Design and Testing of the Form 2

Formlabs makes a pretty dang good SLA printer by all accounts. Though a bit premium in the pricing when compared to the more humble impact of FDM printers on the wallet, there’s a bit more to an SLA printer. The reasoning becomes a bit more obvious when reading through this two part series on the design and testing of the Form 2.

It was interesting to see what tests they thought were necessary to ensure the reliable operation of the machine. For example the beam profile of every single laser that goes into a printer is tested to have the correctly shaped spot. We also thought the Talcum powder test was pretty crazy. They left a printer inside a sandblast cabinet and blasted it with Talcum powder to see if dust ingress could cause the printer to fail; it didn’t.

The prototyping section was a good read. Formlabs was praised early on for the professional appearance of their printers. It was interesting to see how they went from a sort of hacky looking monstrosity to the final look. They started by giving each engineer a Form 1 and telling them to modify it in whatever way they thought would produce a better layer separation mechanism. Once they settled on one they liked they figured out how much space they’d need to hold all the new mechanics and electronics. After that it was up to the industrial designer to come up with a look that worked.

They’re promising a third part of the series covering how the feedback from beta testing was directed back into the engineering process. All in all the Form 2 ended up being quite a good printer and the reviews have been positive. The resin from Formlab is a little expensive, but unlike others they still allow users to put the printer in open mode and use other resin if they’d like. It was cool to see their engineering process.

Building One Thing In China

Conventional wisdom dictates that if you need to make a million of something, you go to China. China is all about manufacturing, and there aren’t many other places on the planet that have the industry and government-subsidized shipping that will bring your product from China to people around the world. Building a million things in China is one thing, but what about building one thing? How do you create a working prototype of your latest product, and how do you make that prototype look like something that isn’t held together with zip ties and hot glue? The folks at Hatch Manufacturing have a guide for doing just that, and lucky for us, it’s a process that’s easy to replicate in any well-equipped shop.

In this tutorial/case study/PR blitz, Hatch Manufacturing takes on constructing a one-off smartphone. The Huaqiangbei markets in Shenzhen are filled with vendors selling smartphones of all shapes and sizes. If you want a miniature iPhone running Android, that’s no problem. If you want a phone that looks like a 1969 Dodge Charger with the Stars and Bars on top, you can find it in China. But how are all these phones made, and how do you show off a prototype to factories begging for business?

The answer, as is always the case, comes from one-off manufacturing. Building, assembling and reworking PCBs is a well-trodden path whose process could fill several volumes, but for this post, Hatch Manufacturing decided to focus on the plastics that go into a smartphone or tablet.

Once the case or enclosure is designed with a few CAD tools, a block of plastic is run through a mill. After that, it’s a matter of painting and finishing the latest smartphone that will show up in the Chinese market. Putting a professional finish on a block of plastic is something that will look familiar to anyone who has ever assembled a miniature plastic model. There’s priming, airbrushing, sanding, more painting, sanding, wet sanding, and still more sanding. After that comes polishing the plastic part to a fine finish. It is extraordinarily labor intensive work even for a skilled hand with the right equipment.

Once the plastics are done, the PCB, display, battery, and everything else comes together in a completely custom one-off prototype. It’s very similar to how this would be done in any small shop with a benchtop mill and a dozen grades of wet/dry sandpaper. It’s also something anyone can do, provided they have enough practice and patience.

Spin a PCB for Your Most Beloved Sensors

sensorstick-breakout

If you follow [Ioannis’] lead you’re going to thank yourself every time you sit down to work on a new prototype. He took all of the sensors which he most commonly uses and spun one dev board to host them all.

As long as you’re willing to wait for delivery, the cost of small-run professionally made PCBs has become unbelievably reasonable. That’s really nice when you need to test your layout before exploring larger production. But it also means you can develop your own dirt-cheap yet reliable dev tools. This example combines three sensors which all communicate via I2C:

  • MPU6050 accelermoter/gyro
  • BMP085 pressure sensor
  • SHT10 humidity sensor

Obviously this is a great idea, but key is the cheat sheet which [Ioannis] included on the bottom of the board. It testifies as to which chips are on the board, but also includes the device addresses for the data bus. We’ve adopted the mantra that if a breadboarded prototype is not working, it’s always a hardware problem. For those oft-used parts this should alleviate some of the heartache at your bench.

