Webposted August 1999. Updated February 2003, April and September 2005 Read about phase 2 here.
by Terry Rea, CAR S146, VE2TXR

Various methods devised to capture in-flight footage from a rocket include 8mm still and movie cameras, 35mm still cameras and auto winders, 8mm on-board video recorders, and cameras with Radio Frequency (RF) downlinks in the 434 and 800 MHz bands. Reduced costs for high performance but miniature electronic components, such as Charge Coupled Device (CCD) board cameras, and low-power video transmitters make the it possible for more people to capture full-motion, in-flight footage using telemetry and a minimum knowledge of electronics. This field of the radio amateur hobby is commonly known as fast-scan amateur TV (ATV), where fast scan refers to a full 30 frames per second video frame rate. This technical report presents the analysis, design, construction, and flight testing of an audio/video system that will transmit data to the ground using a low-power transmitter. A Basic radio amateur qualification and a radio station licence were obtained to support this project. Construction and flight testing was conducted according to the Canadian Association of Rocketry Model Rocket Safety Codes, and the concept is easily upgradeable to high power rocketry.
The objective of this report is to outline the analysis, design, construction, and flight-testing of an on-board CCD camera with a RF downlink.
Essential. The system had to meet the following requirements:
Desirable. The system should meet several optional requirements. In particular, it should:

Transmitters and Receivers - General. An investigation into various CCD cameras and transmitters was conducted on the Internet in 1999. In fact, several transmitters (TX) and receivers (RX) are on the market in Canada. Significantly more TX equipment is now available. I have included valid links at the time of the first report. Microvideo offers a TX and RX in the 900 MHz band. They also offer a very small 434 MHz video TX that is receivable on an ordinary television. These systems are available with an audio sub-carrier at an additional charge. Golden-West Investigative Services offers a 434 MHz video TX. This particular TX does not have an audio sub-carrier, but it is also receivable on a TV. Finally, Electra Enterprises (VideoCommTech) offers several quality Frequency Modulated (FM) TX and RX in the 900 MHz and 2.4 GHz bands. The Electra transmitters are available with and without audio sub-carriers but require a receiver and downconverter to view the signal on a TV. Visit this site in Australia for a brief overview of transmission FAQs, and SuperCircuits for a plethora of electronic stuff. Recently, as of February 2003, Wireless Video Cameras targets the hobbyist of all persuasions, offering complete systems..
In general, the vendors mentioned above provided quick responses to my inquiries, with Microvideo of Bobcaygeon, Ontario, and Electra Enterprises of Concorde, Ontario, providing the best quality support and advice. Additionally, Electra Enterprises offers a 1-800 telephone number.
434 MHz TX/RX. The direct to TV tuner concept, although the least expensive option, was discarded early. Generally, the more gain in an antenna, the "better" the reception distance, and consequently the "better" the received, un-amplified signal. Basically, the transmitting and receiving antenna are probably the most crucial components of a TX/RX system, ( the antenna in a Electro-Magnetic system is analogous to the lens of a camera in an optical system. ) A typical TV dipole antenna has a very small gain of approximately 1.3 dBi. Hence, operation of a TX at this frequency with transmitting and receiving dipole antennas will likely lead to poor results. Construction of a high-gain receiving antenna was not considered due to time and cost. In addition, 434 MHz is an Amateur Television Station (channel 59) frequency and operation on this frequency would be susceptible to reception/transmission interference.
900 MHz TX/RX. Systems operating at 900 and 2400 MHz require RF receiver/downconverters so that the signal can be viewed on a TV. The Microvideo 900 MHz receiver, with a whip antenna, has downconverter outputs to standard RCA audio and video jacks, allowing normal viewing on a TV/VCR combination. The receiver could be connected to a Yagi-Uda gain antenna to improve reception range. This higher gain antenna would improve performance, but would also increase the time and cost for the system development. In addition, dipoles and basic Yagi antennas are linearly polarized; thus, best reception is obtained when the TX and RX antennas are aligned in parallel. Major roll, pitch, and yaw of the launch vehicle would ensure antenna misalignment, and loss of signal, at some point in the flight.
2400 MHz TX/RX. The Electra 2.4 GHz FM TX consists of a 2' dipole antenna connected to a 3" by 1" by .4" heat shrink-wrap IC board. The TX is also available with a microstrip patch antenna. The 2.4 GHz receiver appears to be adapted from a standard Wavecom wireless video receiver unit. The receiver antenna is a 2 7/8" square microwave patch antenna. The RX patch antenna should offer approximately 3 dBi gain, which is a substantial improvement over a dipole. This should translate into increased reception range and improved signal. The patch antenna also offers circular polarization. Thus, RX the antenna is theoretically able to better receive the TX signal component regardless of TX antenna orientation. The RX/downconverter NTSC/PAL output can be connected to a VCR or TV for viewing. In sum, there are no specialized RX preparations to be effected, other than connecting the RX to the VCR and the VCR to the TV.
