Time based Mask and Power of Pixel

traliman · 2016-11-25 16:51:38

Hello, it will be very usefull to make a support of time based mask, not to lose the power the projector when working with slow polymers. B9Creater uses something similiar to this https://www.youtube.com/watch?v=07f7ZfnHCwg. Some parts of the table are cured more time than the others, depending on depth of the corresponding pixels in mask file. The Mask file for time based method file can be the same as usual.
Conversely, for some very reactive polymers it's necessary to reduce the power of light. And it can be done by using gray color instead of white(decreasing the depth of the pixel from 255 to 120 for example).
Sorry for english, not my native.

Shahin · 2016-11-25 18:17:36

Sorry, I am not sure if I understand correctly. You mean instead of dimming pixels we should cut time of cure partially, right?

traliman · 2016-11-25 19:10:31

Yes it is right.
For example mask file:

and
curing of layer looks li this:

Shahin · 2016-11-25 20:15:09

Why it will have different result than mask?
I mean there are lots of different projector types with different light outputs, maybe cure time and amount of light correlation is not linear but in the same time I have not heard any projector could not cure the resin because of weak output, usually more cure time will solve the issue.

traliman · 2016-11-26 12:24:11

I have not heard any projector could not cure the resin because of weak output, usually more cure time will solve the issue

Yes this is the problem when small weak projectors or lage build areas are used, light is weak and every lumen is important. For exapmle, i have problems myself with curing MJ WaxCast and Envisontec pic100 on large build area.

I mean there are lots of different projector types with different light outputs, maybe cure time and amount of light correlation is not linear

I think more correct to speak about the effect on curing than the amount of light and i think there is no exact linear correlation neither in time based method nor in usual method. It is very individual for every projector and every resin. In time based method you need to choose "light off" speed experimentally using a calibration patterns.
Like i said, b9creator uses only time based method:
https://www.youtube.com/watch?time_cont … rOfnV3sdRg
They dont have mask file, they have a lot of parameters, and a software generate mask automaticaly, using this parameters.
But i think it is good to have opportunity to choose method, that you want to use.

mandreas · 2016-11-26 16:22:19

https://youtu.be/usJGwwadHMk

put on the video of an old conversation, I think that extending the exposure time for parts with lower surface, could help a lot

Shahin · 2016-11-26 17:51:36

It required major rewrite. Do you know if it is patented or not?

mandreas · 2016-11-26 20:03:23

Shahin wrote:

It required major rewrite. Do you know if it is patented or not?

I think so,
maybe try on the site USPTO Patent, for 3D systems, to understand the margins of the patent...

Shahin · 2016-11-26 20:27:35

Andrea, I have give a try could not find it. Could you give it a try?

traliman · 2016-11-29 17:07:28

I am not sure, but i think, it is not patented. If B9 uses this freely, i think, anyone can do the same. And the simple mask method seems to be patented by envisiontec. https://www.google.com/patents/US8126580

mandreas · 2016-12-02 18:13:16

in fact I tried, but no results regarding the patents of that company,

it would be a just cause for the call to arms of some willing, skilled in the patent search, it would be great that Shahin could dedicate to this fabulous implementation, without fearing legal repercussions !

Last edited by mandreas (2016-12-02 18:15:04)

jcarletto27 · 2017-01-19 15:59:26

we may have an alternative to a time based mask. Shahin, is it possible to have models loaded in with supports, and then just the support structures? this way we get stronger supports while keeping detail of the model.
look at it like this timeline:

Layer 1 - exposes both support structure and model
Layer 2 - exposes just support structure
cycle z axis
Layer 3 - exposes both support structure and model
Layer 4 - exposes just support structure
cycle z axis

as far as I know, no system I've ever heard of implements that for SLA, however it is pretty similar to how FDM printers handle multiple materials.

It's less dynamic than going completely on surface area, but at least it ensures sturdy supports.

name the slices with prefix to differentiate them ie model-1.png, support-1.png, or seperate them into 2 folders, create some logic that only requires you to display support slices if they exist.

Shahin · 2017-01-19 18:21:40

jcarletto27 wrote:

we may have an alternative to a time based mask. Shahin, is it possible to have models loaded in with supports, and then just the support structures? this way we get stronger supports while keeping detail of the model.
look at it like this timeline:
Layer 1 - exposes both support structure and model
Layer 2 - exposes just support structure
cycle z axis
Layer 3 - exposes both support structure and model
Layer 4 - exposes just support structure
cycle z axis
as far as I know, no system I've ever heard of implements that for SLA, however it is pretty similar to how FDM printers handle multiple materials.
It's less dynamic than going completely on surface area, but at least it ensures sturdy supports.
name the slices with prefix to differentiate them ie model-1.png, support-1.png, or seperate them into 2 folders, create some logic that only requires you to display support slices if they exist.

Awesome idea. But preparing files for such system would not be easy for average users. If people who uses newly added support generator inside nanodlp, we could fully automate such process.
Also I am not sure how many people have issue with sturdiness of the supports.

jcarletto27 · 2017-01-19 18:30:54

fair point Shahin.

Though, it is less about sturdy structure and more about sturdy connection to the model. With this setup it might be possible to use smaller tip diameter for supports, this should mean cleaner models.

nanoDLP

#1 2016-11-25 16:51:38

Time based Mask and Power of Pixel

#2 2016-11-25 18:17:36

Re: Time based Mask and Power of Pixel

#3 2016-11-25 19:10:31

Re: Time based Mask and Power of Pixel

#4 2016-11-25 20:15:09

Re: Time based Mask and Power of Pixel

#5 2016-11-26 12:24:11

Re: Time based Mask and Power of Pixel

#6 2016-11-26 16:22:19

Re: Time based Mask and Power of Pixel

#7 2016-11-26 17:51:36

Re: Time based Mask and Power of Pixel

#8 2016-11-26 20:03:23

Re: Time based Mask and Power of Pixel

#9 2016-11-26 20:27:35

Re: Time based Mask and Power of Pixel

#10 2016-11-29 17:07:28

Re: Time based Mask and Power of Pixel

#11 2016-12-02 18:13:16

Re: Time based Mask and Power of Pixel

#12 2017-01-19 15:59:26

Re: Time based Mask and Power of Pixel

#13 2017-01-19 18:21:40

Re: Time based Mask and Power of Pixel

#14 2017-01-19 18:30:54

Re: Time based Mask and Power of Pixel

Board footer