Use joy.cpl and in-simulator tests to tell whether joystick hardware, USB, calibration, bindings or aircraft logic is causing control problems.
Test the joystick outside the flight simulator, then trace the same input through the simulator's controls screen and into a default aircraft. If the signal is already wrong in Windows, suspect hardware, USB or calibration. If Windows is clean, suspect bindings, profiles, assistance, trim, autopilot or aircraft-specific logic.
This general flight simulator diagnosis applies to Microsoft Flight Simulator, X-Plane, Prepar3D, FSX, FlightGear and RC simulators. The same checkpoint method works on consoles, although joy.cpl is Windows-only. A mistake we see constantly is changing sensitivity before proving what signal the operating system and simulator receive.
What is joy.cpl, and what is the PC joystick test called?
If you forgot what the Windows joystick test is called, it is the Game Controllers control panel, opened with joy.cpl. It shows whether Windows recognises the controller and is receiving its axes, buttons and hat switches.
Press Win + R, type joy.cpl, then press Enter. The dot in the command matters. This command does not work on Xbox, PlayStation, phones or tablets.
- Close the simulator. This prevents the simulator, force feedback or an active flight from affecting the test. Leave any service required for the controller to operate running.
- Simplify the connection. Connect the joystick directly to the PC where practical and temporarily unplug unrelated gamepads or controllers. A direct connection is an isolation test; it does not mean every USB hub is unsuitable.
- Select the correct device. Highlight the joystick in Game Controllers and open
Properties. If several similar devices appear, move a control and confirm that you are watching the correct one. - Watch it at rest. Leave the stick untouched for 20 to 30 seconds. A centred crosshair or axis indicator should remain still or nearly still, not creep, flicker or jump.
- Move every axis slowly. Test pitch, roll, twist, throttle, sliders, rotary controls and pedal or brake axes. The indicators should move progressively in the expected direction without spikes, dead sections or sudden reversals.
- Check the usable range. Move each control to both ends. Incomplete travel or an off-centre neutral point suggests calibration, endpoint or hardware trouble, although some controls are not designed to use the entire displayed range.
- Press every button and hat direction. Each indicator should activate when pressed and release afterwards. A recognised joystick can still have one failed or intermittently held button.
- Repeat after connection changes. Try another USB port, restart Windows and, where possible, test the controller on another PC. Moving the cable gently should not make the device disappear or reconnect.
joy.cpl shows controller data after Windows has interpreted it; it is not raw sensor telemetry. Even so, it is the most useful built-in dividing line between a Windows-level problem and a simulator configuration fault. If the centre or endpoints are consistently wrong, follow our ordered PC joystick calibration checks before changing response curves.
Is the fault traceable to the joystick, simulator or aircraft?
Yes. The first stage at which the input becomes wrong usually identifies the part of the signal chain that needs attention.
| Checkpoint | What you observe | Most likely area |
|---|---|---|
| Windows Game Controllers | The device is absent or repeatedly disconnects | USB connection, cable, receiver, power, operating mode, driver compatibility or controller hardware |
| Windows Game Controllers | An axis drifts, spikes, sticks or loses travel | Calibration, physical obstruction, worn sensor or controller hardware |
| Simulator controls screen | Windows is clean, but the simulator shows no input or the wrong input | Inactive profile, wrong binding, input translation software or device detection |
| Every aircraft | The controls screen is correct, but aircraft response is wrong | Duplicate assignments, trim, assistance, autopilot, sensitivity or simulator-wide configuration |
| One aircraft | Other aircraft work with the same controller profile | Aircraft systems, fly-by-wire logic, control lock, hydraulics or add-on behaviour |
The first failed checkpoint is the strongest lead, not absolute proof, because two faults can exist at once. Change one variable at a time and repeat the same movement after each change.
joy.cpl does not create a permanent history of your control inputs. If the problem is intermittent or must be auditable, record the device, USB port, active profile, aircraft and scenario, then capture the live indicators using the recording facilities available on the PC or in the simulator.
The joystick is still frozen or nothing happens - what should I check?
A control that does not respond must be located at a specific checkpoint before resets or sensitivity changes will help.
- Nothing appears in
joy.cpl: reconnect the device, try a direct USB port and verify any hardware operating-mode switch. For wireless controllers, include the receiver and its connection in the test. - The device appears but its indicators are frozen: confirm that you selected the correct device. If no axes or buttons respond after another port and restart, test another PC where possible.
- It moves in
joy.cplbut not in the simulator: select the correct controller and active profile, then bind an appropriate analogue axis. A device can be recognised without having usable flight-control assignments. - It moves in the simulator's input meter but not in the cockpit: release the control lock, disconnect the autopilot, centre excessive trim and check assistance settings, hydraulics and aircraft-specific systems.
- The cockpit control moves but the aircraft does not respond: the input has reached the aircraft. Check airspeed, damage, hydraulic power, fly-by-wire limits and the flight conditions rather than replacing the joystick.
- It is stuck at full deflection: another controller or a continuously held digital command may be overriding the axis. Search every connected device for the same command.
If the controller works on another PC, the fault is probably in the original PC's USB path, controller software or configuration. If it fails in the same way on two PCs, the controller, cable, receiver or interface becomes the stronger suspect.
What does error 00103 mean?
00103 has no universal meaning across Windows, joystick hardware and flight simulators. The digits alone cannot identify whether the joystick or simulator is responsible.
Record which application displayed it and the complete text around the number. Then run the checkpoint test above: if the device is absent from joy.cpl, investigate the connection and hardware chain; if Windows input is clean, the code belongs to a controller utility, simulator or add-on layer rather than the basic joystick signal.
What if the joystick is clean but the simulator still misbehaves?
