NOTE: The entire site is severely (many months) out of date. An overhaul is planned for some time in September. Thanks for your patience.

Asking Her Out... With Banners and Rockets

This is the story of a hairbrained idea which quickly became a full-blown hairbrained project and spiraled into two weeks of insanity, completely derailing all other projects and throwing the Lab into an incredible state of disarray.

It was worth every moment and every penny.

The Plan

The idea was simple: to present to a certain person, by some entertaining and unusual means, a certain proposition concerning accompaniment to the pre-graduation formal dance known as the Senior Ball.

As in most projects of this nature, the plan began simply, grew to ludicrous complexity, then settled back to practical proportions and was set into motion with little idea of what the outcome would be. It was eventually decided that the proposition should be presented by means of a series of banners which would unroll into view from a hidden position, accompanied by a pair of model rockets which would launch in sequence with the the deployment of said banners.

Preparation

Banners and banner-drop rig

IMAGE: The second banner after painting
The second banner after painting
IMAGE: The third banner, freshly painted
The third banner, freshly painted

Three 7' x 7.5' banners were constructed, each formed from strips of roll paper joined together with clear packaging tape into a large single sheet. Each sheet was painted with a portion of the message to be presented, then fastened to and rolled tightly around a length of 1" thin-wall PVC pipe, attached to a second pipe at the loose end, and tied tightly with a nylon ribbon to prevent the banner from unrolling itself.

The release mechanism for each banner is exceedingly simple. The ribbon tied about the center of the banner is threaded through a simple device consisting of a nickel-chromium filament soldered in place across a hole in the center of a piece cut from a printed circuit board blank. Each of these devices is wired to the controller board, and when a current is applied, the filament becomes hot and cuts through the ribbon almost instantly.

IMAGE: The banner-release rig, ready for deployment
The banner-release rig, ready for deployment

Control boards

A hardware design was established which could be utilized in both the rocket platform and banner-drop systems, and easily reused in future projects. At the heart of each control board is the Atmel ATTiny2313 microcontroller. A ULN2003 NPN darlington array drives three high-current outputs (using six outputs in parallel pairs), and a low-cost ASK radio receiver daughterboard provides a wireless data link. A single LED is used for status indication.

IMAGE: Newly-completed controller board
Newly-completed controller board

The total parts cost per board is approximately US$10.

The firmware for the controllers, written in C, allows multiple boards to be addressed on the same frequency, and includes extensive error-checking. Command packets are sent over the radio link in a serial stream at 1200 Baud, 5 data bits per frame, even parity.

Rockets and launch platform

IMAGE: Remote rocket launch pad
Remote rocket launch pad

A launch platform was fabricated from some scrap aluminum and a pair of straightened heavy-duty coat hangers as guide rods. Two tumble-recovery model rockets were constructed from paper and cardboard, fitted with garden-variety "A8-3" solid fuel model rocket engines, and placed on the guide rods.

The engines used are intended to be initiated by an electric ignitor inserted into the nozzle. Because the cheaply-made ignitors packaged with these engines require a relatively large surge of current to perform reliably and have a tendency to "short out" during or after the ignition process, a separate battery pack and a pair of relays provide an isolated power system for ignition purposes in order to prevent the controller from failing due to supply voltage fluctuations in the case of such a fault.

Control transmitter

A simple transmitter unit was constructed around a premanufactured RF module and connects to the serial port of a PC to allow asynchronous serial data to be transmitted over the radio link to the receiver/controller units. The range was found experimentally to be about 0.5km line-of-sight.

The transmitter was originally designed to mount directly to the back of a Zaurus SL-5500 handheld PC, but it was discovered the day before deployment that the Zaurus's UART (or perhaps its driver) does not support frame lengths less than 8 bits, so ultimately a laptop PC had to be used instead.
Section in progress. Must find photos.

Execution

One of the most exciting parts of the project was working through the night to get all of the hardware finished on time, then sneaking out, loading up the van, and setting off at 2:54 in the morning to spend hours in the damp cold thinking "What the %@#$! am I doing?" while fumbling about in the darkness with a pocket full of tools and a flashlight between my teeth...

Timeline, May 15th

Results

The sequence was initiated at 12:11:06 PST on May 15th, during the lunch break. Unfortunately, both rockets failed to launch due to an intermittent connection to the ignitor battery pack, which was not discovered during testing. However, all other parts of the system functioned as expected and the three banners successfully deployed in sequence. The third banner broke away and fell to the ground shortly after deploying, but the message was visible long enough to be read.

IMAGE: The first and second banners during deployment
The first and second banners during deployment
IMAGE: The third banner after breaking loose
The third banner after breaking loose

Above is a series of still images from a video recording (thanks, Nick) showing the banners during the deployment sequence. Due to distance and camera motion, image quality is less than superb, and certain details are intentionally obscured for privacy reasons.

Conclusion

Despite various technical issues, the project was a success in that it achieved its original goal.

...and the rocket rig will find use some time in the future...