Weekend progress!

I had a very productive weekend. I have the first prototype hardware built. I am going to make a few changes for prototype 1.1…

  • I am going to remove the light sensor and use that analog input for the battery voltage. I currently have no idea how much power is left. This is more important than an ambient light sensor.
  • I am going to change the OLED display to a similar model that offers four mounting holes. The one I am using is designed to backpack on the Raspberry PI, which I separated. Having to epoxy everything in was not the best method. It worked, but not pleased with the results.
  • Lastly, I am going to split the midsection of the box to facilitate the ability to screw in the temp/humidity sensor and the display.

I have the sensor pack powered up for the last 14 hours and still going strong. Of course I don’t have the same load on it that it would have in production, nor do I know the battery level.

I am gong to focus on software now and build the next prototype sometime in the next week or so, as I need to order parts. The next box is on the printer now, since it is nearly a two day print.

Here are pictures from this weekend, machining the display protector plastic insert and wiring up the first prototype.

Box progress

The base of the box is good and the mid section is, too! I have redesigned the lid a few times and it is printing now. I have decided ided I am going to laser cut some nylon to make nice end caps for the box.

I plan to update the MS project file later tonight and have my first meeting with my advisor tomorrow evening.

Box is nearly complete.

I have the 3D printer operational again and completed a nearly 40 hour print. The bottom of the box came out perfect. This holds the QI receiver and the Power boost module. The middle part is 98% there. I still don’t like how the lettering turned out, even though I switched it from protruding letters to sunk letters. Also the LEDs are very hard to install and when you glue everything in, access to the display is nearly impossible.

That said, I have modified the solid again, and once the box top is done printing I’ll run this one, one more time.

Also, acetone does react to PLA, dispite what you may have read. This was also a factor in my choice to reprint the prototype…

Back to hardware…

I spent this weekend having problems with the 3d printer. I tried a different brand a filament and ended up with a clogged nozzle. That left me to damage the hot end and well, I need to walk away from it for a while. A new hot end is on order and we’ll try again.

In the meantime I have the ADC connected to the wind sensor. Because the sensor is a 5v device and the Pi is a 3.3v device, I had to use a bidirectional level shifter to keep the smoke in all the electronics.

This quick snap shows the sample software from Adafruit reading the ADC.

Wind sensing

The mass airflow device from Modern Devices came in today. Small device that is powered off of 5 volts and outputs a logarithmic output from 0 to 5 volts based on airflow. What makes this sensor really neat is it so sensitive it can detect the air of a person’s movement all the way up to like 60MPH. Here is a quick demonstration video.

3D

I’m shifting gears this week as I am moving from testing sensors to designing the enclosure this whole thing will fit in.  The idea is to make a somewhat compact box and 3D print it.

I’ve done 2D CAD for many years and am quite proficient in it, however 3D is fairly new to me.  Thankfully there are simple solutions like TinkerCAD out there that will help me get up to speed quickly.  I have downloaded a copy of Fusion 360 and have played with Sketchup before, but both will take some time to grasp and become fluent in it.

Along side that, I’m printing various models out to see what works and what doesn’t.  Part of it is around space constraints, part of it is around designing a box that is strong enough to withstand the environment in which I plan to put it in.   Normally I’d machine something out of acrylic or nylon and use lots of brackets and spacers, so this is all very new to me.

Here is a timelapse of one of the box prints I did.  This print took 28.5 hours to complete.  I’m using a Monoprice Select Mini V2 printer, Hatchbox filament and this model was printed at .2mm layer height at 20% infill.

Time

Last night’s battery run test was a success. The unit ran about seven hours, then wrote a database record and did a safe shutdown. I am running it again tonight to compare, but the tests aren’t what the real load will be, nor is it the final battery back.

I started building the case today. It shows I am not a 3D design modler, but I’ll get there. I am building the case in sections, which will all sandwich together with screws.

Power supply is done…I think.

It seems my plan for a simple power supply ended up being slightly more complex. That’s ok, as long as it is stable. Having to use a DPDT switch as the shutdown needs to go low for low battery and the enable needs to go high to run. Flip the switch, enable is active and the booster kicks in. PI boots and a shell script holds an output high. If you flip the switch to off now, the PI will hold the booster in enable and the shutdown begins. Once the IO lines have been terminated the booster stays on for a few more seconds thanks to an RC circuit. Then it powers itself down. If the battery goes low, the PI will shutdown safely. The booster itself will eventually power down as a result of an even lower state battery. Right now I have the unit running some basic IO functions while on battery power. I have written a quick script that will log the time the system powered down. We’ll see how long it runs with the biggest LiPo battery I have, which is about 1/3rd the size I plan to use in the final design.