FFB Probe Diagnostic
Current site language English
Download
Hardware database · 06A3:FF12

Saitek Cyborg 3D Force Stick

06A3:FF12 · 3 approved reports · first 2026-08-09, last 2026-10-20 · probe 1.1.2

Device class: joystick

Capability snapshots · latest per OS

What this device reports on each platform

Windows

3 reports · latest 2026-10-20 · probe 1.1.2

Microsoft Windows NT 10
Axes4
Buttons9
POVs1
FF sample period 1000 µs
Min time resolution 1000 µs
Firmware rev0x0080
Hardware rev0x0080
FF driver ver0x20000
Capability flags: DIDC_ATTACHED | DIDC_EMULATED | DIDC_POLLEDDATAFORMAT | DIDC_FORCEFEEDBACK | DIDC_FFATTACK | DIDC_FFFADE | DIDC_SATURATION | DIDC_POSNEGCOEFFICIENTS | DIDC_POSNEGSATURATION | DIDC_DEADBAND | DIDC_STARTDELAY
Supported effects 11
Constant Force
DIEFT_CONSTANTFORCE | FFATTACK | FFFADE | STARTDELAY
Ramp Force
DIEFT_RAMPFORCE | FFATTACK | FFFADE | STARTDELAY
Square
DIEFT_PERIODIC | FFATTACK | FFFADE | STARTDELAY
Sine
DIEFT_PERIODIC | FFATTACK | FFFADE | STARTDELAY
Triangle
DIEFT_PERIODIC | FFATTACK | FFFADE | STARTDELAY
Sawtooth Up
DIEFT_PERIODIC | FFATTACK | FFFADE | STARTDELAY
Sawtooth Down
DIEFT_PERIODIC | FFATTACK | FFFADE | STARTDELAY
Spring
DIEFT_CONDITION | SATURATION | POSNEGCOEFFICIENTS | POSNEGSATURATION | DEADBAND | STARTDELAY
Damper
DIEFT_CONDITION | SATURATION | POSNEGCOEFFICIENTS | POSNEGSATURATION | DEADBAND | STARTDELAY
Inertia
DIEFT_CONDITION | SATURATION | POSNEGCOEFFICIENTS | POSNEGSATURATION | DEADBAND | STARTDELAY
Friction
DIEFT_CONDITION | SATURATION | POSNEGCOEFFICIENTS | POSNEGSATURATION | DEADBAND | STARTDELAY
Reported by OS

Submission count by OS family

Windows 3 reports

Test strength: 35% (1 report), not recorded (2 reports)

Test strength is the share of full force every effect played at, set on the app's checklist and never above 75%. Read weak or did-nothing answers against it. Reports from app versions before 1.1.1 do not record it.

What testers felt · all approved reports

Chip selections by effect

Damper 3 reports
Did nothing x1 Strong damping x1 Weak damping x1
Friction 3 reports
Weak friction x2 Did nothing x1 Sticky x1
Inertia 3 reports
Did nothing x1 Resists changes in motion x1 Strong inertia x1
Polarity sweep — DI east (right) (90°) 3 reports
Pulled forward x1 Pulled left x1 Pulled right x1
Polarity sweep — DI north (away) (0°) 3 reports
Pulled back x1 Pulled left x1 Pulled right x1
Polarity sweep — DI south (toward) (180°) 3 reports
Pulled forward x2 Pulled right x1
Polarity sweep — DI west (left) (270°) 3 reports
Pulled back x1 Pulled forward x1 Pulled right x1
Saturation — 1 co-directional constants 3 reports
Pulled right x1 Soft pull x1
Saturation — 2 co-directional constants 3 reports
Brief jerk x1 Pulled back x1 Pulled right x1
Saturation — 3 co-directional constants 3 reports
Did nothing x1
Sawtooth Down 3 reports
Asymmetric ramp pulses x1 Fast x1 Slow x1 Weak x1
Sawtooth Up 3 reports
Asymmetric ramp pulses x1 Buzzing x1 Weak x1
Sine 3 reports
Rocking back and forth x1 Slow x1 Smooth oscillation x1
Spring 3 reports
Centered x1 Soft spring x1 Strong spring x1
Square 3 reports
Sharp pulses x3
Triangle 3 reports
Rocking back and forth x1 Smooth oscillation x1 Strong x1
Two-axis diagonal — diagonal 3 reports
Pulled back x1 Pulled diagonally x1 Strong pull x1
Ramp sweep — DI east (right) (90°) 1 report
Built up gradually x1
Ramp sweep — DI north (away) (0°) 1 report
Built up gradually x1
Direction consistency · constant-force direction tests

