configuration.h file can be a decimal number (i.e. 79.2 versus 80)? If so I can try fine tuning these axis' also.

 

I was surprised by the correction, as I felt that the 2560 was a calculated number based on the z-rod threads per inch. I originally argued with my son against changing this number. However it seems to be working with the same error percentage (a little over 1%) at 10, 20 and 30mm. So hopefully it will hold with the 80mm z axis items that I am designing. Could there be some inaccuracies in the manufacturing of the z-rods?

 

My calipers are pretty cheap, but that is what it measures. Probably + or - 0.02mm. I was hesitant about changing those variables. Thanks John for the info. I just did a software update and if that doesn't work, I'll try adjusting that number around 2500 or so and see what comes out. I'm trying to print a sphere which is not coming out round, thus the cube test.

ALRIGHT!! New firmware ROBO3DR1V3 worked. It is now slightly over sized, but can be tweaked into place form here. tomorrow.