Project Logs

Collapse

• But the Sun always rises

  Jaime García • 05/16/2016 at 18:58 • 0 comments

  This is an open source project, and that means that even though the company decided to drop it, sources are as available as they have ever been.

  So the project is not dead! and I will take this chance to update the least bits of advance, and talk about this complementary project: DeLePeHost

  First of all, the last update said FEP was the definitive material for the bottom of the vat, an so it remains to this day, giving wonderful results. Further design has been made on the vat, for it to be easier to assemble, and help keeping the tension on the FEP sheet.

  No more physical progress has been made on the design, although there is new commits on the repo that aim to shrink the footprint of the machine.

  Another of the causes the design hasn't progressed, is that I was working on yet another RasPi host for DLP 3D printers (half because I need it for future experiments, half for fun). Yep.

  It's name is DeLePeHost, and I include link to the sources at the end of this post. It's written in Python and a bit of C. It's far from being clean code by the time I write this, but my plan was to make a backend DLP printer host in Python, in a highly readable way, so people can use it give their printers a host they can customize easily for their needs.

  I used Python also because it is a wonderful "glue" programming language, so basically what this projects is, is an amalgamation of several pieces of software I found lying around:

  • On the top level, it establishes a server that allows you to call it's python routines through the RPC protocol, thanks to the wonderful implementation that @Jesús Arroyo Torrens made: Rainbow. This server calls this set of modules, each of one controlling the projector, the motors, and even a slicer if you want.
  • The motors are controled by a extension of the Pyserial class, that adds some routines to interact with the variation of the GRBL firmware that controls the Arduino.
  • It calls Slic3r if you send it an stl file. It's not that much of a burden for the RasPi, because slicing a file into an svg is not that resource hungry. It's still hardcoded to layers of 50 microns, but that could change with the eventual making of an HTML front end.
  • And probably the part I'm more proud of: Projector Control.
    It doesn't depend on a desktop server to control the video signal, because it uses the OpenVG implementation that comes with the RasPi. I got my head around it thanks to the workbench made by @ajstarks, that wraps the implementation in easily callable functions in C. The software still piggybacks on this workbench, but I intend to make my own implementation for performance reasons.
    This C code is wrapped in a Python module, and is called from the Python part of the software. The cool thing about this, is that all the magic happens from CLI (those cold white letters on black background). DeLePeHost doesn't need an X server taking care of the video signal, so this thing is invulnerable to screensavers and windows or mouse pointers popping up into the print.
    That fact, and being that part written in C makes DeLePeHost quite fast when drawing layers onto the screen. This makes DeLePeHost quite a good software to try continuous layer printing, thing that I plan on tackling when I got my life in order again.

  I wouldn't write this huge paragraph without a reason to be excited, and the reason is that 3 weeks ago I retook the project, that was left on an "almost working" state, and finally got it to print. Tried to record a timelapse of it, that my old GoPro messed up almost in the end, but I'll post it anyways:

  The print actually came out well. It was a face of an old man, modelled by a friend:

  So these are my progresses so far. Sunrise may be slowed down to the rhythm of my free time, but definitely not dead. There's still things to do, things that can be done, and things that you can do! because the source is available in GitHub, and if you have any suggestions, or even designs to add, I'll be listening.

  Stay enthusiastic!

  DeLePeHost sources:...

  Read more »

• Things that have been going on

  Jaime García • 05/16/2016 at 17:00 • 0 comments

  Hello maker! Thank you for standing by.

  The lack of updates on this project is more than noticeable, specially when I wrote the previous posts out of thrill. This post here tries to explain that silence, and tries to draw the future lines for this project as well.

  I started working on Sunrise as part of my internship in BQ, in April 2015. I was finishing my studies as an engineer, while working half-time with this wonderful bunch of people, that formed the Innovation Department. By the time I opened this project in Hackaday, I was finishing writing my final thesis, so I fed the log with what I was doing, but didn't have time to update on anything else. I really had the intention to make this an actual log of the progress made on the project.

  I was about to finish the thesis, the department was peaking in creativity, and we were ready to become a quite interactive entity of the open source community, when due to some financial mumbo jumbo the company had gotten into, BQ decided to fire the whole of the Innovation Department

  I won't reflect any opinions upon the subject, because it doesn't matter, and because actually I'm not that mad with them. It's just sad, but for a while they were an opportunity for us to develop this collection of wonderful projects in open source, and I'm pretty happy about that.

  As a TL;DR, I can conclude that this project is currently unsponsored, and all future progress on this project will go on it's own steam.

• So we are trying FEP now

  Jaime García • 03/08/2016 at 15:30 • 0 comments

  Probably one of the most (if not THE most) important component of a DLP 3D printer when it comes to results quality is the material you use for the bottom of your vat. People out there is still looking up for a solution that complies all the specifications. (Transparent, anti-adhesive, non-reactive)

  The Sunrise it's been using this food-grade silicone sheet, that was not-quite as transparent but translucent enough; and not quite as anti-adhesive, but flexible enough to peel off from the part. It's good, gets the job done, but one's always looking for something better.