What the commanded directions actually did

Reports are consistent with an X/Y axis swap

Scored responses12
Matched5
Mismatched7
Consistency42%
No direction0
Several directions0
Unparseable0

Scoring covers the Windows polarity sweep and the Linux constant-force tests. A commanded direction is the direction the force comes from, so 0° should move the stick toward the tester. Scoring uses the documented mapping: 0° → Pulled back, 90° → Pulled left, 180° → Pulled forward, 270° → Pulled right. Answers that name more than one direction are counted separately and never scored.

Commanded Expected Observed Matched Mismatched
0° north (away) Pulled back Pulled left x1 Pulled right x1 Pulled back x1 1 2
90° east (right) Pulled left Pulled left x1 Pulled right x1 Pulled forward x1 1 2
180° south (toward) Pulled forward Pulled forward x2 Pulled right x1 2 1
270° west (left) Pulled right Pulled right x1 Pulled forward x1 Pulled back x1 1 2
Technical results · all approved reports

Create / Play / Stop outcomes by effect

Effect Reports Create Play Stop
Damper 3
OK x3
OK x3
OK x3
Friction 3
OK x3
OK x3
OK x3
Inertia 3
OK x3
OK x3
OK x3
Sawtooth Down 3
OK x3
OK x3
OK x3
Sawtooth Up 3
OK x3
OK x3
OK x3
Sine 3
OK x3
OK x3
OK x3
Spring 3
OK x3
OK x3
OK x3
Square 3
OK x3
OK x3
OK x3
Triangle 3
OK x3
OK x3
OK x3
Polarity sweep — DI east (right) (90°) 1
OK x1
OK x1
OK x1
Polarity sweep — DI north (away) (0°) 1
OK x1
OK x1
OK x1
Polarity sweep — DI south (toward) (180°) 1
OK x1
OK x1
OK x1
Polarity sweep — DI west (left) (270°) 1
OK x1
OK x1
OK x1
Ramp sweep — DI east (right) (90°) 1
OK x1
OK x1
OK x1
Ramp sweep — DI north (away) (0°) 1
OK x1
OK x1
OK x1
Saturation — 1 co-directional constants 1
OK x1
(none) x1
(none) x1
Saturation — 2 co-directional constants 1
OK x1
(none) x1
(none) x1
Saturation — 3 co-directional constants 1
OK x1
(none) x1
(none) x1
Two-axis diagonal — diagonal 1
OK x1
OK x1
OK x1
Diagnostic visibility · current policy

Technical diagnostics are aggregated on this page so driver authors and tool builders can compare what multiple testers saw, including Windows and Linux runs for the same VID:PID. They use closed result names and path-redacted errors; the private response-notes field includes typed text and the probe-measured motion.

What's NOT on this page

The one private field

The per-effect responses[].notes field stays off this page. It can contain optional typed text and the probe-measured motion. Capability flags, supported effects, OS breakdown, test strength, chip labels, result codes, failed effects, and run errors can be published in aggregate or as a latest-per-platform snapshot. Read the full privacy contract →

Measured motion is what the probe saw on the stick's X and Y axes while each effect played: the signed peak movement as a percentage of the axis's half range, for example X peak +25.1%. It is measured from where the stick rested when the effect started, so it can exceed 100% when the stick rests off-centre. Condition effects such as spring and damper are not expected to move a centred stick.

Got this stick?

Run FFB Probe and share your result to improve this page.