You could still make something like this without spinning or etching a board. You’ll just have to be creative with the soldering.

Build a bare bones Arduino clone which maximizes its use of real estate

barebones-arduino-clone-at-home

Check out all the stuff crammed into a small swath of strip board. It’s got that characteristic look of a roll-your-own Arduino board, which is exactly what it is. [S. Erisman] shows you how to build your own copy of his YABBS; Yet Another Bare Bones Arduino (on Stripboard).

The strips of copper on the bottom of the substrate run perpendicular to the DIP chip and have been sliced in the middle. This greatly reduces the amount of jumpering that would have been necessary if using protoboard. A few wires make the necessary connections between the two tooled SIL headers that make up the chip socket. On the right hand side there a voltage regulator with smoothing caps. The left side hosts the obligatory pin 13 LED, and the crystal oscillator can be glimpsed on the far side of the ATmega328.

Pin headers along either side of the board have been altered to allow for soldering from the wrong side of the plastic frames. Note that there’s a three-pin hunk that breaks out the voltage regulator, and an ISP programming header sticking out the top to which those female jumper wires are connected.

Ringing in at as little as $2-$4.75 a piece you’ll have no problem leaving this in a project for the long hall. We can’t say the same for a $30+ brand name unit.

[Ladyada’s] thoughts on quick-turn and small-run PCB houses

So you’ve mastered your PCB layout software, and it’s time to make the board. But if you don’t want to etch your own you’ve got to decided where to have it fabricated. There’s a slew of services out there, most of which you cannot afford, but the short list of those you can is still pretty long. We think this set of PCB fabrication house reviews will help you make your choice.

[Ladyada] — aka [Limor Fried] — knows what she’s talking about. She owns Adafruit Industries and has done the lion’s share of designing the many kits and items they sell. If you’re going to charge money for something it better work right, and that involves lots of prototypes. But even if you don’t need a quick turn-around or numerous testing boards the post is helpful as she also covers some of the batch producers we’re already familiar with. These include DorkBot PDX and BatchPCB to name a couple.

[via Reddit]

DIY cellphone

Here’s an interesting concept. Lets make a kit to build your own super simple cell phone. Thats basically what a group at the MIT media lab is proposing with this prototype. Consisting of an SM5100b GSM module and a 1.8″ 160×128 pixel LCD screen on a very basic board holding some buttons, this thing is pretty bare bones. Barely any features aside from sending/receiving calls. It does have caller ID though. At$150, it isn’t really that competitive compared to the phones you’d get from your provider, but it is just a prototype.

We particularly like the laser cut flex areas for the buttons on the front.

[Thanks Paul]

Dead-bug Arduino is still breadboard ready

Here’s a no-PCB Arduino that doesn’t obscure the DIP footprint of the AVR chip. It’s built on an ATmega88 chip, and includes a programming header, reset button, a couple of filtering caps, and an LED. This is modeled after the Lilypad hardware, and fits nicely on top of the plastic case of the microcontroller, allowing it to be used in a breadboard or DIP socket. You can see a walk through of the components in the clip after the break.

We don’t really need most of the components on top of the chip (especially the status LED on the SCK line), but there are several things that we like about this. First off, the programming header is extremely nice. We could see this coming in handy for prototyping where you don’t want to add a header to your final design. Just use a chip socket, and this chip while you’re developing firmware. Once everything is dialed in, program a naked chip and swap the two. The same goes for the reset button, which is nice when working on firmware but may not be necessary in your final design.

This is quite an old project, and we’ve actually seen a successor to it. This is Rev. 2 and we looked in on Rev. 7 back in March. That one is a full Arduino, but the circuit board has no substrate.

Continue reading “Dead-bug Arduino is still breadboard ready”