Hello everyone. I’ve politely been reminded that it’s been a while since an update was posted so I thought this would be a good time to catch you up.
As I mentioned in the last update, the hardware design was “finalized” on December 7th and I placed an order for the prototypes to be made at that time. However, shortly after placing that order I contacted Octavo, the company that manufactures the processor in our design, to inquire about a quantity discount on that part. They offered to review our design "to maximize the chances of first-pass success" to which we agreed and sent over our design files. They replied back after reviewing the schematic and had a few minor concerns that they suggested we look into a little more deeply.
Fortunately, it was early enough that the production of the prototypes hadn’t yet started so we asked to have the manufacturing hold be put on our order until we had a chance to review. After taking a second look at the design files we were able to resolve all the issues easily by switching to a single battery design and not placing a few of the redundant components on the circuit board. After double checking with Octavo’s engineers to confirm the changes should work we gave the go ahead to resume the production of the prototypes on 12/22/2018.
As you can see, the new hardware prototypes have arrived. We received them at our office on Jan 15, 2019. Everything about them looks very good but we have not yet had the chance to power them up and verify their suitability.
The collage above shows the different stages of the hardware development process. We started out by determining what major components we’d need and found what size each of those components are. Then we made real-size cutouts of each component and made a rough model of how we thought it could be laid out. After giving some consideration to how the electronics design would integrate with the mechanical design we made some edits to the layout and created a dimensioned drawing, showing the size and shape of the circuit board and where we wanted the components to be located. Once we had that information we began working with a 3rd party hardware developer and they did the hard engineering work, such as drafting a schematic, creating 3D models, and creating other files that their production machines would need to make our hardware. Once all those design files were created it was time to go ahead and produce the prototypes.
The next step will be to boot the devices and confirm they meet our needs. Once we can confirm that we’ll need to get the circuit board assemblies certified by the FCC.
NOTE: In the time it took to write this update we were able to get the prototype boards booted up. There is still some work to be done to make the camera and Bluetooth operable but the processor is behaving exactly as we were expecting it to. This is a huge step forward and alleviates 90% of potential concerns we may have had about this new system. Things are looking good!
We’ve put a deposit down for the FCC certification so we can reserve some lab time for the testing. The spot we reserved is about 4 weeks away and should take 1-2 weeks to complete. This shouldn’t slow the rest of our development down at all, it just means that we can’t begin production until the testing is complete. This shouldn’t be an issue as we probably have 5-6 weeks worth of work still to do before we are ready for production anyway.
Enclosure and optics
As we’ve been developing the new hardware we were also working with Josh, the independent mechanical engineer that I mentioned in previous updates, to design and model the plastic enclosure. With the design of the hardware done we were able to make significant strides as we waited to the physical prototypes to come in.
Josh started by taking our original design that we created with Becky and built off of that, stretching and shaping the model to fit around the electronics and battery. The photo above was taken as he was proposing a new opening shape that would tighten up the size of the opening and make it more obvious which portion of the page was being scanned.
After a few more iterations we were able to reach the design shown above. One cool feature of this new design is a viewing window that you’ll be able to look downward through to see the line that’s being scanned. This feature will make it easier to stay on track and will also allow a little extra ambient light to reach the page which will help the camera take better pictures.
Josh also took the lens that I designed long ago and created this lens/light pipe combination what will funnel the light being produced by the LED and project it directly onto the scanned text. Based on my very limited knowledge of optics and the calculations I made when I created the lens, this design seemed like it should work. However, before Josh got too much further along with the enclosure design I wanted to double check with an optical engineer just to make sure this system would work.
It’s a very good thing thing that we took that extra step because there was not a chance in the world that it would have worked. First off, it turns out that in order to produce a real image, the distance from the lens to the sensor must be at least double the distance from the object to the lens, a 2:1 ratio. In the optical system I designed that ratio was only 1:4, which would just not work at all.
So now we’re working with a gentleman named Keith. He’s an independent optical engineer who was recommended to us by Josh. We’ve been working with Keith for about the last two weeks and things are moving quickly. He’s developed around 10 rough design concepts that work to balance the system length, complexity, and resulting image quality. We’re having a phone call later this afternoon with both him and Josh so we can all discuss the pros and cons of each system and decide which path to move forward on.
Looking ahead, we expect it to take about another 2-3 weeks until the optical system is 100% complete. From there it will probably be another 1-3 weeks to integrate the new optical system into the enclosure design and make final changes to that. Once that is done we’ll have a very high quality 3D print done to test fit. If no changes are needed at that point we’ll be cutting steel.
Feb. 11, 2019