I am at it once more, planing and building an LCD based SLA printer. As before, the struggle is wavelength and I believe I have valuable information regarding wavelength and resins, so lets get to it. I will past info from a PDF I wrote a while ago so that what I know and/or have in terms of resources is as transparent as possible since I want to contribute to this build. Yes, I am planing on using the very same LG Retina display.

If you know about the Morpheus 3D printer, they use a monochrome aka Greyscale display. This panel type will pass UV. I believe the panel they use is the Innolux / CMO R300M1-L01.

Here is a link with info regarding the panel: CMO R300M1-L01 Overview - Panelook.com

PPI of active area: 160.24 ppi, so much higher than the 96.2 ppi or 96.2 micron on the XY plane one achieve with retina display.

Greyscale.
As you might know, one big contender in this area is Barco and Joshua which both produce medical displays and neither of them are cheap. It also turns out that neither of them manufacture the display – this is very common in the monitor business no matter greyscale or RGB.

So, with the information available from Barco model; Coronis 5MP LED (MDCG-5221).

Grayscale medical display with 2560 x 2048 resolution.

Screen technology: a-si TFT active matrix dual domain IPS
Active screen size (diagonal): 540 mm (21.3”)
Active screen size (H x V): 422.4 mm x 337.9 mm (16.5 x 13.3“)
Maximum luminance: 1200 cd/m² typical
DICOM calibrated luminance: 600 cd/m²
Contrast ratio: 1200:1 typical

Greyscale is interesting because No RGB pixels, this allow us to essentially pass any color we want. If the resin cures best at red, use red. Best at blue, use blue etc. One “problem” that I see with greyscale is the limit in XY resolution since these displays is generally not made to be repurposed for SLA or 4K media content, there is no real reason for a higher resolution than 2560 × 2048. Another issue is that UV curing is high energy content. This is noticeable when you read about machines that have VAT problems with resin that bonds with the peel material PDMS and that the PDMS need to be replaced after perhaps 80-100 prints. This is not good.

There exist photopolymers that cure using ranges within the visible spectrum. The advantage is ofc that we entirely bypass the passage of UV (405nm) of LCD displays and this should also work for DLP (since DLP is not out of the question). So before I dig deeper into LCD based solutions, I want to check what is available regarding daylight curing photopolymers which is very important.

Bucktown Polymers:
ZVE200-V470 - UV to Visible blue light cured resin.

ZVE200-V470 is a single component zero volatile organic compound (VOC) resin and does not require any weighing or mixing. It cures rapidly when exposed to ultraviolet light and visible blue light up to 500nm and offers nearly unlimited working time before exposure. This is a tough, hard and low shrink resin. Sensitive enough to cure by the blue light from standard unmodified Pico and Pocket sized LED projectors.

Low Viscosity, Low odour, Zero VOC Epoxy, High Performance and tough Corrosion resistant, Cures by UV up to visible blue led, lamp or laser <500nm, Compatible with SLA, laser, LCD and DLP 3D printers.

Custom colors available by special order.

Bucktown Polymers - ZVE200-V470

MADESOLID:
MADESOLID have a range of resins: CastSolid, Vorex and MS Resin. The common factor for MADESOLID is that they are compatible on variety of SLA/DLP printers, including the Form1+ and B9 Creator. The common factor for DLP is mercury vapor lamps such as used in the NEC NP41, since this model use mercury vapor lamp, MADESOLID wrote this on their webpage regarding the LittleRP DLP printer.

Here is some technical info regarding LittleRP and MADESOLID: Technical Support | LittleRP.com

At this stage of me collecting info, I finally found a PDF with info - go to link and page 4.
http://www2.units.it/ramponi/teaching/DIP/materiale/dip04_Display.pdf

Here we can see that the peek region for blue wavelength is roughly 460 nm, meaning this is the wavelength of desire or in other words: light at 460 nm is transparent to the blue led in the LCD.

