For your viewing pleasure

In case you are interested and have about 8 minutes to kill, here is a video I shot while the Dalek was printing. I have a float on a lever in the resin tray (on the left side) that blocks an opto interrupter. Every 10 layers the printer checks the sensor and if it is unblocked it triggers a relay that turns on a fish tank air pump. Air is pumped into the resin jar and forces resin up the tube to the vat. the hissing sound is a pressure relief valve (a small hole in the cap) on the jar. Without it the resin kept flowing for quite a while after the pump stopped. You can see I eliminated the power tilt and went with a passive design. The tray is hinged on one side and allowed to lift and fall back by itself. The thud is the sound of the tray dropping back down. Interestingly, You see that after the first few layers there is almost no movement of the tray. Once the build platform gets some distance from the floor it doesn't have the suction when it lifts. Exposures with small surface areas release so easily that the tray often doesn't lift at all. Unfortunately that also makes it difficult to tell if there has been a printing failure, such as the part separating from the build platform. I changed the front of the resin tray to clear so I could spot a failure before it made a big permanent blob and ruins the Sylgard.

The first couple layers get a longer exposure (about 9 seconds) to stick it to the build plate. After that the exposure is 5.5 seconds per layer. The Z axis lift is also higher for the first few layers. I found that after about 5 layers I only had to lift about 1/8". That helped cut some time off the total print run.

 I also changed the vat floor. My first floor was a piece of 1/4" Plexiglas with a Sylgard 184 coating. I have since switched to regular double thick glass. Nothing special just plain glass. By using glass I can heat cure the sylgard in a toaster oven in about 40 minutes instead of waiting 24 hours for it to cure at room temp. I made the Sylgard layer a little thicker and so far it has held up well, 5 or 6 prints with no noticeable degradation. Also added is a red plexiglas filter/shutter that blocks the lens before and after printing to eliminate accidental exposure during set up and clean up. you can hear it being moved out of the way just before the projector comes on the first time.

Doctor! Doctor! Give me the news.....

Here are some things I have printed with my completed DLP printer. I'm still trying different resins and pigments to find the right combination of performance and cost.

There is a limit to the printer's ability to produce really small details. surface details like the lettering on the call box print ok but details that extend away from the larger body like the door handle or the posts around the light  don't get enough energy from a tiny dot of light to cure the resin.


TARDIS

EX-TER-MI-NATE!

The Dalek came out really well. The TARDIS and Dalek were both from Thingiverse.com. I printed them both hollow to conserve resin. I used MakerJuice subG+ Orange resin from makerjuice.com It works pretty well but I did notice some over curing of the layers. I think I need to add some more pigment. I'm also not a fan of the translucent orange. It doesn't photograph well and it looks awful. The orange does do a good job of blocking the wavelengths that the resin is sensitive to. I'll try adding something to make it more opaque and less fluorescent tangerine.


The Dalek took about an hour to print. 470 layers @ 5.5sec exposure + z axis cycle time (about 2 sec) and pauses to top off the resin level.

 

Here are some tests I did with the MakerJuice SubG+ Black resin.

 

My first  attempts to print more "mechanical" stuff.

You can see the over curing problem on the undersides of the bolts.

 

This one came out really well. The problems I am having with the over curing don't seem as noticeable on more organic shapes. It's only the hard edged overhangs where it is noticeable.

 

 

I wanted to see how small I could go.

 

 

 

This was my first successful print. I set it up so there were no supports needed. I used Spot-A GP unpigmented resin with a mix of black and white pigments. I was going for a neutral grey similar to the Form1 resin. I like the more opaque pigments, it is much easier to see details. The spotA resin cures faster than the SubG+ but is more expensive.

 

IT LIVES!

Wow, I don't know where to begin. Where did the last 18 months go? I made the mistake of putting this project on the "back burner" and then forgot all about it. It's not that I didn't intend to get back to it, but somehow I never found the time or motivation.

The good news is that I finally dusted it off and finished the build! There were a few small changes to the design.

The active tilt mechanism with the springs and stepper motor has been replaced with passive tilting. Now the tray is hinged on one side and is allowed to lift on one side as the build platform is lifted. The problem I had with the springs was there was too much play in the vat location. If I tightened it down so it didn't "wiggle" then it wouldn't tilt smoothly. I tested the passive tilt with some resin and it seems to work fine. If I lift the platform 1/8th inch the resin separates from the vat floor and the vat falls back to the level position (isn't gravity great!). For the vat floor I'm using a piece of 3/16" Plexiglas covered with a thin layer of Sylgard 184. I don't know how long the Sylgard will last so I have the frame set up so I can swap out the floor easily. I'll probably go ahead and make several  from glass so I can scape off the Sylgard and re-coat as needed.

Since I no longer needed the stepper and axis for the tilt, I used them to add a red filter between the projector and the vat. It's just a piece of transparent red plexi that swings in front of the projector lens. This is useful because before I start the VB script Macro in Mach3 the projector is showing the Mach3 interface. Also, when the script is done it goes beck to a mostly white screen that would expose a lot of resin. I tested the filter and it seems to block all of the wavelengths that my resin is sensitive to. I also used a large piece of the red plexi to make a door for the front so I don't have to worry about ambient room lights.

I'm still tweaking the control macro and need to reposition the home and limit switches but the dry runs look good so far.

I promise it won't take another year for me to post some more. I've already started playing with some resin and should have some results soon.

