When cross-braced canopies re-emerged onto the skydiving scene in the mid-late 90's, jumpers began experimenting with wingloadings much higher than had previously been thought practical. With the increased wingloading came high-speed landing approaches designed to maximize speed, and therefore available lift, for landing. And with these high-speed approaches came the concern that an AAD could be "tricked" into firing at an inopportune moment—low to the ground under a working parachute.

Some of the early adopters of the new cross-braced technology performed tests in which they were unable to fire a CYPRES AAD during a high-speed landing approach. This was something of a relief to those who were pushing the envelope of canopy wingloading. The increased popularity of over-water swooping prompted many to assemble special "pond-swooping" rigs without a CYPRES installed, further rendering the issue moot. The CYPRES 2, however, has a water resistance feature, presumably to address the growing pond-swooper population. It is this fact combined with some ad-hoc harness turn testing that prompted me to revisit the question.

Jumping an Icarus Extreme VX 86, I did a high pull from 13K. I initiated a toggle turn and then kept flying the turn using only harness input from that altitude down to about 4K. Counting the seconds in my head, I measured the time it took to descend from 11K to 10K, 8K to 7K, and 6K to 5K. I calculated my descent rate at the highest altitude to be about 113 MPH (yes – freefall speeds!), and around 85 MPH in the lower, denser air.

The question in my mind became "can the speeds necessary to fire a CYPRES be generated at the lower altitude at which the CYPRES would actually fire?"

Canopies: Icarus Extreme VX; 99, 86, and 74

Wingloading: 2.3 (99), 2.6 (86), and 3.1(74)

AAD: CYPRES (original model) DOM 08/96

Descent rate measurement: Video camera trained on an analog altimeter

DZ Field elevation: 250 ft. MSL

Weather: Sunny, Low 80’s

I placed the CYPRES unit in a CamelBack backpack with the control unit taped to a position closely approximating an installation in a standard rig. I tied a small loop of spectra through the cutter assembly hole and taped it down to the cutter.

Prior to each jump, I set the CYPRES unit for +1500 foot landing field elevation. This would allow me to discontinue the diving turn at an altitude sufficient to enter a normal landing pattern.

On each jump, I exited the aircraft at 13K and deployed at 9K. I then stowed the slider, turned on the camera to record the descent rate, and flew to a clear area for the descent. At approximately 4K, I initiated a 180 toggle turn and continued the turn using only harness input. At 1500 feet, I discontinued the turn, found my place in the landing pattern (I ended up landing with the mid-point group from the Otter load), and landed.

I started on the largest of the 3 canopies, the 99 loaded at 2.3, then worked my way down to the 86, then the 74. Approximate descent rates for each of the canopies are shown in Table 1.

After landing the 99, I went to the packing area and opened the CamelBack. The spectra loop on the cutter was intact, and everything looked fine. The CYPRES had not fired. It was time to downsize and try again.

The 86 descent was noticably faster than on the 99. After landing, a number of people were curious about the results. I thought there was a good chance that the unit had fired, and upon inspection, found that I was correct. The unit had, as designed, fired sometime during the descent from 4K to 1500 feet.

I no longer had an intact CYPRES unit to jump, but I decided to jump the 74 in order to get the descent rate data for comparison. Surprisingly, the 74 does not descend significantly faster in a harness turn than the 86 (refer to Table 1), which raises the question of what the upper limit ("canopy terminal velocity") may be.

Table 1: Descent Rates by Canopy Size (MPH and ft./sec., rounded)

The original model CYPRES is designed to activate when a descent rate of 78 MPH is detected at an altitude of 750 feet (the same descent rate "trigger" is specified for the new Vigil, as well). According to the manual (Page 9), "In the event of a breakaway below this hight [sic] Cypres will operate down to apx 130 feet AGL." Thus, we can conclude that any high-performance landing approach that builds enough downward speed prior to 130 feet AGL could result in a Cypres fire. I think that my test demonstrates that this is not outside the realm of possibility.

If you are jumping an aggressively-loaded canopy and are using a Cypres (or similar "expert model" AAD), consider carefully your landing approaches.