More Fans Did Not Fix the Projection System
An architectural LED projector packs a light source, power electronics and finned heat sinks into one compact housing, and the first reaction to a hot unit is often a second fan or a larger free-air airflow number. The installed result did not improve, because nobody changed where the air enters, how far the fans sit from the wall, or what the discharge does.
That pattern repeats across projector and large-format lighting programs. Two fans are fitted where one was fitted before, the housing is drilled a second time, and the measured LED temperature barely moves while the unit gets louder. The reason is physical, not mysterious: a fan delivers what the air path around it allows. A fan pressed against a panel, or sharing its inlet with a neighbour's discharge, cannot produce the flow its curve promises, no matter how many of them are installed.
The figure below shows one housing, two installations. The fans, the heatsink and the LED module are identical on both sides. On the left, both fans sit close to the inlet wall with no gap to speak of, and the warm exhaust loops straight back into the lower inlet. On the right, the same fans sit farther from the wall, spaced apart, with a defined transition into the heatsink. The difference in delivered cooling comes from geometry alone.
What the LED Junction Demands From the Air Path
The reason layout matters so much in a projector is the customer at the end of the heat path: an LED junction whose light output and lifetime both fall as its temperature rises.
High-power LED datasheets rate luminous flux at a fixed junction temperature, commonly 85 degrees Celsius for the Cree XLamp class of emitters, and the same datasheets derate the current the junction may carry as temperature climbs above that point. An optical engine that runs hotter than its rating does not fail loudly; it dims quietly, shifts colour, and reaches its end of life - the point where output has fallen to 70 percent of its initial value - far earlier than the hours printed on the box.
Lifetime data from LED vendors follows a consistent pattern: for a given emitter family, each additional ten degrees at the junction roughly doubles the speed of lumen depreciation. That is why a ten-degree error in the air path is not a ten-percent error in the product. It is the difference between a projector that holds its brightness specification for its rated life and one that visibly fades within the warranty period.
The air path is the last link in the thermal chain. Heat leaves the junction through the solder point and the board into a heatsink, and from the heatsink fins into moving air. Published guidance for unmanaged lighting assemblies puts the solder-point-to-ambient thermal resistance in the range of 15 to 25 degrees Celsius per watt before any deliberate thermal design; the fins and the fan exist to cut that number. Every flow shortcut or recirculation in front of the fins leaves the junction hotter, and nothing downstream of the LED can recover it.
Why Layout Decides Before the Fan Model Does
Fan catalogue curves are measured in laboratory conditions: a straight duct, open space in front of the inlet, a controlled discharge. A projector housing is none of those things, and the gap between the two is what the air moving industry calls system effect.
The Air Movement and Control Association documents the pattern across fan types: obstructions, elbows and abrupt transitions placed near a fan lower its delivered performance below the catalogue curve, and the penalty grows with the severity of the disturbance. Two identical fans in two different positions in the same housing can differ in delivered airflow by a wide margin at the same speed and input power. The fan did not change; the air path did.
This is why adding a second fan so often disappoints. In a projector, the second fan usually goes next to the first, in the same cramped inlet zone, sharing the same preheated air and the same sharp turn into the heatsink. Both fans now operate deeper into their disturbed-inlet region than the single fan did. Total flow may rise a little; flow that actually crosses the LED heatsink may not rise at all, because more of it short-circuits through the path of least resistance.
The corrective sequence is to review the layout first and the fan selection second. Our guide to system effect and installation clearance sets out the clearance distances and transition shapes that let a fan deliver close to its catalogue curve, and the same rules apply inside a projector housing at small scale.
The Wall Behind the Fan Is a Real Cost
The most common layout fault in compact projectors is a fan mounted with its inlet face close to a wall, a panel or a neighbouring unit. The inlet looks open; the velocity field is not.
Sanyo Denki's technical report on its 60-millimetre San Ace fans quantifies the effect directly. Two fan thicknesses were compared inside the same 25-millimetre installation space with a wall on the inlet side: the thicker fan left a 10-millimetre gap to the wall, the thinner one left 15 millimetres. In open air the thicker fan moved more air. Installed against the wall, both fans delivered the same maximum airflow of 0.58 cubic metres per minute - the wall gap, not the fan, set the limit - and the installation with more clearance ran 2 dB(A) quieter at the same time.
