LCD Based 3D Printer - Introduction

[dpizzle]

Yeah, you're probably right. I was thinking Endlighten solves the evenly distributed light without needing to use a diffuser part of it, but we'd probably lose some power from it. I might give it a shot just to see, since it's reasonably cheap.

[Barill]

Yeah, I'm curious to know the result.

I think the best solution is the LED array under the LCD.

xxxxxxxxx VAT xxxxxxxxxxxxxx
xxxxxxxxxLCDxxxxxxxxxxxxxxx

xxxxxxxxLED_ARRAYxxxxxxxxx


But I fear that the LED ARRAY creates a luminous flow in all directions which could be problematic in terms of accuracy.
So I devised a polarizing film between the LED ARRAY and the LCD to direct the flow in one direction only. But I read a polarizing film lost half the light output!

[Phife]

Hey Guys, sorry for the late reply.

IMO there is no point in using color LCD's. The Green and Red color filters block all the light from the array. So 2/3's of your pixel will be black. This would probably leave holes in your print, or at least have some sort of screen door effect. That being said, I havent tested a color LCD with my new array, so I really dont know what the effect would be. You will need more light power to make up for the 66% loss in the color filters.

Using a TIR reflective backlight ive tried, you need a lot more power than you can put into the edges. Heat becomes an issue as well. And the light is scattered not collimated.

LCD Polarizers lose 90% of the light you put through, so optics and power are very important. Using bare emitters does not work, you need the LED lenses to direct the light through the LCD.

[Barill]

Result of the very first archaic test, just "one layer"...

It sounds promising.
You can see a allipse, a square and a sort of "+" sign on the left, under the ellipse

[dpizzle]

Very cool Barill! Possible to share any details on your set up?
Monochrome or color LCD? Are you using LEDs, if so what wavelength? How long does it take to cure?

[Barill]

Hi dpizzle,

Please note, this is the very very first test, very far from the Phiffe status...

But...this is a color TFT LCD ! from my old stock (like the ones you can find to connect with Arduino shield)

For the test I made a 9 leds 3W array, 405 nm, with the built lens.

I did not measure time to cure for the moment, but I think about 2-3 seconds.

As I said, this is the very first test. I did not work at all about improve wavelength, lenses, etc. But I think that by improving these points it will be possible to make something work.

[Phife]

Nice results Barill,

I had similar with my initial color LCD tests, but I was using a 400w bulb and it took about 15-20 seconds for something like that.

I should go back and test color LCD with LED arrays.. Do you notice any banding, missing spaces in the print?

Even with my monochrome LCD I can see a slight banding from the space between the pixels, looks like a grid on my print bottom. Cant see it on the surfaces though, just the bottom of the layer.

How thick is the cured resin? is it fully cured or is it still soft?

[Barill]

I can see the pixels shapes but all spaces seems hard cured.

For the moment I only made "one layer". I don't have yet Z axis, the aim was to see if it was possible to get something with color display.

[Phife]

Youre test made me wonder about the color LCD's again. I just picked up a 10" 1920x1200 LCD panel and controller online. Pretty cheap, i'll do some tests with my LED arrays with it and see how it turns out.

Im also switching to top down, bottom up with this large build area just seems too hard to get reliable prints.

[Barill]

a other test.
Can you guess what it is wrote ? Not easy !
Otherwise, the squre is well done.

I have to work on exposure time, I think to much here, about 3 seconds.