Disclosure: Some links in this article are affiliate links. We may earn a commission on qualifying purchases at no extra cost to you.
A force feedback joystick for flight sims does something a spring-centered stick cannot: it moves the point the stick wants to rest at, and it pushes back with force that changes with what the aircraft is doing. Trim shifts the physical neutral instead of only the virtual value; a stalled wing loads the stick; a helicopter and a jet feel different through the same grip. That movable center and those sustained directional forces are the reason to buy one. Rumble and vibration are a side effect, not the product.
The base creates none of that force on its own. Your simulator or a middleware layer does, and the routes are not interchangeable between DCS, MSFS 2024 and X-Plane. This guide covers what force feedback actually changes at the stick, where the forces come from in each sim, and how the three realistic options — the MOZA AB9, the VPforce Rhino and the long-discontinued Microsoft SideWinder Force Feedback 2 — compare once mounting, lever length and software setup are part of the decision.
A desktop joystick and a force-feedback base are not the same purchase. One bolts to a desk and works immediately; the other is a control system whose force quality depends on how rigidly it is mounted and which software is feeding it.
As an Amazon Associate, we earn from qualifying purchases.
What force feedback changes at the stick
The behavior worth paying for is force you feel continuously, not a burst of rumble. A force-feedback base can hold a stick against a trim setting, load it up as airspeed rises, add damping that resists quick inputs, and move the point the stick naturally sits at. That last one — a movable physical neutral — is why FFB matters most for helicopters and aircraft flown with conventional mechanical controls. On a spring-centered stick, trim only changes a number inside the sim while the spring keeps dragging the grip to the same center. With FFB, trim moves the physical center to match the aircraft.
FFB does not guarantee that those forces are aerodynamically accurate. There are two fundamentally different ways the forces arrive, and the distinction matters more than a compatibility line on a box:
- Native DirectInput FFB — the simulator or aircraft module calculates an effect and sends it to the device through Windows force-feedback APIs.
- Telemetry-driven FFB — a middleware layer reads airspeed, control position, trim, angle of attack, ground state or engine vibration and calculates the force itself.
A base can support one, both, or neither for a given aircraft. If you want context on what a yoke, a HOTAS and a stick each do before adding force feedback, the flight sim controls guide covers the control architectures.
Where the forces come from in DCS, MSFS 2024 and X-Plane
DCS supports native DirectInput force feedback, and how strong or detailed that feels varies by aircraft module. A base that speaks DirectInput — the Rhino and the old SideWinder, for example — can use it directly.
MOZA is the exception worth flagging. Its documented DCS configuration tells users to let MOZA Cockpit and its telemetry force model drive the base and to switch DCS's built-in force feedback off, so the sim's native forces and the software's calculated forces do not fight each other. Do not copy Rhino or SideWinder DCS setup instructions onto an AB9.
MSFS 2024 does not drive these bases through a universal DirectInput path. Each one needs middleware: MOZA Cockpit for the AB9, TelemFFB for the Rhino, and community FFB-Bridge software for the SideWinder. FFB-Bridge reads flight data through SimConnect and synthesizes the effects, and it also documents a specific quirk — MSFS 2024 does not always expose a clean ready-to-fly state, so automatic force arming can trigger during loading and manual arming after you are in the cockpit is the safer habit.
X-Plane follows the same pattern. The Rhino uses a plugin and data-reference path through TelemFFB, MOZA uses its own telemetry, and FFB-Bridge supports the SideWinder over X-Plane's UDP/RREF data route.
The practical consequence: Windows and the sim seeing your X and Y axes is not proof that force feedback is working. The chain runs simulator physics → native FFB or exposed telemetry → middleware → Windows/HID command → the base's servo loop → the mechanical stick, and a live axis only confirms the first and last links.