The same report shows the practical consequence: against that wall, the thinner fan's higher static pressure curve gives it the better installed cooling performance once the load rises, even though it loses on open-air airflow. Selection against an installed condition overturned the open-air ranking. Any engineer who has watched a bench winner fail in the housing has seen this exact mechanism.
The rule that survives testing is simple to apply and easy to skip: state the distance between the fan inlet face and the nearest obstruction, and treat a small gap as a design constraint with a measurable cost - in flow, in sound, and in which fan curve actually wins.
Layout Choices Show Up in the Noise Spectrum
A projector is often specified for occupied spaces, which makes the acoustic consequence of layout as commercial as the thermal one - and the two share the same root cause.
A fan operating with a disturbed inlet sees uneven blade loading: each blade passes through a different velocity field once per revolution. The measured result is more broadband turbulence noise plus stronger tonal components at the blade passing frequency and its multiples, even though the speed command has not changed. Sanyo Denki's wall-gap comparison put the difference at 2 dB(A) in an installation no worse than a typical projector, from clearance alone.
Structure adds to air. Fans mounted close together on a shared thin panel couple vibration into the housing, and panels close to the fan can amplify tones at their own natural frequencies. A projector that seems loud 'for no reason' is usually radiating a tone that the layout, not the fan, selected.
Because the mechanism is layout, the diagnosis is acoustic measurement at the installed condition, not a louder fan swap or a firmware speed cap that gives away cooling authority. Our article on cooling fan noise, RPM and resonance walks through separating airborne tones from structure-borne vibration, and the medical equipment noise validation article shows the same discipline applied where a hard acoustic limit is written into the product standard.
Size the Airflow From the LED Heat Load
Layout decides how much of the flow reaches the fins; the heat balance decides how much flow that needs to be. The two numbers meet in one equation.
An LED converts roughly 30 to 40 percent of its electrical input into light; the rest is heat that must leave through the air path. A 200-watt optical engine therefore dumps on the order of 120 to 140 watts into the housing, plus the driver losses. The airflow needed to carry that heat away at a chosen temperature rise follows directly from the heat balance, and it is the first number to write down before any layout or fan discussion.
The rise you permit is a design choice with consequences. A small rise keeps the junction low and the lumen curve flat, but demands more airflow through a housing that is already tight. A generous rise saves flow and costs lifetime at the junction, at the rate described in the junction section above. State the rise, in degrees, next to the flow number - an airflow target without its temperature rise is not a specification.
One correction belongs in every projector calculation: the temperature of the air available at the inlet. If the exhaust can recirculate, or the inlet sits in a warm ceiling plenum or a sunlit façade, the air arriving at the fins is no longer room air, and the permitted rise shrinks by exactly that difference. Our guide on fan margin at high ambient treats this correction as a selection input rather than an afterthought.
A Layout Comparison You Can Run in One Afternoon
None of the evidence above requires a wind tunnel. Three configurations, one afternoon, controlled conditions - and the right layout decision usually becomes obvious in the data.
Keep the fan model, the voltage and the heat load fixed; change only the geometry, one factor at a time. Measure at thermal steady state, and record the acoustic level at a fixed distance from the inlet side, because flow gains that double the sound are rarely wins. The comparison below is the minimum set; the numbers in the table are the acceptance record, not decoration.
| Configuration | What changes | Record at acceptance |
|---|---|---|
| A - current layout | Baseline, nothing changed | LED case temperature, air temperature in versus out, sound level at 1 m, fan current |
| B - spacing added | Wall gap at the inlets and fan-to-fan spacing increased within the housing | The same four measurements, plus the inlet air temperature to detect recirculation change |
| C - transition added | A defined inlet or discharge transition fitted between fans and heatsink | The same four measurements, plus airflow direction check at the heatsink face |
A Validation Plan With Numbers Attached
A projector that passes the layout comparison still needs an acceptance record that will survive the next housing revision. Four measurements carry that weight.
