I flew my plane with its newly fastened lights on Monday, November 17th in the gym our indoor club uses. The plane flew really well and drew some encouraging comments. All went well until the last flight (it's always the last flight). I met a 3D head on and we both tumbled to the ground. Me because I had lost my main wing, him because he graciously and quickly cut the throttle! The damage actually was not as bad as I feared. The main wing had a number of tears as shown below, however, my main fear was the delicate wires and lights I had just added. The wires were wrapped around his prop and motor. This took some careful and delicate work with needle nose pliers, but eventually they came free. I retrieved the wing and stuck it back on. Between the tears in the wing and motor being knocked off its mount it was amazing the plane actually flew! However the flight was not pretty and the landing snapped the propeller. So off to CellarDwellar for a new wing and prop. Glued the motor back in place and stuck a new wing on, re-wiring the lights back in place. I tested the results outside a couple nights later, but at -5C the batteries only lasted 2-3 minutes.
On November 28th at the golf dome, the plane flew wonderfully. I had also altered the elevator authority in order to improve my inverted flying ability. It worked - although I think I could go one more notch.
I have decided to try and put lights on my Parkzone Citabria. I wanted at a minimum wing markers that flashed. The process involved
researching a light, low power circuit that could pulse 2 LEDs about
once a second. In my research I found a site that describes a
fascinating circuit that uses only a single transistor
(and not even all the pins!) However, I was not able to recreate the
flash as the voltage of the Citabria battery is only about 4V. However,
in writing to Mr. Cappels, he offered a novel solution I had not
thought of - microprocessor control. He generously programmed an
ATTiny12 to flash 100ms pulse at 4Hz, plus a few
others - which turned out to be a good thing.
Flashing at 50ms 2Hz
turned out to be more eye catching. Before committing to anything, I first plugged the chip into a experiment board along with a number of LEDs to confirm power consumption, flash rate, circuit wiring etc. The resistors in line with the LEDs are all 150 Ohm - being from surplus they are different power ratings (not a real concern at the low power I am using - 0.1W).
The 50ms at 2Hz was the ideal rate, so the process of soldering everything together began. To save weight Dick Cappels suggested surface mount LEDs. I agreed - then I saw how big they are. SMT LEDs are small! Soldering them was to be a challenge.
To further save weight, I opted for enamel coated wire from a solenoid. This did not make the soldering process easier...
Soldering leads to the LEDs was the first major step. To begin with I unwound a length of wire and formed it into a loop as follows:
Then I stripped the enamel off with a small lighter (burn the enamel carefully, then wipe the ash off with an abrasive cloth). Here is the result:
Next I carefully wrapped the loop around the LED such that each side was wrapped around a contact:
I then carefully soldered the wire under a magnifying glass:
Then I snipped the excess wire from the loop leaving the LED at the end of 2 wires.
Once that was done, I tested the circuit - only having a few spare components to work with, I did lots of testing at nearly each stage. Here's the yellow LED blinking:
Once the yellow LED was complete, I repeated the process for the red LED. I also wired up a white LED direct to the power (via a 1K resistor) for a landing light. The power consumption of the complete circuit was very satisfying. Between blinks it is only a couple of mA. The flashing LEDs only draw around 40mA during the pulse.
That was satisfying - next, soldering the circuit. I opted to wire direct to the chip without using a socket. There was an increased risk of damaging the circuit, but again, I'm adding this to a plane that weighs 20+ grams! I'm nervous about what the extra weight will do. So I cut the unused leads and solder the components direct to the chip:
The leads from the red and yellow LEDs were soldered to the resistors, leads from the power and ground were soldered as well. The whole package was then taped into the top inside cavity of the Citabria. I attached the circuit near the idle CG point (highest point on the wing curve) to retain the balance. I also noted which side the antenna wire from the receiver protruded into the cavity and attached my circuit to the opposite side.
Take care when running the wires to avoid getting tangled in the servo mechanisms.
Here's a peak through a hole in the belly (that happened the first time I flew the plane...):
The leads were run out to the wing tips and taped in place:
The power leads were run through the underside of the fuselage and soldered to the battery leads (no power switch installed):
Here is the final test showing the LEDs in action. The purpose of the setup in the chair was to also permit a full throttle battery test. The engine ran over 7 minutes at full throttle, which is still lots of flying time. Actual flights at 1/2 to 3/4 throttle have approached 10 minutes with the lights one!! Enjoy! (Sorry, no sound on my basic still camera with movie mode)