The MOZA AB9: current hardware that can clamp to a desk
The AB9 is the practical starting point for most buyers because it is current retail hardware rather than a built-to-order boutique product. MOZA specifies a dual-servo motor system, 12 N·m peak torque, ±20° of pitch and roll travel, a 1,000 Hz USB refresh rate and 216 W rated power from a 36 V DC supply. The base measures about 227.9 × 157 × 204.5 mm and is PC-only, with documented indoor operating guidance of 5–35 °C.
MOZA lists it for MSFS 2024 and 2020, X-Plane 10/11/12, DCS, IL-2 and War Thunder. It supports both rigid cockpit mounting and an optional desk clamp, and it takes MOZA's own flight grips while accepting third-party grips through adapters.
Two limits matter before you buy. First, MOZA publishes the 12 N·m figure as peak torque and does not give an equivalent continuous rating, so long, heavily loaded pulls are not the same as the peak number. Second, the unit is large enough that owners moving from a normal joystick often find they need to reposition or rebuild their mount. Some owners also report occasional telemetry dropouts and force interruptions, and MOZA's own documentation warns about running other SimConnect-heavy software alongside MOZA Cockpit.
One hardware detail is worth confirming on the exact unit you buy: the current MOZA PDF manual shows a USB Type-B connector while one MOZA setup guide refers to USB-C to PC, so do not build a fixed cockpit cable run before checking.
The VPforce Rhino: a rig component, not a desk joystick
The Rhino is aimed at cockpit builders rather than desk users. It is specified at 9 N·m maximum torque per axis with ±22° of travel (44° total), weighs about 5.2 kg, and stands on a roughly 205 × 180 mm footprint with about 290 mm to the grip connector. Inside is a documented 1:6.2 belt-drive transmission, and current hardware uses 14-bit-per-revolution position sensing — about 13-bit effective at the stick — while units built before the second quarter of 2023 used lower-resolution sensing.
It is a bolt-down product. Installation needs four 6 mm bolts into a rigid structure, the partially exposed transmission needs clearance, and the emergency stop has to stay reachable. The base accepts Thrustmaster- and VIRPIL-style grips directly, while VKB and Winwing grips need adapters — and neither the grip nor the mounting hardware is included.
The Rhino's force is not constant. VPforce documents thermal derating: as the motors heat under sustained high load, available torque is progressively reduced and returns after cooling. That matters more than the headline nine newton-metres if you run a long extension with strong centering. On the software side, TelemFFB adds a lot of capability but is another layer to configure, with DCS Export.lua, X-Plane plugins and aircraft-specific telemetry all in play.
The Rhino is manufacturer-direct and is not sold on Amazon. If you want a mechanical, spring-and-cam base instead of active force — a different set of trade-offs — the VKB Gladiator NXT EVO review covers what a non-FFB gimbal buys you.
Microsoft SideWinder Force Feedback 2: the cheap used experiment
The FFB2 is not a modern competitor; it is the cheapest legitimate way to feel real force feedback. It is a self-contained desktop stick with an integrated two-axis force-feedback mechanism, twist rudder, a throttle slider, an 8-way POV hat and eight programmable buttons, running from USB with its own onboard power electronics. It replaced the earlier Force Feedback Pro around 2000 and has been discontinued ever since.
On modern Windows it is more usable than its age suggests. Windows 11 can run it through the built-in USB HID/PID driver without Microsoft's obsolete SideWinder software, and its USB identifier is 045E:001B. In DCS it can use native DirectInput force feedback directly. For MSFS 2024 and X-Plane it leans on community FFB-Bridge, which reconstructs the force effects from sim data.
The catch is condition. Every surviving unit is decades old, and used-market reports include weak or asymmetric centering tied to worn potentiometers, gears or motors, plus a few dead hat or button inputs. Microsoft never published a modern torque figure, so it should not be ranked against the AB9 or Rhino on force. It is also discontinued and not sold on Amazon. Buy one as an experiment, test every axis and control before paying, and confirm the listing is a USB Force Feedback 2 rather than an older product carrying "Force Feedback" in its name.