One interesting comment in the Blue light cured resin thread over at buildyourownsla.com

Re: Blue light cured resin
Postby Cation » Thu Mar 19, 2015 11:56 am

The Bucktown Polymers - ZVE200-V470 is sensitive to over 500nm. That is why it works with unmodified Pico, LED and LCD projectors. Many of the smaller Pico projectors use LED's vs a wide spectrum lamp. The LED's used for Blue are centered around 460nm. The other resins tend to use BAPO or TPO as photo initiators that are not sensitive above ~430nm.

@Phife: that’s why they won't cure at 470nm. Stronger or more power at 470nm won't cure resins with photo initiators that are only sensitive below 430nm.

So what we know at this point is that the photoinitiator used in the resin is the key for maximum performance regarding daylight curing resins and 460 nm is or must match in order for the LCD based SLA to work in an optimum situation.

Spot-A Materials:
Our Photoactive Resins are liquid resins that harden (cure) through exposure to a given window of light, visible or UV. These resins cure quickly to deliver polymers with a wide range of available properties. Our Photoactive Resins are liquid resins that harden (cure) through exposure to a given window of light, visible or UV. These resins cure quickly to deliver polymers with a wide range of available properties.

Photocentric article.
Photocentric’s Entirely Different Approach To Resin 3D Printing

Photocentric’s Entirely Different Approach To Resin 3D Printing You’d think a manufacturer of liquid photo-polymer resins would know much about 3D printing with resins, and this company certainly does.

Photocentric is a 13-year old UK-based company that apparently is one of only five companies on this planet that manufacture liquid photo-polymer resins. They produce around 500 tons of such resins each year, much of which is resold by others, including 3D printer manufacturers in their proprietary cartridges. Currently they employ eight researchers dedicated to 3D print applications, with more researchers being added frequently.

The goal of Photocentric seems to be an increase in sales of resin, as their success depends on increased use of resin. They’re now taking some bold steps to do so: their big move is to market a pair of demonstration resin-based 3D printers that involve a very different use of resin technology.

Both units employ a unique approach to resin 3D printing that could prove very interesting for not only Photocentric, but also other makers of 3D printing gear, since the company intends on open-sourcing the approach.

Here’s how it works: a standard LCD panel is placed under the resin tank to shine pixels of plain light - and this is the important part - not UV, into the resin. The unmodified light source uses very low energy to solidify the resin.

But what of the sticking-to-the-bottom-of-the-tank problem that occurs in almost every other 3D printer that uses the bottom-up approach? Photocentric explained that their system’s very low energy is very different from others who use “far too much energy” when blasting UV at the resin. For them, this is what fuses the solidified resin to the tank. In Photocentric’s approach, they use very little energy so the resin solidifies without fusing to the PFA tank. Thus, these inexpensive resin 3D printers should operate very fast, similar to the Carbon3D method. They call the method “Daylight Polymer Printing”, or “DPP”. Yes, they have a patent pending on this approach.

There are other advantages to DPP as well. Since it employs only normal light, there is no danger to eyes as you might find in a laser-powered device. The LCD panel is so close to the tank that no focusing or optical elements are required. And it’s very inexpensive, too, as they’re using off-the-shelf components. The energy used is so low they say it cannot go beyond the cured area, implying good printing accuracy.

That seems to be the case, as you can see with the very good quality sample prints shown here.

The two “LiquidCrystal” 3D printers (The LiquidCrystal and LiquidCrystal Mini) differ only in build volume size, and are priced at the ridiculously low prices of USD$599 and USD$1,199. The printers are mechanically very simple and should prove quite reliable, and are available for pre-order now.

If you want to wait for three months, they say they may have a 17” (431 x 431mm) version available, no doubt using a standard 17” LCD panel to drive it. They’re even talking about a massive 40” version.

Q: What is the difference between daylight and UV 3d resin?

A: Daylight resin 460nm is for use with the Liquid Crystal and LCD range of printers only. UV resin is for use with printers that use the UV spectrum – DLP and laser based printers will require UV resin.

