Cohesion3D MicroFactory: Electronics Inside Plastic

[Cohesion3D]

Hello everyone,
I'm Ray from Cohesion3D. I'm working on a machine that can print any model using UV-cured resin and then print a circuit inside it with conductive inks. I've worked with circuitry since I was very young but I've actually never had a circuit board manufactured because it requires either toxic chemicals or costly and time consuming fab houses. Even though services like OSHPark have made this cost quite small, there's still a waiting period and pressure to get it right the first time, which can be detrimental to a healthy design cycle. I like how 3D printing lets me produce a design very soon after having an idea, without having to iterate tens if not hundreds of times before being able to justify the expense of having it manufactured. Being able to hold a working part right away gets the creative juices flowing and improves productivity, and I want to extend this ease of use to electronics prototyping as well.

Cohesion is a whole new way to prototype and manufacture hardware products as there is no circuit board at all. You just designate a cavity inside any CAD model and add your circuit board design there. At it's core, the Cohesion3D Microfactory is a multi paste extruder, so just slice the model as you normally would for a dual extruder print and go. It's still up to the user to place the electronic components by hand, as adding pick and place functionality would increase the complexity many times and add hundreds upon hundreds of dollars to the cost of each unit.

This is going to be an extremely easy to assembly kit. I've built other printers from kits before and have experienced the frustration of trying to line things up, having to measure, and perhaps worst of all, having to guess. I've addresed the weaknesses of 10 other printers I've seen in detail, as well as combined their strengths as I designed Cohesion. Assembling the frame is as simple as snap together and tighten some screws, and the toolheads are modular and interchangeable, which means you'll be able to add functionality such as a filament extruder, several different kinds and sizes of syringe and pump systems, and a lot more by simply snapping a module onto the carriage.


I really am trying to make this the best printer you can buy at the really low price point of $1199 considering all the functionality and customization it will offer. I'm planning to launch a crowdfunding campaign early 2015. I would be extremely grateful for any suggestions, complaints about your existing printers that I can avoid with mine, and questions that I can answer.
Please check out http://cohesion3d.com and tell me what you all think!

[soofle616]

As a circuitry design hobbyist this is VERY interesting to me. I all too well understand the hassle and cost of having pcb's custom made for my various projects and would love to have a way to make functional prototypes at home. I will offer the following thoughts to you:

1) Cost of the machine. While certainly not unreasonable for a 3d printer, given the cost of custom PCB prototypes, it's a little... hopeful. I can have a custom pcb design fabricated and shipped to me in about 5 days from several different sources for around 100 bucks (for 3-5 boards depending on the supplier) So at 1200 dollars I'd need to generate 13 different designs at a minimum before this machine would pay for itself, and that doesn't take into account material/maintenance/repair costs for your machine. While this is certainly not a deal killer, it does reduce your potential market to those who create a lot of designs. Someone like me who only makes maybe 1-2 a year is just going to continue buying them from someone else for $100 a pop

2) Your website needs a lot of help (that sounds like I'm being mean but that's not my intent as explained further). You have a single picture of a rendering of a machine, no pictures of the actual physical machine, even in a prototype stage, no multiple angles to allow people to get a good look at what you're building, no videos of it moving (or better yet, printing). You also have no real useful information about the machine or the parts it produces, just a generalized statement of it's capabilities with no specifics. Ok great, no pcb, how do I attach my components to the printed part? Can i use SMD's? just how conductive is the conductive paste/ink? Can it be used for logic level parts or is it only good for higher voltages (more than 0-5v)? What kind of current capacity does the material have? Etc., etc.

3) how bout a sample part with electrical components installed showing that the finished part does actually do what it's supposed to do. At the very least some cutaway views of sample parts showing the passages with conductive ink in them.

4) what are the expected costs of ownership? Material costs, repairs, maintenance, etc?

It really is a great idea and I'm not trying to be a naysayer here. I'd love to see something like this come together and see the light of day and would be interested in having one myself. I realize that you're probably very early in the design/build stage of this project so a lot of my questions will be answered as you progress so again, please don't take this as me bashing you or your idea.

[Cohesion3D]

As a circuitry design hobbyist this is VERY interesting to me. I all too well understand the hassle and cost of having pcb's custom made for my various projects and would love to have a way to make functional prototypes at home. I will offer the following thoughts to you:

1) Cost of the machine. While certainly not unreasonable for a 3d printer, given the cost of custom PCB prototypes, it's a little... hopeful. I can have a custom pcb design fabricated and shipped to me in about 5 days from several different sources for around 100 bucks (for 3-5 boards depending on the supplier) So at 1200 dollars I'd need to generate 13 different designs at a minimum before this machine would pay for itself, and that doesn't take into account material/maintenance/repair costs for your machine. While this is certainly not a deal killer, it does reduce your potential market to those who create a lot of designs. Someone like me who only makes maybe 1-2 a year is just going to continue buying them from someone else for $100 a pop

2) Your website needs a lot of help (that sounds like I'm being mean but that's not my intent as explained further). You have a single picture of a rendering of a machine, no pictures of the actual physical machine, even in a prototype stage, no multiple angles to allow people to get a good look at what you're building, no videos of it moving (or better yet, printing). You also have no real useful information about the machine or the parts it produces, just a generalized statement of it's capabilities with no specifics. Ok great, no pcb, how do I attach my components to the printed part? Can i use SMD's? just how conductive is the conductive paste/ink? Can it be used for logic level parts or is it only good for higher voltages (more than 0-5v)? What kind of current capacity does the material have? Etc., etc.

