By Javier Rollon · 2022-03-15
The Jetstream 32 took three years to build. That's a long time for any software project. For a one-person operation working evenings and weekends? It felt like a decade. But I wouldn't trade a single frustrating night for what came out the other end.
Everyone wanted me to build a CRJ-700 after the CRJ-200. Bigger plane, more popular in real-world operations, obvious commercial choice. I picked a turboprop that most people had never heard of instead. Typical Javier decision, my friends said.
Here's why. The BAe Jetstream 32 is a pilot's airplane. Twin turboprops, conventional flight controls, no fly-by-wire safety net. You feel everything through the yoke. In turbulence, you work. On approach in gusty crosswinds, you earn your landing. That's the experience I wanted to give simmers — not another glass-cockpit autopilot exercise, but real stick-and-rudder flying.
Turboprop engines are strange beasts. The relationship between power lever position, torque output, prop RPM, and airspeed is nonlinear in ways that catch people off guard. You can't just push the throttle forward and go — you manage torque limits, ITT limits, prop sync, beta range for ground operations. Each of these had to be modeled individually.
I spent four months just on the engine model. Not the 3D model — the thermodynamic simulation. Compressor stages, combustion efficiency curves, turbine section behavior at different altitudes and temperatures. The real Jetstream's Garrett TPE331 engines have very specific quirks. Pull power too aggressively at altitude and you get a temperature exceedance. Forget to sync the props and you get a persistent vibration that's subtle but annoying, exactly like the real thing.
The CRJ was about systems — FMS, autopilot, electrical buses. The Jetstream is about feel. The flight model reacts to weight distribution, to fuel burn shifting the CG aft over a three-hour flight, to the difference between flying in rain versus clear air. X-Plane's blade element theory handles most of this automatically, but getting the prop model right required endless iteration.
I tested over 200 different propeller configurations before the torque response matched the pilot's operating handbook. Two hundred. Each one required loading the aircraft, flying a standard departure, checking climb rate, checking torque values at different power settings, comparing against the real numbers. Most developers would've stopped at twenty. But "close enough" isn't good enough when a real Jetstream 32 captain is going to fly your aircraft and immediately know if the torque doesn't feel right.
Turboprops have a sound. Not the generic whine of a jet — a rhythmic, mechanical pulse that changes with prop RPM and blade pitch. Getting that right was the moment the Jetstream stopped being a project and became an aircraft. I recorded reference audio from three different angles, mapped the frequency patterns, and spent two weeks in the audio editor making sure the cockpit sound at cruise power was distinct from takeoff power. Because it is in real life, and pilots notice.
The Jetstream taught me that building aircraft isn't about features — it's about authenticity. Every hour invested in getting one small thing right pays dividends in immersion. That's not a business strategy. It's an obsession. And it's why three years felt like the right amount of time.
Javier Rollon develops aircraft for JRollon Planes. Follow on Twitter.