Q: Why do I need VAT film? – Probably PDMS

A: The film is the base of the VAT and needs replacing from time to time if it gets damaged. You should be printing over 2kg of resin before you need to change the film.

Q: Does Daylight resin work with any LCD screen?

A: Yes, depending on the screen brightness the cure speed will change.

- Once again 460nm is confirmed. Hope this helps and good luck with the build. I will follow the progress.

 

This is a fair point, and I need to do a bit more research on what intensity LEDs one can get that are specifically in the 460nm spectrum. if those LEDs are not high power enough, even with the lost efficiency you might end up with a higher overall intensity. But I certainly have not done enough research to know if that is the case.

 

A question for evilc66: Why do you/we need to remove any polarizing film ?

 

I'm back .. he he and I have a few things to share.

First. Found a controller board for the 9.7" retina display, only $49 and is made by TinySine.

Link: Retina Display Controller Board - DE2KTOP Air

You need a HDMI -> mini DP if you are using the Raspberry Pi. Which brings me to one of the issues of Raspberry Pi, its native resolution is only 1920 x 1080p. Some have managed to re-code the Pi to produce 4K video at 15Hz. But there are better solutions than the Pi - I'll get to that shortly.

But first, I want to address the physical build area of the SLA.

SLA printer on steroids.
The 9.7" retina panel presumable is chosen due to the high PPI rating which in turn give us a low XY resolution of 96.2 microns. Smaller is better. FDM printers usually produce 200-300 microns on the XY plane while Z can be as low as 25 microns.

If we look at the Form 1 and 2, they achieve 155 and 140 microns on the XY plane. Build volume Form 1 (XYZ); 125 x 125 x 165 mm or 4.9 x 4.9 x 6.5 inch. Form 2 (XYZ): 145 x 145 x 175 mm or 5.7 x 5.7 x 6.9 inch.
They cost $2688 and $4437

Envisiontec Ultra SP3 have a resolution of (XY) 50-100 microns. Build volume: 266 x 175 x 193 mm or 10.5 x 6.9 x 7.6 inch. This one costs: $20 000 or more.

The 9.7" retina have a XY build volume of: 197 x 147.8 mm and any Z your heart desire. Yes, this is larger than the Form 1 and 2, but smaller than the Envisiontec.

While I was searching for a 9.7" controller board and found the TinySine version, I stumbled upon IPad Pro 12.9" thinking wow, this is better. I would get a XY build volume of: 262 x 196 mm or 10.32 x 7.74 inch and that is actually almost spot on the Envisiontec build volume - keeping in mind that the Z is not yet fixed.

However, the Pro 12.9" panel is expensive, coming in at $285-300 and even if I can purchase it, I am unlucky locating a controller board. So what to do ?

Well, why not look at 4K. This would be a natural jump to maintain a low XY micron resolution while increasing the XY build volume and I knew there are 15.6" 4K panels available. Its nothing new. But they are only found in laptops which is kinda sad since a PC monitor would be ideal.

So, while searching for a controller board for the IPad Pro, I found a very nice LCD + controller package on Ebay. $185.99.

Lets look at the size comparison

Panel: LG LP156UD1-SPB1
Type: IPS
Resolution: 3840 x 2160 px
Size: 15.6"
XY micron: 89.9
XY Build volume; 345 x 194 mm or 13.6 x 7.65 inch

So, by going up in size we at the same time go down on the micron size, isn't that sweet ?

The controller board model isn't mentioned but it gets the job done and there are many more available on the internet. Info will be included in the end of this entry.

Argument for and against going up on the physical build volume.
For: We get a much lager build volume, in fact, twice that of the 9.7" retina. There are plenty of machines that offer a smaller build volume and they are often advertised to people who are working in the jewellery industry, but what if and when you want a part that is much larger ?

