nanoDLP

Thank you for wanting to to evaluate this function,
you're right it would be more technically speaking an hollowed, unless there is a way, with the dimming algorithm, which instructs it to start, for example, (xx) mm from the tangent of the outer edge,
(Just an 'idea, from a totally ignorant person like me about  programming)

How will you remove the resin out of this structure after print ?

the solution to this problem could be the following:
instruct the program that executes the dimming / hollowed, but from point A to B runs the Honeycomb...

Agreed, even a 2D honeycomb or hatching pattern as used in the FDM slicers would be greatly helpful.  In those slicers they give you an option for how many solid bottom and solid top layers to have, the middle layers are hollowed with a patterned infill

As you mentioned if the structure get closed, only for tall objects you could save some resins as lots of resins will get blocked inside.

But the main issue is the performance and memory issue. Lets say you need wall thick as 5 layers. So for each layer we need data belong for next 5 layers and 5 previous layers.
Each layer we have 1920x1080 pixels each pixel is equal to 4 bytes so we will require 88MB memory. For 4K display it will be 352MB.

There is a workaround but it will need lots of time to develop. Will try to put some time into it after finishing new support generator.

Hexagonal infill is available on the beta version.

every time I have tried to check out the infill feature my Pi crashes...:(..

web interface stops responding...some time ssh too and I have to power cycle...
when it comes back up slicing resumes followed by a crash, so I have to power cycle and quickly delete the plate.

looks like a another great feature....would love to try it out...for now given the above I think I will wait for the next release.

here is a link to see the dashboard right before the Pi crash...as you can see it does do some layers, it should have been into the 'Infil' section at this point I think.
This was a small model (same one shown in my posting about the missing support types)
I let it run all night just to be sure it was actually crashed and not just super super busy...but this morning I can't connect so I think it is a crash.

I also previously tried a super simple model (just a small cylinder 20mm diameter 10mm high but it crashed too.

https://www.dropbox.com/s/plx78l8asqwff … M.jpg?dl=0

you can see CPU climbing am memory usage too but they are flattening out I think...this is hard to capture because at some point it just dies....
I guess I could wait until I get the disconnection message or stop getting log entries but it looks pretty much like this.

By the way  this was after a fresh reboot to update to the latest build.

I just tried to ssh to the pi and while this is not always the case I got in....so even though nanoDLP seems to have died the system
is still up.... CPU is basically idle, memory is not locked up and shows a typical usage...drive space is only 15% used with 12G available.
I will try to copy out the logs.....

Last edited by macdarren (2017-05-18 16:52:56)

I just sent you an email with the log I was able to "scp" from the Pi, not sure how to get the debug file once nanodlp has failed (I can't get into the web interface)
Also just to be sure I send matching items is there a way to get the STL file from the Pi...the plate file must be on the pi somewhere?

If you still have the SSH connection alive , you can navigate to "/printer/public/plates/"(no really sure about the exact path) and delete the problematic plate.  restart your Pi and should be fine .

An update about better reinforcement structure for use in 3d printing, the spaceclaim program produces beautiful structures and shell, my favorite is this image, solves the problem of resin drain and creates an extraordinary structural strength