If Windows shows stable, proportional movement, test the simulator configuration before altering calibration or replacing the controller.
- Confirm the active device and profile. Simulators can store several presets for one controller, sometimes with aircraft-specific variants. Editing an inactive profile changes nothing in the flight.
- Use an analogue axis assignment. Pitch, roll, rudder and throttle controls should normally use their corresponding axis commands. Separate up, down, left or right commands are digital and can create abrupt or apparently frozen movement. Our axis assignment and profile procedure explains how to identify the correct commands.
- Check direction and range. Reverse an axis only when its live indicator moves opposite to the physical control. A response curve cannot correct a reversed binding.
- Remove duplicate assignments. Check the joystick, throttle, pedals, gamepad and any virtual controller. Two devices commanding the same axis can make the controls fight, jump or return unexpectedly; our erratic control troubleshooting sequence covers the usual conflicts.
- Check controller translation software. A physical device and its virtual output can both appear in the simulator. Bind one intended input path and remove the duplicate assignments from the other.
- Return outside influences to a baseline. Disconnect the autopilot, centre the trim, disable automated control assistance and begin with default response settings. Use calm weather so gusts do not resemble input noise.
- Try a default aircraft. Load a simple aircraft in a ready-to-fly state. If it behaves correctly with the same profile, the original aircraft or add-on is the changed variable.
- Build a clean profile if needed. Preserve a known-working preset, then assign only pitch, roll, yaw and throttle. Test those controls before adding secondary commands in small groups.
If Windows recognises the joystick but Microsoft Flight Simulator does not, use our Microsoft Flight Simulator device and profile checks for the simulator-specific part of the diagnosis.
Why does the problem affect only one aircraft?
A fault confined to one aircraft usually comes from that aircraft's systems or control implementation, not a failing joystick.
Check for a control lock, engaged autopilot, extreme trim, unpowered hydraulics, fly-by-wire protection or custom add-on logic. Do not judge a hydraulically powered control surface solely while parked with the aircraft unpowered; some controls will not move normally until the relevant systems are operating.
Should I calibrate, add a dead zone or replace the joystick?
Choose the remedy from the shape of the signal, not simply from how sensitive the aircraft feels.
| Test result | Appropriate action |
|---|---|
| Smooth movement with a repeatable centre or endpoint error | Use the calibration method intended for that controller, which may be Windows, a controller utility or automatic calibration |
| Minor noise only around the centre | Add the smallest dead zone that keeps the centred input steady |
| Clean input, but the aircraft reacts too quickly or slowly | Adjust sensitivity or the response curve in the simulator |
| Sharp spikes, changing centre, button flicker or disconnections | Retest the USB connection and another PC; persistent faults indicate hardware, cable or receiver trouble |
Some modern controllers should not be put through legacy Windows calibration, so check which method the model is designed to use. Sensitivity and curves reshape valid input; they cannot repair missing travel, intermittent wiring or a worn sensor.
A dead zone that must keep growing is evidence of increasing centre noise, not a lasting repair. Large dead zones also remove useful precision around neutral, particularly during trimming, formation flying and approaches.
Does a FlySky transmitter use the same joy.cpl test?
A FlySky transmitter connected to a PC follows the same isolation principle, but the transmitter, USB interface and cable must be treated as one signal chain. FlySky is a transmitter brand, and connection methods vary by exact model and interface.
- Confirm joystick support. A USB socket or physically compatible cable does not by itself prove that the transmitter can operate as a Windows game controller. Some arrangements use a compatible simulator interface rather than a normal USB cable.
- Check
joy.cpl. Windows may display the USB interface under a generic controller name rather than the FlySky model name. Move each transmitter channel and identify the corresponding indicator. - Use a simple transmitter setup. Unexpected mixes, rates, trims or channel reversals can change the signal before it reaches Windows. Begin with a clean model memory where appropriate, then calibrate the channels using the method required by the interface and simulator.
- Map channels inside the RC simulator. If every channel is stable in Windows but the wrong surface moves in the simulator, correct the channel assignment, reversal and endpoint settings there.
If no FlySky input reaches joy.cpl, changing simulator sensitivity will achieve nothing. Test the transmitter-to-interface connection, the interface's compatibility and any required operating mode first.
Our simulator scores look inconsistent between trainees - how can we troubleshoot them?
Make the test conditions identical, then use repeated crossover runs to determine whether the score difference follows the trainee, controller station, profile, aircraft or scenario.
- Standardise the scenario. Use the same aircraft state, fuel and payload, weather, failures, starting position, assistance options, realism settings and scoring rules. Random weather or traffic can make nominally identical runs different.
- Lock the software configuration. Record the simulator build, aircraft or add-on version, active control profile and scenario revision. A changed aircraft model or scoring script can alter results without any hardware fault.
- Verify every controller before testing. Run the same centre, range and button checks in
joy.cpl. Record calibration status and make sure no station has duplicate axes or a larger dead zone. - Use comparable hardware. Different stick travel, spring force, pedal geometry and response curves can affect task difficulty. If different hardware is unavoidable, document it and do not assume the scores are directly equivalent.
- Swap trainees between stations. If unusual scores follow one station, investigate its controller, profile and simulator installation. If they follow one scenario across stations, examine the scenario or scoring logic. Several repeated runs are more useful than one comparison.
- Capture evidence. If the simulator supports replay, telemetry or an instructor log, preserve the relevant input and scoring data. Otherwise record the live control indicators, aircraft state and final result. Remember that
joy.cpldisplays live input but does not log a session.
Do not compensate for unexplained score differences by quietly changing sensitivity for individual trainees. Accessibility adjustments may be necessary, but they should be documented and accounted for in the assessment protocol if scores are meant to be comparable.