Monthly Archives: April 2010

Industrial Design, thoughts on Portability

Goals for OpenPCR industrial design:

  1. Sturdy – can be put in a backpack without fear, can be dropped from 6 inches without shattering
  2. Portable – can pick up OpenPCR machine with one adult sized hand, and it will fit in a backpack.
  3. Cost effective –  optimize materials cost + assembly time. ponoko is good because you can cut vents and whatever sized holes you need. This would cost hundreds of dollars from a traditional fabricator, and can’t be done by places like TAP plastic that can’t cut holes or parts <1″.

What is portability? I spent some time in the San Francisco Botanical Garden today thinking about this.

  • Components such as rubber bumpers or handles technically make a device bigger and heavier, but add portability. Which would you put in your backpack, a cube with sharp metal corners or a cube with rounded rubber corners?
  • Dimensions that fit in the hand are much more portable — i.e. what if your digital camera were 2x as long but 1/2 the thickness. Same “mass” from a technical perspective but the 2x long one sucks in your pocket!
  • Portability means that “dirty” parts are covered during a move — i.e. which would you put in the back of your car, a broom with a cover, or a broom with no cover that gets dirt all over your seats. I say a broom with a cover is much more portable.

Case design with Ponoko

One of the engineering hurdles for the OpenPCR project is the case and industrial design. I’ve been looking into ponoko.com over the past couple weeks, and am finally ready to give it a shot.

Google Sketchup is amazing for quickly diagramming your design in 3D. To get ponoko to laser cut the shapes, you’ll need to export the shapes to svg. The tricky part is when you try to add “teeth” to the edges of your design — for us, this is important because we want the case to be easily assembled and sturdy. The open source tool for this, called InkScape, is a total time sink (can you get Illustrator?). That said, as a beginner to ponoko I had my fair share of hair pulling frustrations — your design needs to pass through their uploader and I ran into many many many little issues being new to all the tools involved. Last Sunday I spent 5 hours putting together a simple cube, with the goal of getting a simple price quote from ponoko.

To that end, I worked for 5 hours on Sunday night using Sketchup and Illustrator to pull together a more accurate case for our OpenPCR machine. I shot it over to ponoko and we’ll see how it goes.

For the first edition, I’ve decided to go with bamboo as the base material, which ponoko has ~1/4″ sheets of. Let’s see more tools for biotech made from biology!

Heat sink dimensions

Getting dimensions after the fact..

So I ordered 2 heat sinks from Newegg, hoping that they will fit the 4×4 cm or 3x3cm block that we end up using. These heat sinks have an interesting approach to heat disposal, which lends itself to our desire for a horizontal fan setup.

http://www.newegg.com/Product/Product.aspx?Item=N82E16835233001 ($23)

  • This says it fits a various set of CPU sockets. Looking at the specs for the socket, they range from 3.75 cm to 4 cm, so I think we’re in good shape (I’m assuming this means that the heat sink contact surface is ~4 cm square).

http://www.newegg.com/Product/Product.aspx?Item=N82E16835103074 ($15)

  • hmm, might be cutting it close on this one. The intel sockets it’s design for are 3.75 cm square.

Goals for OpenPCR

We want to produce an open design for a PCR machine. Our goal is to start this project up quickly and finish it. After that, we would like to move onto other things, such as SNPs or PCR kits or synthetic biology.

A DIY Xerox machine for DNA: A fast, computer controlled PCR machine that uses normal PCR tubes and may be built mostly with off the shelf components + free schematics. It does thermal cycling as well as boil, cool, and freeze (4C) samples.

1. Track the development process through blog posts and photos, to follow the design process from idea to reality, and how that reality changes week by week
2. Publish a walkthrough of assembling everything, and a simple DNA test to check that it works
3. A parts list + links to order the parts we use. Schematics for parts that need to be custom made.
4. A basic understanding of PCR patents and how they affect this project (why didn’t this project happen in 2000? can someone sell these as kits without infringing? link)

That said, we are still in the “will this work?” stage rather than the “refine the design” stage of the project. Can a thermal cycler be built from mostly off the shelf components?

Digital reading

Found a $3-$4 chip component that can do an 18-bit (as opposed to Arduino’s 10-bit) analog to digital conversion, and thus can solve our temperature read issues. Requires an $12 Arduino Proto Board to mount, but that was going to be required anyways. Will order and hopefully have working by next weekend.

-Josh

Search for fan/heatsink

Here’s an Intel fan+heatsink that presents an interesting design
http://www.newegg.com/Product/Product.aspx?Item=N82E16835203002

Others:
http://www.newegg.com/Product/Product.aspx?Item=N82E16835154001&cm_re=cpu_heat_sink-_-35-154-001-_-Product
Interesting design.

Similar to intel, half the price:
http://www.newegg.com/Product/Product.aspx?Item=N82E16835103074

All from:
http://www.newegg.com/Store/SubCategory.aspx?SubCategory=574&name=CPU-Fans-Heatsinks&Pagesize=100&Page=1