Against: Okay, so bigger also mean it possibly creates some issues. The issue I see is the much larger VAT and its an issue due to the fact that each time the resin is exposed to light, even small amounts, it degenerates the quality. This is also an issue with small VAT. So is it a problem ? One way to solve this is by planing how much you print and when.

If we go back to the native resolution for Raspberry Pi, which is 1920 x 1080 p, we already have problems running the 9.7" retina panel and while we can re-code the Pi to obtain 2048 x 1536 p why not look at a different unit all together.

UDOO X86
This can easily run 4K resolution, in fact 3x 4K monitors. It have HDMI as well as DP out, so compatible with the TinySine board and the Ebay controller board I will link to. It is also more powerful than the Pi 3 which isn't a negative thing.

The basic version cost $89

Link: UDOO X86 - UDOO



Since I will use nanoDLP for my SLA, I naturally had to check if I can run this software on the UDOO X86 and yes, the UDOO X86 can run Windows, all Linux distributions and Android. Raspberry Pi often use Raspbian which is based upon Debian Linux

Shahin who is the creator of nanoDLP had this to say: Jessie lite is fully compatible. Jessie was the first version with systemd, early version of nanodlp could not control jessie's (full desktop version) graphic stack completely due to systemd. I think I have solved the issue by adding some systemd configurations to setup procedure. I have not tested that enough to know if it working perfectly or not.But it should not be problem for guys with linux skill to fix any possible issue.

Point being, whatever the nanoDLP desire, the UDOO X86 should be able to run it.

So, so far we have the LCD + controller board for $186 + UDDO X86 for $89 with a total cost of $275. Remember that Adafruit sell the display + controller for $179.90 ($180) which is on its own without the computer only $6 lower than the 4K LCD + controller board .... something to think about huh.

Anyway, here are the links to the Ebay solution.

LCD + controller board. $185.99
1. DisplayPort DP Board + 15.6inch 3840x2160 4K UHD LCD LP156UD1-SPB1(Matte)

Controller board only. $75.88
2. DP Board for 15.6inch 3840x2160 4K LCD LP156UD1-SPA2 LP156UD1-SPB1 LP156UD1-SPC1

LG panel info.
http://www.panelook.com/LP156UD1-SPB1_LG Display_15.6_LCM_overview_26868.html

nanoDLP SLA software.
nanoDLP - Raspberry Pi Based Software for DLP 3D Printers

 

ONO resin. Daylight curing.
Resins - ONO

 

I certainly have the same question - why remove the polerizer when the polerizer is part of the mask, aka make certain area black and other areas lit. Without the polerizing film, the panel is uniform.
I'm not saying there isn't something to removing it, I just can't see the benefit - am I missing something ???



Also, it should not impact the 460nm performance that much since the passage of this wavlength is not so much due to the polerzing film but the transmission in the blue matrix of the RGB.

Polerizing Function.
The function of the polarizer is to improve color and definition , making it possible to see the screens of LCDs. If polarizers were removed from LCDs, it would be impossible to recognize letters or graphics.

Source: LCD Parts

 

I totally get your argument and I definitely share that one with you, you have my full support. My line of thinking is this: If the initial cost for the 9.7" Retina + controller board is that close to a 4K panel and controller board, why not jump up in scale ? I know I want the possibility to print large on something with a respectable size, however, that does not mean a 9.7" will not work, one can always cut the part into smaller sections and glue together afterwards.

And for the TinySine, perhaps you are right and in that case, one should order the whole package from them so that one have a match. And I asked if the controller board would work on the 12.9" Pro version, but it does not - so no controller for that one as far as I know.

 

Hey evilc66, I came in late into this project of yours but I am no less interested / contender. I contributed quit allot over at buildyourownsla forum for a while, even considered building a 4K projector out of the 4K SONY/Sharp display found in the SONY Zperia Z5 Premium, but it turned out too complicated, so it was terminated. So now I am back at LCD which turns out to be the best resort after DLP. Texas Instruments will release a 4K DLP projector this fall, but It can't beat the LCD approach. So, full steam ahead Captain

Regarding FDM, I am actually considering a combo machine, meaning they would share the Z axis but can't be used at the same time. So I have a few options since I have not invested in anything yet, but I must wait until early next year due to other financial areas. But will keep an eye on this one, help when I can and contribute / ask the hard questions when I can.