3) how bout a sample part with electrical components installed showing that the finished part does actually do what it's supposed to do. At the very least some cutaway views of sample parts showing the passages with conductive ink in them.

4) what are the expected costs of ownership? Material costs, repairs, maintenance, etc?

It really is a great idea and I'm not trying to be a naysayer here. I'd love to see something like this come together and see the light of day and would be interested in having one myself. I realize that you're probably very early in the design/build stage of this project so a lot of my questions will be answered as you progress so again, please don't take this as me bashing you or your idea.

I appreciate the response and I of course understand the value of honest feedback. I wouldn't get very far if people only told me what they thought I wanted to hear. Thank you.

1) I'm basing my costs on a few factors. The first one is the state of the market. Several months ago a good 10 inch x 10 inch build platform printer with a single filament extruder cost $999. Some still do, but cheaper ones are starting to appear. CartesianCo has a $1600 printer that will make print circuits on a variety of substrates using silver inks. BotFactory has a $2500 machine that prints the circuit, applies conductive glue (instead of solder paste), and Picks and Places the components. Both of these companies did quite well on Kickstarter, certainly proving the market for rapid prototyping electronics. I'm trying to take it a step further.

2) I absolutely agree with you regarding the website. A lot of things are in flux (no pun intended) right now so I don't want to put up images of the current state only for it to change in a few weeks. That said, I do hope that I have development reach the next milestone in a few weeks and can start publishing that. For example, the cantilever approach wasn't rigid enough, the other end of the arm would shake a bit too much during travel. I had planned to mill the relevant plates out of a rigid material like Garolite but instead decided to keep all the parts printed, this way it's easier for other people to make/ repair/ upgrade theirs. Thus, there is now a vertical beam on each side and this keeps everything much more solid.

3) Again, will work on this in a few weeks. I have designs anticipated such as the bunny with a blinking nose, which leads me to functionality:

I anticipate 3 modes of usage, depending on user skill level and how involved they want to be in the design.

1. Basic: Just as a multi material 3D Printer or basic PCB prototyping tool. Export a CAD model of your traces from your preferred circuit design suite, design a plastic floor under it, and hit print
2. Intermediate: Separate a CAD model into several separate prints (horizontally slicing it). Essentially, every circuit board section is the top layer of a print job. Take the bunny with a blinking nose: a bottom section with a battery holder, a middle section section with the timer circuit, and a top section with the LED. What's cool is that it's possible to print wires with a vertical component using the conductive ink to connect all the sections, or just leave grooves in the plastic to run conventional wires. When the print finishes, place the components by hand, and assemble the print sections together using pegs/ glue. Both SMD and Thru-hole components can be used, for thru holes it may be necessary to designate a small hole in the plastic under the circuit for the legs of the component.
3. Advanced: Instead of chopping the model into several pieces that go on top of each other, simple create a rectangular cutout inside the 1 piece model and place the circuit there. Standard procedures apply for generating gCode for a multiple material print. The user must find the layer at which the circuit is printed and insert a pause command after it. This way, the printer will stop after creating the circuit and prompt the user to place the components. After this, the user hits resume and the printer continues to create a seamless object. With Repetier-Host, it's possible to scroll through the layers of gCode so it's not a difficult task but some people may find it a bit intimidating.

4) Costs of ownership. You'll want spares of tubing, needles, and syringes for printing in different materials (each set is a self contained system as to not contaminate another material). I'm sure eventually these will get gunked up and you'd want to replace them but the costs are fairly nominal. As far as material - filament can be had for $30 - $50 per kg and will be useful for larger or more rigid prints, replacement/ upgrade parts for Cohesion. UV Resin, which is safer (as no heat is required), and can produce more detailed prints, is a bit more expensive at $65/ liter. For both of those, that's quite a lot of printouts for that quantity of material. For the conductive inks, I'm currently using Bare Conductive but the resistances are way too high for functional TTL circuits. It's also much too viscous to extrude out of a fine needle. Still, great way for me to experiment and I'm about to order more suitable inks for testing. I'm ballparking $10 worth of conductive ink for a board of several inches in length each side, plus $1 worth of plastic for the floor under it as described in the Basic scenario. Clearly, electronic components are a separate ballgame.


As you've correctly pointed out, I am in the early stages of my venture. I wanted to get the conversation going as soon as was possible.

Some pictures of the current state are included. I'm changing the design a bit to make it better looking and easier to assemble. Clearly I want to finish the functionality modifications and clean it up before posting to the site.

Thanks,
Ray

2014-10-24 23.02.19.jpgCohesion Original Bowden.jpgCohesion Kraken Carriage Render Full Unit.jpg