Torque means little without lever length
Comparing peak newton-metres across bases is close to useless on its own, because the force your hand feels depends on how far the grip sits from the rotation axis. Hand force is roughly torque divided by lever length.
At a 0.30 m lever, the AB9's 12 N·m works out to about 40 N at the hand and the Rhino's 9 N·m to about 30 N. Move the same figures out to a 0.50 m center-stick extension and they fall to roughly 24 N and 18 N — before mechanical losses, software gain or thermal limiting. A long extension therefore changes the feel far more than the difference between the two rated figures, and it makes mounting rigidity and thermal headroom more important, not less.
This is also why the AB9's 12 N·m needs a footnote: it is a peak figure, and MOZA does not publish an equivalent continuous rating. The Rhino documents its reduction under sustained load rather than pretending peak equals sustained.
Mounting, first setup and the settings that matter
For any high-force base, the mount is part of the control system. Flex in a desk, a mounting plate or an aluminium extrusion absorbs movement and makes strong force feedback noisier and less precise, so a rigid surface matters more here than with a spring stick. Plan the swept volume too: ±20–22° of travel plus an extension occupies more space than a desktop stick, and the base also needs room for its own calibration and self-test movement. Keep the emergency stop reachable from your normal seating position.
Start with low force and raise it only after you have verified axis direction, trim behavior and soft limits; beginning calibration at full gain is how people get surprised by a sudden movement. The controls worth understanding are overall force or gain, spring/centering force, damping, friction, inertia, trim and center handling, soft limits, and vibration or telemetry effects. MOZA's own general starting point is roughly 70–80% overall force, 50–60% spring for fixed-wing aircraft, 30–40% damping, 20–30% inertia and 10–15% friction, with the spring removed for helicopter-style setups — a manufacturer baseline, not a universal setting.
If the base is going on a rig rather than a desk, the flight sim cockpit guide covers how yoke, HOTAS, center-stick and pedal geometry change what you need to mount.
The buying mistakes that cost the most
- Choosing from the torque number alone. Peak force says nothing about continuous output, lever length, or how an aircraft's force curve is generated.
- Assuming axis recognition means FFB works. The sim seeing pitch and roll is not the same as the sim driving force.
- Running two force generators at once. The clearest case is an AB9 in DCS, where MOZA's telemetry model and the sim's native force feedback must not both be active.
- Leaving duplicate pitch and roll bindings in place. MOZA explicitly warns about automatic default bindings in MSFS and X-Plane.
- Pricing the base before the rest of the install. Grip, adapter, extension, clamp or cockpit plate are often separate purchases.
- Underestimating clearance. A high-force base plus an extension needs space for the full range and for calibration movement.
- Putting a Rhino, or a heavily extended AB9, on a flexible mount.
- Buying a used FFB2 without testing every axis and button.
Alternatives worth knowing before you commit
The MOZA AB6 is the lower-force entry into the same ecosystem, positioned at 6 N·m peak against the AB9's 12 N·m. The VIRPIL VPC ATLAS is a current high-force entrant that is still in its initial rollout, so treat it as a product to watch rather than a field-proven equal to the Rhino yet. FFBeast is a boutique, partly DIY force-feedback ecosystem with several motor-power options that demands more component decisions than an AB9. BRUNNER's CLS-E active controls sit in a different price category altogether. Older force-feedback sticks such as the Logitech G940 turn up used and are mainly relevant to buyers willing to lean on modern bridge projects.
For most PC simmers building a first force-feedback setup, the MOZA AB9 is the sensible default: current hardware, broad sim support, and the option of a desk clamp instead of a full rig. Choose the VPforce Rhino only if you already have a rigid cockpit, compatible grips and the appetite for a build and an open middleware setup. Treat a used SideWinder Force Feedback 2 as a cheap way to learn what FFB feels like, not as a permanent answer.
Whichever you pick, the sequence is the same: the software generates the force, the base applies it, and the mount decides whether you feel it cleanly.