So, what ever it is you are doing, however small or "meaningless" the effort is (to you) keep us updated.

 

Glad you enjoy my company .. he he.
I'll restrain myself from pulling you into a direction this project is not intended to go.

 

Funny you should say that because THAT is how my mind works and your are in for a ride for the next entry.

I might - or not - have a solution for "our" panel types. What I mean is if everything works out, the IPS panel might transmit UV light at any frequency our hearts desire. I need to go into very technical stuff so I'll compose the text before posting here.

 

LCD based SLA printer is not new, it’s been around for a while now and we know there are 3 commercially available products. There is the Morpheus which comes in at $5000 – yes, five thousand. This one achieves 174 microns on the XY plane. It uses a monochrome panel, aka B&W only, allowing the transmission of UV light. The monochrome panel lack the RGB section. Monochrome panels are often above 21” in size and only feature 2K resolution, that is why the XY micron size is well above 100 microns – I’ll get back to RGB.

There are two that I am aware of which use non altered LCD panels and that is ONO/OLO phone sized printer and then there is Photocentric which achieves 137-96 micron depending on model of choice, ranging from $900 - $4914.

Ideally we could pass UV through our IPS panel. Even if you can, the amount bellow 400 nm is tiny and so, makes no sense. That is where the 460 nm wavelength comes into play since this is the peek wavelength in the blue (B) section of the RGB color filter. While some near UV photons will pass true the (B) filter, the energy is diminished the further down you go on wavelength, its more or less only half at 430nm which is an area where many resins cure poorly since they have a photoinitiator that require < 410nm.

It’s now that I want to become technical for a bit, go in depth of the technology we have available. Basically, can we remove the RGB color filter all together?

Before you read any further, I really recommend that you watch EEVblog LCD panel teardown that I have linked to, its very useful since I will talk about another solution which revolves around light. But let’s get back to the RGB subject.



If you look at the picture bellow, you can see the average construction of LCD panels.
You have a transparent panel, then a polarizer, then the LCD screen, the RGB color filter, yet another polarizer and finally a front glass.



OBS: The technology behind Plasma and OLED panels is very different, there, each color emits the color directly rather than as a filter. Plasma and OLED cannot be used for SLA. As I mentioned, a monochrome display lack the RGB color filter and use the LCD only. The next picture is very helpful in understanding the next section that I will introduce.



Figure 1: Left is a closeup of the Nexus One screen, with a pixel grid superimposed to make it easier to see pixel boundaries. The right image is a shot of the same screen region taken through the debugger.

Source: Secrets of the Nexus One's screen: science, color, and hacks

So what we have is the LCD behind the color filter. It’s not only that the LCD is in an on or off state, it uses a grayscale. If you watched the EEVblog teardown video you could see the individual RGB sections which exist on top of the LCD. The grayscale aids in transmitting each individual color and as we know, white is false positive since it’s a mix between 100% RGB. This is also important.

Let’s assume that the panel use same size on for each LCD matrix. The Retina is 96 microns for each subpixel, we would then get 96 / 3 = 32 microns, but since there is a tiny separation between each RGB section, let’s say that it is 5 microns. 2 dividing section per subpixel, we would get 96 – (2x5) = 86 / 3 = 28,6 microns for each color section or LCD matrix.

Essentially, if we could remove the color filter and access the LCD, we could control the micron resolution on a level no one so far can offer. Sub 50 microns. Envisiontec have the best so far at 50 microns.

Here is a better picture of the LCD matrix.



Source: Color Filters for Displays | Toppan Printing Co., Ltd. Electronics Division

If you check some of the information available, they give more specifics about how the colors are added. Here is a cross section.



and here is an example from patent US7362394 showing how each color is added.

