XBee module woes

I have been using XB24-AWI-001 modules for Airball so far. During testing I noticed that if they didn't have line of sight (e.g. if the probe was above the car roof and the display was inside the car), the data would cut out.

Probe pose!

The official Airball spokesmodel poses with the probe.

Go for road test

Road test

No fancy pictures or videos of "flight" testing since we did not have an independent camera person and did not wish to sacrifice safety.

Melissa and I had a pre-flight safety briefing, established the "knock it off" protocol and the chain of command, and went over our test plan and what could go wrong. We conducted some tests, took some data, then returned home and had a quick debrief.

Our main safety issue is that holding a 5-foot pole with a thingey at the end of it is really hard work for the muscles, so any further testing should be done with some sort of mechanical fixture.

On the other hand, holding the pole made me aware of how turbulent the air is. It is not clear if that or my wobbly hold or something else caused the effect, but the airball display was wobbling all over the place. Actual flight testing will determine how much, if any, lowpass filtering we need in the device.

The XBee units seemed to need line-of-sight -- we would lose data if the probe was over the roof of the car. This might require a change in the form factor, e.g., an antenna that you stick onto the side windshield of the airplane. We'll see.

I have yet to determine the stability of my static pressure measurement. I don't think it's really worth messing with right now because the probe body is very poorly sealed. A future, properly sealed probe should give us better information. The ideal of course is that the static pressure remains the same going from zero to cruise speed and back. I'm sure most static pressure sources don't achieve that much, but it'd be interesting to see how well we do.

I have a bunch of data to crunch over the next couple of days and will be posting graphs as I have them. Stay tuned.

Probe noses

A trio of Airball probe noses, fresh from the 3D printers at Shapeways!

Probe innards

On/off switch and panel mount USB port added. Port allows both charging and programming. Probe innards complete; now waiting on mechanical parts.

Sensor milestones

TWO Airball milestones this evening:

1. First closed loop aero-derived alpha/beta/Q.

2. First battery powered probe.

This is the first time we've put the whole system together and derived actual data from the airflow using the sensor we have designed, and displayed it wirelessly on the display unit we designed.

This is not an *operational* milestone (yet) because we don't have things wrapped up for proper use, but it is a really important technological milestone and means we are almost ready for our first car-borne road testing, followed closely by airborne tests!

Video here:

Sensor board

Some complete and work-in-progress pictures of the Airball probe "sensor board". This puts all the pressure sensors onto one modular unit with 4 signals going to it: power, ground, I2C clock, and I2C data. This way I can experiment with the "compute board" as I want to without risking damage to my expensive sensors: The stuff you see here is $200 worth of kit!

The soldering could be better as always since I'm a n00b but it wires up properly and works!

Now waiting for 3D printed probe parts to assemble the first functional probe and do a full system test.

Breadboarded sensor

Breadboarded Airball sensor probe. Two differential pressure sensors and one gage sensor, all 10" H2O range. One BMP280 barometer. Things of note:

* My 3D printed 5-hole probe nose turned out to be clogged; need to make more.

* The Adafruit BMP280 library for reading the sensor seems to crash whenever I use it with my code. That's sad because the BMP280 is a pain in the neck to read -- you have to download all these calibration parameters from the thing to correct its readings.

Next steps are to figure out if we can get the BMP280 library to work, and to get some more probe noses built so we can try out very simple alpha/beta/Q measurements. And hook up an XBee to the Arduino's serial port to get this information to the display.

Wireless pressure milestone

Airball milestone: first complete path of pressure signal wirelessly to display.

Damaged board

And now a story from Lamella, the Lame Day of Lameness.

The Sparkfun Fio V3 is an Arduino breakout board that has built-in connections for an XBee wireless module, and can charge a single cell LiPo battery automatically through the USB port. Nifty, eh?

Well so where's to buy a LiPo? It's really expensive to get them shipped, because LiPo, and fire, and stuff like that.

So I went instead to Aero Micro, in Sunnyvale, the well-loved hobby shop. There I had the folks find me a LiPo battery with the correct JST connector on it. They made me buy it before I was allowed to connect it to anything, which turned out to be a wise precaution....

Perry (the well-known proprietor) plugged it into my board, and there was an immediate loud pop. A sliver of something flew off the board and hit Perry in the forehead, alarmingly close to his eye. :( And the magic smoke was released from the board, never to return.

On subsequent investigation, it turns out the JST connector on the battery, while compatible mechanically, was wired with the opposite polarity to what was on the Fio V3 board.

Electronic parts are replaceable, but I'm really scared that Perry could have gotten hurt in this stupid misadventure. I'll have to be careful myself when wiring up new and interesting power electronics together.

Sensor shipment

$150 worth of pressure sensors from Digi-Key. Came in a gift box.

Beginning probe work

Beginning work on probe. Pressure sensor integrated with Arduino. This is an All Sensors DLHR-L10G 1in. H2O gage sensor with an I2C interface.