First, LED case temperature at the point the emitter vendor specifies, at thermal steady state - this is the number the light output and lifetime actually follow. Second, air temperature in and out of the housing, whose difference times the flow gives the heat really being removed; when that number falls far short of the electrical input, the flow is recirculating or bypassing the fins. Third, fan current against the speed command, which rises when the fan works against a blocked or recirculating path. Fourth, sound pressure at a fixed point on the inlet axis.
A tachometer reading inside specification does not prove any of the above. RPM confirms the motor is commanded and turning; it says nothing about the flow crossing the fins, which the layout controls. Why normal RPM does not prove cooling airflow sets out the acceptance measurements that do. Repeat the four measurements at the worst ambient the product will ship into, because a layout that passes in a 22-degree laboratory can fail in a 40-degree conference room ceiling.
Where MEGA Tech Fits In
MEGA Tech builds the fans your projector layout has to work with - and we review the layout with you before the housing is cut, because the layout decides what any fan can deliver.
What we do: custom DC and EC axial fans and backward-curved centrifugal fans, selected against your installed operating point rather than a part number. For LED projection and architectural lighting that means we size the impeller and motor for your heatsink pressure drop at your worst inlet air temperature, we advise on inlet clearance, fan spacing and discharge transitions from the airflow side, and we put samples on your unit so the validation numbers above are measured in your housing, not ours.
What we need from you: the LED and driver heat load in watts, the permitted air temperature rise, the housing drawings with the inlet and outlet geometry, the acoustic target and its measurement distance, the voltage window and control interface, the ingress protection the product must hold, and the life assumption in hours.
Where our boundary sits: we do not supply heatsinks, LED modules, optical engines or the luminaire housing, and we do not certify photometric safety or electromagnetic compliance of the finished product. Those belong to the optical and electronics teams. And we will not tell you a fan is equivalent because its outline drawing matches; the installed operating point decides.
Engineer FAQ: LED Projector Fan Layout and Airflow
Short answers to the questions this layout review raises most often.
Q: Why can adding a second fan fail to cool the LED module?
A: Fans can compete for the same inlet air, short-circuit flow around the heatsink, or recirculate warm exhaust back into an inlet. Check the complete path from inlet to heat source to outlet before changing fan count.
Q: Is more airflow always better for a projector?
A: No. Flow that crosses the heatsink cools the LED; flow that bypasses it only adds noise. Past the point where the fins are supplied, extra airflow buys lifetime at a sharply diminishing rate and costs sound immediately.
Q: How much clearance does a fan inlet need from a wall?
A: There is no single number that fits every housing, but the Sanyo Denki comparison shows the cost of skipping the question: two wall gaps of 10 and 15 millimetres produced equal maximum airflow and a 2 dB(A) sound difference. Measure the installed performance at your actual gap instead of assuming the catalogue curve.
Q: Can we fix noise with a speed cap in firmware?
A: A cap trades cooling authority for sound without removing the cause. If the tone comes from a disturbed inlet or a vibrating panel, correcting the layout reduces sound while keeping the flow; the cap only hides the symptom until the summer installation demands the speed back.
Q: What single measurement best tracks projector cooling health in the field?
A: LED case temperature at the vendor-specified point, logged at steady state. It integrates everything upstream - inlet condition, recirculation, filter dust, fan wear - into the one number that follows light output and lifetime.
Questions for a Design Review
Bring measured numbers from the housing, not catalogue values from the datasheet.
- 1. What is the total heat in watts the air path must carry, LED plus driver, and at what permitted temperature rise?
- 2. What is the distance from each fan inlet face to the nearest obstruction, and was the fan curve measured or corrected for it?
- 3. What spacing separates the fan frames, and where does each discharge go relative to the neighbouring inlet?
- 4. What did the four acceptance measurements read for each configuration in the layout comparison, and which configuration won?
- 5. What is the worst inlet air temperature the product will see, including ceiling plenum air, sunload and recirculation?
- 6. What acoustic level did the installed unit measure at the specified distance, and which tones dominate the spectrum?
- 7. Who owns the housing geometry changes that the winning configuration requires, and by what date?