Source; Patente US7362394 - Color filter for IPS and liquid crystal display apparatus

So, Imaging now that both polarizing films, any protective film and RGB film was removed, what you would be left with can be seen on the left in the next picture. The right panel is untouched.



So with a naked LCD panel, we would than add 1-2 polarizing films – restore it – and finally add the front glass panel and we would have a monochrome panel.

At this point in time and space, I don’t know if we simply peel off the RGB film or wash it off with some chemicals like an alcohol. I did contact LG for information regarding the type of color used and how to remove it. Might get an answer or not.

If I had a display, I could start experimenting, but I don’t have one. I am considering purchasing the 9.7” to try to remove the color filter… unless someone is up for the task ?

Going back to the EEVblog teardown video, see how everything is optimized for the light to travel up to the LCD panel. I suggest that this technique is used. Swap out the white led’s with blue or UV – depending upon how it goes with the color filter. Fill all 4 sides with led’s to maximize the energy content and keep the panel slim. Ofc, there is an energy max level, we don’t want to much, but 2-3 sec per 10-25 micron level (Z) would be nice.

As a bonus, here is a video explaining polarizing films.



Any suggestions, thoughts, research is more than welcome, I am not an expert in this area .. yet

 

I never said it would be easy, I expect nothing to be easy about this ... LOL, however, one can look at it as everything ells besides the LCD as garbage. One can always add a new glass front, new polarizing filters and so forth. Science require us to to boldly go where no man has gone before and as far as I know no one has attempted at removing the color filter.

I did check for 4K monochrome LCD's in the 12-17" range and there is non. Place an order at LG for the LCD part only and a few panels is just not going to happen. Perhaps 500 or 1000 could be made, I am sure. Custom orders happen all the time. But we are far from such a position, even combined.

So if you wana have a go at it and share what you find in working on the display(s), I take my hat off and salute your sacrifice. I will probably buy one and give it a go myself in 1-2 months, which seams an eternity away since I want to know something asap.

 

Very interesting read especially with all the activity over the last few weeks. I have almost all the electronics needed to go down this road. The one piece of information I cannot seem to find is what the cure time is on the various resins with a stock Retina display.

I think photocentric is 55 seconds for 100 micron layer and obviously less if you can get a smaller layer size
What about the other resin manufactures I think the only other player in the game at the moment is bucktown has anyone actually tested this resin?
The ONO resins seem interesting if they will ever ship and at what price point.

I am thinking the goal may not be to increase light output but to increase Z-Axis resolution and bring print bed as close as possible to the LCD which i have a few ideas on.

 

What ideas are they ?

One thing that I've been thinking about is the problem that surrounds the sticking of the object to the VAT. We have the build platform, when light hits it, the resin starts to cure, growing towards the VAT. If we know at what speed this occurred, then that is in itself the Z axis speed and resolution. Like with Photocentrics 55 seconds for 100 micron, that is equal to 0.55 s or 550 ms / micron. With that you can have a 1 micron resolution on Z and get layerless since the shift between any layer is more or less continuous. The time to cure 100 microns is the same, but you are not printing at 25 or 50 microns.

So, if you know the curing speed for a few layers, what is the problem of stickyness to the VAT ? and wouldn't one be able to keep the build platform at very close proximity to the VAT as well ?

 

My idea to bring the print bed as close as possible to the LCD is this.

Imagine if you can a sheet of say MDF with a routed out section the dimensions of the LCD. This section would be just a bit deeper then the LCD. Then routing another groove all the way around the LCD about an inch away. Then using a flex vat the dimensions of the groove you can apply downward pressure on the vat to tension the film almost against the LCD.

I like your idea about continuous printing.

 

I am not talking cm here i am thinking 100 micron or less. You could even be level and have the print bed slightly larger then the LCD when you level the bed to the screen you are applying the pressure to the MDF not the panel.

 

Good luck and too bad on the display.