  The community it's been experimenting with this material called FEP, that has similar properties to PTFE, but somehow it differs in the way it's produced. This material comes in transparent sheets, but it's price makes people choose sheets 0.1mm thick, and that looked way thin compared to the 0.5mm I'm used to. I was hesitant to use it for fear it would rip, but that was until we saw a FEP membrane with a 11kg weight on top, and not ripping at all.

  So we tried FEP.

  It's a 0.25mm film, but 0.1 would have performed good as well. Image is completely clear on the other side of the image, and resin doesn't stick at all. It's not as elastic as silicone, but it just adapts to the shape of the vat wonderfully, and stays there. It gives a nice clear surface, that resists the attack of the resin quite good.

  Because it doesn't blur the image, print quality is increased. Take a look at the hair of this sappho, you can actually see the perimeter of the last layers:

  It so happens that I moved the Sunrise to a room with a window, and it's been sunny lately. With the new optical upgrade, natural light level on the lower side of the vat was quite high, and because the Sunrise a naked kind of machine (for now), the resin cured a bit on the bottom:

  You can see, a resin patch with the same shape of the vat window. There. Sitting on my dear FEP. I thought this would be the end of that sheet, as it happens when resin cures on top of methacrylate, but for FEP, this is actually no big deal.

  @paclema showed me that FEP can do better than that: took a cutter, and peeled of the resin, leaving the FEP intact. That actually happened while I was writing this post:

  So we are back in business! FEP shows no damage, so we will print with it until it does. In the meantime maybe I should design a specific vat for this kind of membrane.

• Sunrise v0.3 hopeful objectives

  Jaime García • 02/28/2016 at 01:11 • 0 comments

  Latest modification on Sunrise were focused in ease of assembly, but then we encountered these new ideas, that could actually lead to a redesign more than simple tweaks. I'll condense them in two main ideas:

  Put the projector sideways

  Is that simple. By putting the projector sideways we get rid of all the empty space on the left of the Optoma, making the machine thinner.

  But there's another advantage.

  The Sunrise is scalable and everything, and you could build a big Sunrise that held a Cinema projector inside (heat generation issues aside) by choosing another length for the aluminium profiles. But, if you put the projector sideways, this may not even be necessary.

  You see, projectors generally have the output lens to one side. and the projector grows the other direction. By putting the projector sideways, (like printing in "portrait" mode) you put that direction, pointing to the front, where there's nothing but maybe the user.

  The point is that we could hopefully fit more projectors in the printer that way, without having to change it's structure.

  Removable printing environment

  This one is difficult to explain so bear with me. If TL;DR go the hyphened paragraph.

  Most of the printers end up having an enclosure to protect the resin an everything, but because this is RepRap inspired, that just hasn't been included yet.

  There seems to be no place to fix an enclosure but just covering the whole machine up, but that was undesired because of hackability purposes. But what if we only enclosed the vat and the platform? like, only the place where the action happens.

  Then we would have the thing that, there's no place to fix the enclosure bu the very vat, but then we would have the axis' arm going through the back wall. OK! let's make that happen, we can make like a bellow so the axis can go up and down without losing the enclosure.

  That would work, but taking out the vat and leaving the enclosure there, dripping onto the projector didn't feel right. Soooo... what if we took the whole enclosure with the vat, and the axis' arm just stuck in there like those gloves you have in labs, passing through a window, for handling hazardous materials?

  -We would have this kind of methacrylate box, that you just put on the Sunrise, make the print, and take it out and away from delicate components for clean up.

  It's utility is still in discussion but, when we build it we will find out.

  You can find an early assembly of this ideas on the Development branch in github.

• Thoughts on the current release

  Jaime García • 02/25/2016 at 18:09 • 0 comments

  When I came up with the Sunrise v0.2, the machine finally acquired the stability it needed to actually use it as a tool. With the proper resin, times went down to something bearable, and really started to give consistent results.

  But, the assembly was still quite difficult: Hammerhead nuts kept rotating before going into the profile slot, some parts needed to be forced into place, some nuts were placed in sort of a "stunt" manner, and slicers with no tolerance adjustment would render the parts with smaller dimensions than intended.

  The next release is intended to solve this issue, but it also has a lot of other features to implement. This last release, v0.2.5, iis somewhat in the middle of 0.3 and 0.2 and changes the style the new parts are made, so they are compatible with tolerance adjustment from source. Also, it implements holders for the nuts not to rotate upon assembly, and some other tweaks. Still, all parts are not updated, so assembly difficulty is expected...
  

  ...but we are the brave ones, aren't we?

View all 5 project logs