Where does the time go?

I can't believe the Summer is over and winter is just around the corner. If you haven't already guessed, I have way too many interests and find myself too easilly distracted by the project of the moment. The good news is that now that the weather is turning nasty and it gets dark early, I'll be working on indoor distractions (like the 3D printer).

I'm going to try and force myself to stay focused and actually finish this project before I flit off to the next one. I have been thinking on some changes to the Z axis design. I would like to be able to buy a precision linear stage but the cost is prohibitive. I'm pretty certain I can build something that will be "good enough" but I'm still working out the details. There are a few other changes I want to make to my original design concept. That often happens when my concepts meet the reality of fabrication.

Once I work out the Z axis It shouldn't take too long to assemble. I hope to have a prototype ready for testing by the end of the year.

Thanks for checking back, I'm sorry for staying away for so long.

Now with Real Tilting Action!

I've been working on the resin tray. I needed a way to be able to test different coatings/coverings for the glass at the bottom of the tray. To that end I have made a frame that will hold the glass. I can prepare glass pieces with different coatings and change them easily. I do feel that I cheated a bit as I used my CNC to make the frame. I probably could have made it with my table saw and router table but the CNC was faster and much neater.

The tray will slide into a holder that floats on springs. It can be leveled by adjusting the bolts at the corners. The holders were made with much more mundane tools. A hacksaw for the aluminum and a table saw to cut the HDPE blocks. A drill press was used for the holes. What you don't see is the threaded inserts I found at the hardware store. They screw into the MDF and the bolts are threaded into them. The blocks slide up and down nicely on the bolts (I found some that had the top 2/3 unthreaded) but I found that I needed to open up the holes on the underside of the blocks so the bolts could "wiggle" a bit . Otherwise they would bind up when only one side was pressed down.




Here you see the finished assembly with a stepper motor to provide the tilting action. There is a eccentric (off center) wheel that presses down on a lip on the back side of the tray. The tray will be tipped between layer exposures to help separate the cured resin from the vat floor. Then the Z axis gets raised slightly before the next slice is exposed. I may need to relocate the stepper so it won't interfere with the Z axis.

Baby steps

I have started the modifications to turn my CNC router into a test bed for my 3d printer design.

I added a platform to the Z axis that will hold the build platform. The resin tray will hang off the edge of the table so I can have the projection from below. For now I just have a piece of Plexiglas as a placeholder for the tray.

Watch the video to see it in action!




What you see is Mach3 running a macro. The macro opens a PowerPoint slide show. Then lowers the z axis quickly to the top of the tray then slowly to the bottom so it doesn't splash down into the resin. The macro then shows a slide for the set exposure time alternating with a blank(black) slide while the z axis repositions itself for the next layer. Then the next slide is shown . Lather, rinse, repeat.... When the last slide is done the z axis is raised to its full height so the build tray can be slid out of the holder. The images you see on the screen will be projected up from below the Plexiglas. Of course all the timings will need to be adjusted once I start testing with actual resin.

Next up... I'll add the projector. I just need to get a VGA splitter and a front surface mirror.

Details, Details

I've been doing a lot of brainstorming with a kindred spirit (Jon) I met over at the Chemshapes.com forum (formerly Snygro).

My ultimate goal here is to come up with a 3d printer that can be built cheaply from easily obtainable materials without access to a full machine shop. That being said, I need a practical test bed to try some ideas. To that end I am going to make modifications to my diy cnc router and add 3d printing capability to it. The router already has a functional z axis and all the electronics. This way I can work out the design details and transfer that to my dedicated desktop printer design.

On the software, I had planned on using MS PowerPoint to create a slide show of my layer slices but was having a very difficult time figuring out how to time each slide with a blank(black) transition that was long enough. It could be done with custom animations but not to all the slides simultaneously. If there were hundreds or thousands of slides that would really suck! I couldn't find a slide show program that would let me set the dark time between slides so I was trying to learn Visual Basic so I could write one for myself. The problem is I was having trouble wrapping my brain around it and was getting frustrated when I couldn't get it to do what I wanted.

In rides Jon Watson to the rescue.
He solved two problems in one stroke.

1. I needed to create a slide show with the appropriate timings and 2. then coordinate that show with the motion control of the printer. At first I was thinking of having some sensor on the printer that "looked" for when a blank slide was displayed so it knew when to make its moves.

Jon came up with a macro to apply the animations to all the slides automatically and then use Mach3 to call up the slides and coordinate them with the motion control software. Anyone with a background in CNC knows about Mach3 but for the rest of you, It's a program that uses something called g-code to control a CNC machine. I own a copy but wasn't planning on using it for this project because it is a significant investment (for a hobbyist) and I wanted to keep the final build cost as low as possible. The really cool thing about Jon's solution is that he uses Mach3's scripting ability (something I hadn't even looked at) to do all the work. That means that the trial version which has a limit of 500 lines of code will work because the script doesn't actually need to run any g-code. Mach uses Cypress Enable a much more (I think) intuitive scripting. After studying Jon's script and going over the Cypress manual I was able to make a few tweaks and removed the need (I hope) for the custom animations in PowerPoint. Now you just place a blank slide as slide #1 and Mach3 will alternate the blank slide with the slice slides. Of course I still need to set the exposure and delay times. That can only be determined once the test bed is up and running and I have some resin to test.

I'll post all the code when I'm sure it isn't buggy.