Scaling Problem

MSURunner · Post by **MSURunner** » Mon Mar 31, 2014 12:34 pm

Okay, I have an issue with the dimensions of prints that has just surfaced I believe (Or it has just been called to my attention). My prints are approximately 2-3% too small, which didn't make a noticeable variation on smaller calibration prints. However, now that I'm printing things at a larger scale, it's definitely a problem. What variable is the correction for such? I moved the arm length from 269 to 271 as I saw on the forum, but it had no change. Running .91 and EEPROM values are fairly close all around to "stock" other than my Horizontal Radius is 127.9...

Nylocke · Post by **Nylocke** » Mon Mar 31, 2014 12:40 pm

Do you have EEPROM enabled? Is the arm length an EEPROM variable? That should adjust the scaling... Last time I calibrated a delta, I had to decrease the arm length to increase part size I think :/ maybe I'm just crazy, that was over 6 months ago

MSURunner · Post by **MSURunner** » Mon Mar 31, 2014 12:45 pm

Yes, and I tried, though it didn't seem to make any difference. Maybe I'll try a much larger scale factor

joecnc2006 · Post by **joecnc2006** » Mon Mar 31, 2014 12:46 pm

I would measure all the points for delta arm length, the effector offset, Carriage offset and delta radius and make the changes in the Firmware (make sure to write the defaults down), I did this when I built the Magnetic arm mod and the prints for a 25mm cube came out to be 25.01mm-25.03mm. of course you do have plastic cooling shrinkage to take into consideration but it is very small.

MSURunner · Post by **MSURunner** » Mon Mar 31, 2014 12:53 pm

joecnc2006 wrote:I would measure all the points for delta arm length, the effector offset, Carriage offset and delta radius and make the changes in the Firmware (make sure to write the defaults down), I did this when I built the Magnetic arm mod and the prints for a 25mm cube came out to be 25.01mm-25.03mm. of course you do have plastic cooling shrinkage to take into consideration but it is very small.

I had done that at one point, and maybe I didn't check the values as carefully as I had previous thought... Breaking out the calipers now

MSURunner · Post by **MSURunner** » Mon Mar 31, 2014 1:54 pm

Carriage offset was off slightly, with the difference being the corresponding difference in what I had in the horizontal radius so I think those two almost cancelled out. However, it looks like my diagonal rod was set much lower in the Configuration.h file than in the EEPROM. Since the manipulation of the diagonal rod in the EEPROM didn't make any change and bringing the Configuration.h value to what it is the EEPROM helped out, I'm guessing there is something in the firmware where the EEPROM value is not being recognized... Anyone concur?

joecnc2006 · Post by **joecnc2006** » Mon Mar 31, 2014 2:24 pm

MSURunner wrote:Carriage offset was off slightly, with the difference being the corresponding difference in what I had in the horizontal radius so I think those two almost cancelled out. However, it looks like my diagonal rod was set much lower in the Configuration.h file than in the EEPROM. Since the manipulation of the diagonal rod in the EEPROM didn't make any change and bringing the Configuration.h value to what it is the EEPROM helped out, I'm guessing there is something in the firmware where the EEPROM value is not being recognized... Anyone concur?

what were your difference in settings? Did you do a test print to check it out? Maybe a pen plot would be more accurate to measure.

Nylocke · Post by **Nylocke** » Mon Mar 31, 2014 2:33 pm

Cant you just change the EEPROM value of the rod length? I'm pretty sure thats what it was meant for...

MSURunner · Post by **MSURunner** » Mon Mar 31, 2014 2:46 pm

joecnc2006 wrote:
MSURunner wrote:Carriage offset was off slightly, with the difference being the corresponding difference in what I had in the horizontal radius so I think those two almost cancelled out. However, it looks like my diagonal rod was set much lower in the Configuration.h file than in the EEPROM. Since the manipulation of the diagonal rod in the EEPROM didn't make any change and bringing the Configuration.h value to what it is the EEPROM helped out, I'm guessing there is something in the firmware where the EEPROM value is not being recognized... Anyone concur?
what were your difference in settings? Did you do a test print to check it out? Maybe a pen plot would be more accurate to measure.

Test print is a 100x100 mm square, going larger to get better precision than the 20/25 mm cubes.
Configuration.h was at 267 for some reason and I left it alone, figuring it would accept the EEPROM value. Adjusting the EEPROM didn't seem to make changes. I moved that to the 269 that is "normal" and what the tricklaser arms should be and it made about a mm/1.5 mm difference.

MSURunner · Post by **MSURunner** » Mon Mar 31, 2014 5:09 pm

Okay so here's the problem I'm currently having. The part comes into Cura with dimensions (64.6 x 121 x 13.5 mm), yet prints with dimensions (64.1 x 118.4 x 13.25 mm or 99.23% x 97.85% x 98.14%). Thoughts? What am I missing here? The Z-height isn't as big of a concern for me as I'm slightly smooshing it into the glass so I should be off a little bit, but the fact X and Y direction don't correlate?

Nylocke · Post by **Nylocke** » Mon Mar 31, 2014 8:54 pm

What material are you printing, and you should try printing a few to get more data, 3 times is a trend

MSURunner · Post by **MSURunner** » Tue Apr 01, 2014 3:58 pm

Okay, so here's what's going on. I've got the Y-direction correct, but the X-direction is now about 101% of what is should be... There's got to be a solution besides split the difference and call it good. My new working theory: differences in the delta radius? Yes/No?

joecnc2006 · Post by **joecnc2006** » Tue Apr 01, 2014 4:05 pm

Are your towers parallel to the bed when using a square, and is the center of the nozzle the same distance to all three towers? May be an alignment issue, just throwing out some more things to check.

MSURunner · Post by **MSURunner** » Tue Apr 01, 2014 4:36 pm

joecnc2006 wrote:Are your towers parallel to the bed when using a square, and is the center of the nozzle the same distance to all three towers? May be an alignment issue, just throwing out some more things to check.

I assume you meant perpendicular and they were pretty damn close when I went through the re-assembly. There is a SLIGHT difference in the delta radius of the Z-tower (about .5 mm) but I have also accounted for that in the EEPROM...

MSURunner · Post by **MSURunner** » Thu Apr 17, 2014 1:23 pm

Any ideas on the uneven X vs. Y? I'm still at a loss and have checked just about everything I possibly can...

lordbinky · Post by **lordbinky** » Tue May 06, 2014 3:43 pm

My X/Y variance was corrected when I got my Z=0 center (0,0,0) equidistant from all the towers. Are you still having the X/Y issue? I believe it was your post led me to correct that in the first place.

Welcome to the SeeMeCNC Community

Scaling Problem

Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem

Re: Scaling Problem