A color wheel is a rotating disc divided into pie-shaped optical segments — typically dichroic filters or phosphor coatings — that a light beam passes through to produce a timed sequence of colored output. You'll find one spinning inside every single-chip DLP projector from brands like BenQ, Optoma, and older Texas Instruments reference designs. The wheel synchronises to the imaging chip so each color frame lands during the matching segment. The result is a full-color image from a single monochrome modulator, or in astronomy, calibrated narrowband exposures from one CCD.
Color Wheel Optics Interactive Calculator
Vary wheel speed and segment angle to see segment dwell time, color switching rate, and a synchronized animated color wheel.
Equation Used
Segment dwell time is the time one filter slice stays in the light beam. Convert wheel speed from RPM to revolutions per second, then multiply the revolution period by the segment fraction theta/360.
- Wheel speed is steady.
- Segments are equal angular slices.
- Spoke blanking time and sensor jitter are ignored.
Operating Principle of the Color Wheel (optics)
The wheel itself is a thin glass or aluminium disc, usually 40-80 mm in diameter, divided into angular segments. In a DLP projector each segment is a dichroic filter → a thin-film stack that reflects one band and transmits another — bonded into a metal hub. The hub rides on a brushless DC motor spinning between 7,200 and 14,400 RPM, which corresponds to 2× or 3× the video frame rate (120 Hz or 180 Hz for 60 Hz content). An optical index pulse, usually a small magnet plus Hall sensor or a notched encoder track, fires once per revolution so the DMD chip knows exactly which segment is in front of the lamp at any microsecond.
Why spin it that fast? Because the eye integrates color over roughly 16 ms. If the red, green, and blue segments cycle slower than that, you see the rainbow effect — colored fringes trailing fast eye movements. Texas Instruments solved this by going from 1× wheels (single RGB sequence per frame) to 6× wheels with RGBRGB or RGBCYW segment patterns. The trade is segment dwell time: faster wheels mean shorter time on each color, which means the DMD has fewer microseconds to modulate gray levels, and bit depth drops.
Tolerances matter. Segment angle error above 0.5° throws off the spoke time — the brief moment when the beam straddles two segments and must be blanked — and you get color bleed at segment boundaries. Wheel runout above 50 µm axial causes the beam to walk off the filter edge near the hub. Bearing wear shows up first as audible whine around 4 kHz, then as image flicker as the index pulse jitters by more than 10 µs. A failed wheel motor is the single most common reason a 5-year-old DLP projector ends up on the bench.
Key Components
- Dichroic Filter Segments: Pie-shaped thin-film interference filters bonded to the disc, each passing one color band (typically R: 610-700 nm, G: 500-570 nm, B: 430-490 nm). Edge transition width must be under 15 nm to keep colors saturated. Thickness is usually 1.1 mm Schott B270 glass.
- Aluminium or Magnesium Hub: Carries the segments and bolts to the motor shaft. Concentricity to the bore must be held inside 25 µm or the disc walks at speed and beats the bearings. Magnesium is preferred for high-RPM wheels because it dampens resonance better than aluminium.
- Brushless DC Motor: Spins the wheel at 7,200-14,400 RPM with speed regulation tighter than ±0.1%. Lifetime is usually 20,000-30,000 hours of continuous duty. Motor cogging torque must be low or you'll see 12-pole-frequency banding in the projected image.
- Index Sensor: A Hall-effect or optical sensor that fires once per revolution to give the DMD controller a phase reference. Pulse jitter must stay below 10 µs to keep the spoke-blanking interval aligned. A second sensor on some wheels watches segment edges directly for finer phase lock.
- Spoke-Light Recovery (optional): Some wheels add white, yellow, or cyan segments to boost lumens on bright scenes — the BrilliantColor approach Texas Instruments introduced around 2005. Extra segments raise peak luminance by 30-50% but desaturate primaries, which is why home-cinema purists prefer RGBRGB six-segment wheels.
Who Uses the Color Wheel (optics)
Color wheels show up wherever you need to time-sequence colors or narrowband filters through a single optical path. The economic logic is simple: one detector or modulator plus one wheel costs less than three of each, and registration is automatic because everything shares the same lens.
- Home & Business Projection: Single-chip DLP projectors — BenQ TK700STi, Optoma UHD35, and the original Texas Instruments DMD reference designs from 1996 onward all use 6-segment RGBCYW wheels at 7,200 RPM.
- Cinema Projection: Older Christie and Barco DLP cinema projectors used phosphor-and-color wheels in laser-phosphor light engines to convert blue laser light into broadband white before color separation.
- Astronomy: ZWO and FLI filter wheels carry 5-9 narrowband filters (Hα 656.3 nm, OIII 500.7 nm, SII 672.4 nm) in front of a cooled CMOS sensor for deep-sky imaging — same mechanism, just stepped instead of continuously spinning.
