I use Simplify3D and I want the print times it gives me to be accurate. Right now they are about 80% too fast. My prints end up taking 5 hours even though S3D shows 4 hours or even less. What are the contributing factors here and how do I correct this for accurate estimations?
switch on accelerated printing in your firmware - if you've got it.
I think s3d assumes this is on.
Quote:
Originally Posted by curious aardvark
I have no idea what you are talking about. Where in S3D would this be? Where in the firmware would this be? How does this change what S3D uses to predict a print time vs. what it actually comes out to be?
My prints take 1.5x as long as are shown to me.
DBFIU, what CA is trying to get at is that there's a lot about your printer that the estimator doesn't know, and thus can't factor into the estimate. It won't factor hot end and/or bed heating, homing, or bed leveling as obvious starters. It also assumes all axis moves occur with instantaneous acceleration - probably the biggest gotcha for those of with i3v printers since Colin suggests fairly slow acceleration rates for them. You may not have an i3v, but it could still be a big factor. I personally believe that there are periods where Marlin just briefly pauses printing while it processes the next handful of moves - those pauses won't be factored either. With time, you just sort of get used to mentally factoring the estimate to something more believable.
Ohh I see. The computer doesnt know what my firmware has programmed for accelerations and dwell times and ABL and stuff.
How can I tell S3D all these things so it can spit out more accurate numbers? Is this program not evolved enough to the point to be able to do these predictions with the correct inputs?
None of the slicers are good at estimates. Cura has a percentage factor you can adjust, but the bit I played with it I wasn't convinced it did much. Again, you just sort of get used to mentally adjusting the slicer estimate. This likely won't help you feel better, but I did notice that when I migrated away from Marlin and MEGA2560 to Smoothieboard running smoothieware, the S3D estimates are a lot more accurate. I'm maybe within 15% now. Nearly all the speeds and other settings are the same as before - the one big difference is acceleration. I've left smoothieware at the 3000 m/Sec per second default instead of the much slower values I had in Marlin. I also no longer observe those very brief pauses in the printing either.Quote:
Originally Posted by DBFIU
I have no idea what smoothieware and smoothieboard are. All my machines run arduino 2560 marlin and I cant switch over at this time.
They're just a different electronics and firmware suite. I wasn't trying to imply you should switch to them. Just providing the difference I noticed as a way to emphasize that there are delays in firmware that can't be properly estimated by the slicer...Quote:
Originally Posted by DBFIU
I see, thank you!Quote:
Originally Posted by printbus