Given these factors, the Electra Enterprises TC-2400 TX and RX system was chosen over the Electra 900 MHz and Microvideo 900 MHz systems.
The Electra TX specifications are:
Camera. CCD camera provide generally superior performance over less expensive CMOS cameras; hence several cameras were investigated, with a variety of power consumptions and current drains. A small, "no-name" 100 mA@9 Volt pinhole B&W CCD board camera was found a local security company. This was the least expensive CCD camera I could find as my budget was already being stretched. Very little analysis went into the selection of this item. In the end, it has a relatively high resolution, infinite focus range and auto-gain control for low to bright light environments. It has a small 3-pin connector, of the same type used on Pentium II, III heat sinks, making wiring very simple. SuperCircuits offers a wide range of small CCD board cameras that you may want to investigate.
Both the RX/downconverter and the TV monitor can be run from an automobile battery via a 300 Watt inverter plugged into a cigarette lighter. This provides complete portability.
Note. Since the initial report in 1999, I now capture video imagery direct from the downconverter to a handheld digital video camera, significantly reducing the required field equipment.
The TX payload was designed to fit a previously built 48.2 mm diameter rocket - an Aerotech Cheetah that has been flying for several years. The transmitted picture should remain upright on descent; however, most payload sections hang upside down from the parachute on deployment. This would result in an inverted image during descent. In order to receive upright video throughout the flight, the payload section was designed to have an ejection charge duct to allow the ejection gases to blow by and push the nose cone and parachute out above the payload section.
As I could not decide whether to look out the side or put a mirror to look at the ground, I tested a compromise. The camera field of view (FOV) was split with a mirror. The top half of the video picture frame looked down the tube to the fins and the bottom half out the side towards the horizon. Testing showed the resulting imagery to be a little bizarre, and there is an opaque zone across the image, caused by the lower edge of the mirror. Given that approximately 15% of the FOV is obscured along the line caused by the mirror edge, the mirror was removed so that the camera FOV is strictly out the side of the rocket.
The payload wiring schematic is shown in figure 1 and the layout in figure 2.

Figure 1. Payload Wiring Schematic
Figure 2. TX and CCD Camera Layout
The camera requires a minimum of 9 Volts and the TX requires 9-12 Volts. The camera and TX are wired in parallel. Together, a single 9 Volt battery will just provide the minimum electro-motive force (EMF) to operate the system.
Figure 3 shows a modified body tube (with mirror shroud) and coupler. Also visible is a mock-up of the camera mounting board, camera, and TX. Note the black ejection gas duct through the coupler. A single 9-volt battery illustrates the scale in this photograph.
![]() |
![]() |
| Figure 3. Transmitter, Camera and Coupler Prototype with split FOV | Figure 4. System showing mounted Camera and battery |
Figures 4 and 5 shows the front and back, respectively, of the nearly complete flight model with the IC board. The TX is on the reverse. An ON/OFF slider switch is visible behind the camera. The camera is fastened to the payload frame with two small nylon bolts. The TX is fastened at flight time using double-sided adhesive tape. Future high power rocket applications will require a more rugged and permanent mounting methods. Figure 6 shows the prototype mounted in the airframe. The dipole antenna is mounted vertically. The antenna was later mounted externally on the airframe as an inverted-V. The battery, just visible in the airframe, is held in place by the coupler, the mounting board, and the airframe.
![]() |
![]() |
| Figure 5. Reverse showing transmitter (in black), ON/OFF switch, and antenna with coaxial feed line. | Figure 6. Prototype mounted in 1.9"airframe |
Static testing shows that the camera alone will operate for 45 minutes on a single rechargeable NiCd 9 Volt type battery; however, the NiCd battery nominal voltage is 7.2 Volts. The current drain for the CCD camera and the transmitter together is significant. During early testing, it was found that the complete assembly is functional for approximately 30 minutes on a single, fully charged 9 Volt NiCd battery. Results from additional tests are shown below at table 1.
Point to point reception range, across open ground is given below for a clear dry day over asphalt. Both TX and RX antennas were approximately 2.5m above the ground.