- Stage & Architectural Lighting: Martin MAC Aura and Robe Robin moving-head fixtures use motorised color wheels with dichroic segments to swap colors mid-cue without the energy loss of subtractive gel.
- Microscopy & Fluorescence Imaging: Sutter Lambda 10-3 filter wheels position excitation filters in front of a mercury or LED source for live-cell fluorescence work, switching channels in 40-100 ms.
- Spectrophotometry: Older HACH and Thermo Scientific benchtop colorimeters used a low-speed color wheel with discrete bandpass filters to step through wavelengths in front of a single photodiode.
The Formula Behind the Color Wheel (optics)
The number that matters most is segment dwell time — the microseconds the beam spends inside one color segment per revolution. This sets how many gray-level bits the DMD can resolve and whether the rainbow effect becomes visible. At the low end of the typical range (3,600 RPM, 1× wheel) you get long dwell and clean bit depth but bad rainbow artifacts. At the high end (14,400 RPM, 6× wheel) the rainbow disappears for almost everyone but dwell drops below 1 ms and gray-level shaping gets aggressive. The sweet spot for consumer DLP sits at 7,200 RPM with a 6-segment wheel, giving roughly 1.4 ms per segment.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| tdwell | Time the beam spends in one segment per revolution | seconds (s) | seconds (s) |
| NRPM | Wheel rotational speed | revolutions per minute | revolutions per minute |
| θseg | Angular width of the segment | degrees (°) | degrees (°) |
| fcycle | Color refresh rate (full RGB cycles per second) | hertz (Hz) | hertz (Hz) |
Worked Example: Color Wheel (optics) in a 4K home-cinema DLP projector
A custom installer in Calgary is specifying the color wheel for a prototype 4K DLP projector running 60 Hz HDR content. The wheel carries 6 segments — RGBRGB — each occupying 60° of arc. The motor runs at 7,200 RPM nominal. The installer needs to know segment dwell time at nominal and at the low and high ends of the realistic operating range, then check whether the DMD has enough time to resolve 10-bit gray levels.
Given
- NRPM,nom = 7200 RPM
- θseg = 60 degrees
- Segments = 6 RGBRGB
- Frame rate = 60 Hz
Solution
Step 1 — convert nominal RPM to revolution period:
Step 2 — segment dwell at nominal 7,200 RPM with 60° segments:
That is the sweet spot for consumer DLP. 1.39 ms gives the DMD enough time to PWM through all 1,024 levels of a 10-bit channel using a roughly 1.3 µs minimum bit time, and the wheel cycles each color twice per 16.7 ms frame, which kills the rainbow effect for the vast majority of viewers.
Step 3 — at the low end of typical operation, 3,600 RPM (a 1× wheel design from the late 1990s):
Long dwell, plenty of bit depth, but each color shows once per frame and the rainbow effect is brutal during fast eye movements. This is why nobody ships 1× wheels anymore.
Step 4 — at the high end, 14,400 RPM with the same 60° segments (a 4× wheel):
Rainbow effect is gone for nearly everyone, but dwell drops below 1 ms and the DMD struggles to resolve full 10-bit gray smoothly — you start seeing posterisation in dark scenes unless the controller uses dithering. Bearing life also halves roughly each time you double RPM, so 14,400 RPM wheels typically rate 15,000 hours versus 30,000 hours at 7,200 RPM.
Result
Nominal segment dwell is 1. 39 ms at 7,200 RPM with 60° segments. In practice that means the DMD has enough headroom to render a clean 10-bit grayscale ramp without visible posterisation, and the rainbow effect drops below the perceptual threshold for roughly 95% of viewers. Compare that to 2.78 ms at 3,600 RPM (great bit depth, terrible rainbow) and 0.69 ms at 14,400 RPM (no rainbow but visibly crushed shadow detail) and you can see why 7,200 RPM became the industry default. If your measured dwell is shorter than calculated, the most likely causes are: (1) motor speed regulation drifting high under thermal load — check tach output against the index pulse, (2) segment angle error above 0.5° from sloppy bonding at the hub, which steals dwell from one color and gives it to another, or (3) spoke-blanking interval set too wide in the controller firmware, which the user perceives as shortened color time even though the wheel itself is fine.