Table 1. Field Test Results for 14-15 August 1999, Innovation Drive, Kingston, Ontario
| Test | Battery Power Supply | Configuration | Range |
| 1 | 9V Lithium | Full payload with parachute. Antenna mounted vertically, internally above camera with shield | 400m |
| 2 | 9V NiCd rechargeable (7.2 V nominal) | Full payload with parachute. Antenna mounted vertical, internally above camera with shield | 250m |
| 3 | 9V Lithium | IC board only, with antenna in free space | 500m |
| 4 | Two NiCd rechargeable in series (~14 V nominal) | Fully configured payload, with antenna configured as Inverted V dipole externally on the airframe. No shield | ~600m |
Flight simulations were performed using WINROC 4.0 and yielded satisfactory results. Figures 7 and 8 show the simulation outputs for two flights, the first on an Aerotech E30 and the second on an EconoJet F20. A four second delay was selected for the E30 motor and a 7 second delay was selected for the F20.
Figure 7. Aerotech E30 Simulation
Figure 8. EconoJet F20 Simulation
The flight model of the transmitting camera system was flown 4 September 1999, at Blazing Archer III, near Picton Ontario. This launch was hosted by the Ottawa Rocketry Group, and held at the CFD Mountainview airfield. The ORG obtained the airfield and an airspace waiver for the event. The launch ceiling was 11500 ft and no impulse limit was imposed!
I have no objective data on the weather conditions, except to say that the weather conditions were clear, very hot, and quite humid. The wind was calm to still. In essence, both flights had a good straight vertical boost and ejection was at or very near apogee. Both flight tests are considered successful. The boost footage in the first test is rock stable, with virtually no roll and excellent signal. The second sequence, equally good in reception shows that the airfield, launch tower and exhaust trail are clearly visible.
Sequences were recorded on VHS tape and digitized using the Digital Processing Systems (DPS) Perceptions Video Recorder and Adapter. The video was sampled at 30 Hz using the DPS hardware CODEC and then encoded to MPEG. The MPEG file does not do justice to the original sequence.
Test |
Motor |
Altitude (calculated) |
Still |
Video |
Comment |
1 |
AeroTech E30-4 |
210 m (4 sec delay) |
|
(4 M) Lift-off to apogee |
good flight excellent audio/video signal with little platform roll minor airframe damage on recovery |
2 |
EconoJet F20-7 |
360 m (7 sec delay) |
|
(5.3 M) Lift-off, parachute deployment and partial descent |
satisfactory flight excellent audio/video signal significant platform roll during boost minor airframe damage on recovery |
Additional comments: Following the first flight the payload and airframe were examined. Minor body tube crinkling was evident in the airframe; however, the payload was in perfect working condition. It was flown again two hours later with no modification, and used the same batteries. Inspection of the F20 motor after the second test showed a small piece of copperhead igniters in the divergent passage of the nozzle. No other nozzle abnormality is evident and interestingly, roll is dampened completely during the coasting phase. Airframe damage upon landing excluded further flight tests; consequently, maximum transmission range was not determined. The signal was lost upon touchdown in both cases.
Cost Breakdown
In Canadian dollars, component costs were, approximately:
This document outlined the requirements specification, research, design, construction and flight results of a B&W transmitting CCD camera with a microwave downlink. The camera, transmitter, and receiver were purchased as separate components and integrated into the overall payload. The payload design easily met the minimum requirements, as well as some desirable requirements. Maximum transmission range was not determined during these tests. A microwave downlink provides a high bandwidth, line of sight signal. Overall, quality in-flight video and audio was obtained via the microwave downlink.
Before increasing the output power, antennas must be improved. A high gain receiver antenna is the easiest first step, and would increase the gain from 3 dBi to over 14 dBi. Noting that gain is logarithmic, this would be a substantial range improvement. At the payload, radiated power can be better directed towards the ground, and gain improved with a microstrip or patch antenna on the rocket. The net result would be an improvement in reception range and signal quality. A flexible microstrip can also be conformal to the airframe, reducing aerodynamic drag and improving launch vehicle performance. Similarly, a quarter wavelength patch antenna could likely be fabricated from a section of printed circuit board, although the higher gain would also make the antenna directional, causing reception problems with roll, pitch and yaw. A conformal, phased array microstrip would make an excellent engineering project for college or university students. The transmitter has sufficient bandwidth to handle colour video, so an upgrade to a colour CCD camera can be performed. This would yield excellent in-flight video, and offer a great way to study flight dynamics from the launch vehicle.
![]() The author, with ground crew The red rocket is the launch vehicle used in the article above. Watch my CAR Level 2 certification attempt - an all fibreglass ARCAS clone - fly on an I-211 and disappear into the wild blue yonder. I hate spam! Email me at terry@engarde.caX but remove the X from the address. I have done this to prevent web robots from harvesting my email address and sending me all that unwanted garbage at my expense. |
All available in-flight video footage:
If you made it this far, you must really be interested in this stuff. Here are some links to related websites:
14 September 2005 |