Choosing the Color Wheel (optics): Pros and Cons
A spinning color wheel is one of three ways to get full-color output from a single imager. The other two are three-chip systems with permanent dichroic prisms and LCoS panels with field-sequential LED illumination. Each trades cost, color volume, and reliability differently.
| Property | Color Wheel (single-chip DLP) | Three-Chip DLP with Prism | LED Field-Sequential LCoS |
|---|---|---|---|
| Typical operating speed | 7,200-14,400 RPM mechanical | Static — no moving optics | 0 RPM, LED switching at 180-360 Hz |
| Color refresh rate | 180-360 Hz (depends on wheel multiplier) | Continuous, all 3 colors simultaneous | 180-360 Hz electronic |
| Rainbow artifact susceptibility | Low at 6× and above, severe at 1× | None — colors are simultaneous | Low — limited by LED switching speed |
| Bit depth ceiling | 10-bit at 7,200 RPM, drops at 14,400 RPM | 12-bit straightforward | 10-12 bit |
| Cost (light engine, USD) | $80-300 for consumer wheels | $2,000-15,000 prism block | $400-1,500 LED engine |
| Service life | 20,000-30,000 hours wheel motor MTBF | >50,000 hours, no wear parts | >50,000 hours LED-limited |
| Audible noise | Wheel whine at 120-240 Hz fundamental | Silent optics | Silent optics |
| Best application fit | Consumer & business projection, $500-5,000 price tier | Cinema, professional venue, $20k+ projectors | Compact pico projectors and high-end home cinema |
Frequently Asked Questions About Color Wheel (optics)
Because dwell time per segment shrinks as you add segments at constant RPM. A 6-segment wheel at 7,200 RPM gives 1.39 ms per color, while a 4-segment wheel at the same speed gives 2.08 ms. The DMD modulates gray by pulse-width-modulating mirror flips, and the shortest reliable bit time on a typical Texas Instruments DMD sits around 1.3 µs. Shorter dwell means fewer PWM bits available before you hit that floor, so the controller has to dither the bottom 1-2 bits. Dither is invisible in midtones and highlights but visible as faint moving texture in dark scenes.
If shadow detail matters more than rainbow rejection — say you're watching mostly cinematic content with a steady gaze — a 4-segment RGBW wheel can actually look cleaner.
That's almost always a phase-lock problem between the wheel index pulse and the DMD frame buffer, not a wheel defect. White text demands all three primaries at full intensity, which means the controller has to drive the DMD hard during every segment. If the index pulse jitters by more than about 10 µs — typical when a Hall sensor magnet has loosened or the bearing is worn — the spoke-blanking window walks across the segment boundary frame to frame, and you see chromatic shimmer on high-contrast edges.
Quick check: scope the index pulse against vsync. Anything above 15 µs jitter on a wheel rated 7,200 RPM means the bearing or sensor is going. Full-color images mask the problem because no single primary is at peak.
Decide by exposure time. If your exposures are shorter than about 100 ms — fluorescence imaging of live calcium transients, for instance — a continuously spinning wheel synchronised to the camera works well because the time lost during segment transitions is negligible. If exposures are longer than a second, which is normal for narrowband astrophotography (Hα subs of 300 s are routine), use a stepped filter wheel. Continuous spin would average colors across the exposure and destroy your spectral selectivity. Stepped wheels also let you carry 5-9 different narrowband filters that don't need to be in any particular angular sequence.
The ZWO EFW series and FLI Atlas are the standard stepped wheels for that reason.
Yes, schedule replacement now, because the whine is the bearing telling you it's drying out. The first audible signature is usually a 4-6 kHz tone that wasn't there when the projector was new. The image stays clean for another few hundred hours because the motor controller compensates for rising bearing drag by drawing more current. Once it can't compensate anymore — typically when current draw exceeds the controller's limit — the wheel slows below its phase-lock window and you get sudden color breakup, often mid-movie.
Wheel motors are usually $40-120 in parts and the swap takes an hour. Waiting until failure means you risk a full optical engine teardown if the wheel fragments at speed.
Thermal expansion of the segment bonding adhesive shifts segment angles by a fraction of a degree, and the dichroic coatings themselves drift slightly in their cutoff wavelengths with temperature. Both effects widen the spoke-transition zone, which the controller has to blank — and blanking time effectively subtracts from each color's perceived duration. The eye reads that as a shorter color cycle, which is exactly the condition that produces rainbows.
If the projector cools down and the rainbow goes away, the wheel itself is probably fine. If the rainbow persists cold, the segment bonding is permanently shifted and the wheel needs replacement. Most consumer projectors don't have enough airflow over the wheel to keep it cooler than about 45°C in a sealed cabinet — pulling it out of the cabinet often fixes the symptom.
You can, and several research groups have — particularly for hyperspectral and multi-primary display work — but the controller firmware has to know about the segment count to align the PWM scheduler. With closed Texas Instruments DLP controllers you'll need to use a development board like the DLPC900 EVM where segment count and timing are exposed. The mechanical wheel build itself is straightforward at any prime number of segments as long as total angle sums to 360° and you balance the disc to under 0.5 g·mm at the rim or it'll vibrate at speed.
One trap: avoid segment counts that share a common factor with the frame rate. A 6-segment wheel at 60 Hz and 7,200 RPM works because 6 divides cleanly. A 7-segment wheel at 60 Hz needs the wheel running at 8,400 RPM (or some multiple) to keep an integer number of full color cycles per frame.
References & Further Reading
- Wikipedia contributors. Color wheel (optics). Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
