Your Cheap Projector has 1 BIG Problem!
video description
Date: 2025-04-07
Related videos
Comments and reviews: 20
RegisMichelLeclerc
I heard that 10dB was half the noise, so I jumped to Wikipedia to check, they're still about accurate when it comes to physics. Here's what it says: A change in power ratio by a factor of 10 corresponds to a change in level of 10 dB. A change in power ratio by a factor of 2 or 1/2 is approximately a change of 3 dB. Checked, 3dB lower is half the noise. I'm not great at maths, but I kinda understood that the scale is not only log of power, it's square root of log of power. The good thing is that it makes your job that much a bigger feat: 10dB is huge!
Another source of noise is the DLP itself, but I'm not sure how it can be tackled. On my very old projector, it's actually most of the noise (some rattling. There is a step motor inside the projector, setting a colour change 75 times per second (R, G, B alternating 25 times per second, maybe faster) by spinning a wheel, which forced me to setup the room so the sofa was surrounded by the 5. 1 and the projector was at the complete back of the room, there would be no way of sitting anywhere close to the projector. Do you think there might be any way to reduce the noise from the DLP, or has it been taken care of these last 10 years
A noise of 57 decibels sounds a bit high to me, especially that the scales don't add up: mathematically speaking, a noise of 22dB peak added to a noise of 22dB peak makes, at most, a noise of 25dB peak for a constant noise, which it isn't, one noise may compensate the other, there's more than the fans to reach 57dB; there may be a complicated idea of noise cancelling there, if you can grab the DLP's wheel signal, say with a piezo loudspeaker emitting the opposite noise.
reply
I heard that 10dB was half the noise, so I jumped to Wikipedia to check, they're still about accurate when it comes to physics. Here's what it says: A change in power ratio by a factor of 10 corresponds to a change in level of 10 dB. A change in power ratio by a factor of 2 or 1/2 is approximately a change of 3 dB. Checked, 3dB lower is half the noise. I'm not great at maths, but I kinda understood that the scale is not only log of power, it's square root of log of power. The good thing is that it makes your job that much a bigger feat: 10dB is huge!
Another source of noise is the DLP itself, but I'm not sure how it can be tackled. On my very old projector, it's actually most of the noise (some rattling. There is a step motor inside the projector, setting a colour change 75 times per second (R, G, B alternating 25 times per second, maybe faster) by spinning a wheel, which forced me to setup the room so the sofa was surrounded by the 5. 1 and the projector was at the complete back of the room, there would be no way of sitting anywhere close to the projector. Do you think there might be any way to reduce the noise from the DLP, or has it been taken care of these last 10 years
A noise of 57 decibels sounds a bit high to me, especially that the scales don't add up: mathematically speaking, a noise of 22dB peak added to a noise of 22dB peak makes, at most, a noise of 25dB peak for a constant noise, which it isn't, one noise may compensate the other, there's more than the fans to reach 57dB; there may be a complicated idea of noise cancelling there, if you can grab the DLP's wheel signal, say with a piezo loudspeaker emitting the opposite noise.
reply
INUID
It looks like you could simply attach an additional (standard) CPU cooler to the backside (the exterior side) of the aluminum plate (where the LED is mounted, thereby extending the cooling volume outside the case. Just like some do with mini PCs (e. g, Intel NUCs. I’d definitely take cues from those mini PC solutions since they’ve been grappling with this very issue for years. Cooling the LED (was it 10W or 20W of waste heat) completely passively is relatively straightforward. You can 3D print a fanductor even design your own side panel with modifications for the external supplemental cooling. Heatpipes are available in all shapes and sizes/lengths, and you can already get large CPU coolers for a very reasonable price. And by the way, has anyone mentioned the Noctua brand in the comments yet
The potentially problematic parts are those that MUST be cooled by the internal airflow. But when it comes to projectors, I have no idea.
Anyway, I’m really looking forward to a follow-up videothose projectors must be somehow entirely passively (or at least ULTRA quietly) coolable.
reply
It looks like you could simply attach an additional (standard) CPU cooler to the backside (the exterior side) of the aluminum plate (where the LED is mounted, thereby extending the cooling volume outside the case. Just like some do with mini PCs (e. g, Intel NUCs. I’d definitely take cues from those mini PC solutions since they’ve been grappling with this very issue for years. Cooling the LED (was it 10W or 20W of waste heat) completely passively is relatively straightforward. You can 3D print a fanductor even design your own side panel with modifications for the external supplemental cooling. Heatpipes are available in all shapes and sizes/lengths, and you can already get large CPU coolers for a very reasonable price. And by the way, has anyone mentioned the Noctua brand in the comments yet
The potentially problematic parts are those that MUST be cooled by the internal airflow. But when it comes to projectors, I have no idea.
Anyway, I’m really looking forward to a follow-up videothose projectors must be somehow entirely passively (or at least ULTRA quietly) coolable.
reply
oasntet
- less thermal paste is better, not more. thermal paste conducts heat worse than aluminimum or copper; lap both surfaces perfectly flat and use a miniscule amount of paste, you'll get better results.
- there are plenty of replacement fans that spin just as fast but are far less noisy; the Noctua line are known for it, and it's down to high quality bearings and vortex disruptors on the fan blades.
- the best way to reduce heat is to reduce power to the LED. the light dropoff is not linear - cut power by 50% and you'll usually lose only about 25% of the light.
Personally, I'd 3d-print a fan shroud that clips to the outside of the fan grill, goes through several bends with internal features that break up the sound waves, and ends in another bigger and slower fan to make up for the lost air velocity. You can make a projector significantly quieter by putting it in a larger box with a snaking air path, so long as you help the air along with a second quiet fan.
reply
- less thermal paste is better, not more. thermal paste conducts heat worse than aluminimum or copper; lap both surfaces perfectly flat and use a miniscule amount of paste, you'll get better results.
- there are plenty of replacement fans that spin just as fast but are far less noisy; the Noctua line are known for it, and it's down to high quality bearings and vortex disruptors on the fan blades.
- the best way to reduce heat is to reduce power to the LED. the light dropoff is not linear - cut power by 50% and you'll usually lose only about 25% of the light.
Personally, I'd 3d-print a fan shroud that clips to the outside of the fan grill, goes through several bends with internal features that break up the sound waves, and ends in another bigger and slower fan to make up for the lost air velocity. You can make a projector significantly quieter by putting it in a larger box with a snaking air path, so long as you help the air along with a second quiet fan.
reply
DelticEngine
For an application like this, where I don't want to be opening it up again, I'd consider a graphene pad instead of thermal compound.
Depending on the method of conduction, it may be worth trying some sort of thermal insulation technique between the LED and the light processing part of the projector. This is assuming that it is the LED COB module that is a significant source of heat in the optical chamber. If it is the LCD itself that needs cooling then you may be able to reduce the fan speed if the airflow dynamics were improved.
I you don't mind a a more significant modification and a 'frankenstein' solution you could try a CPU cooler or maybe even a GPU cooler on the LED COB module. I may be a bulky solution, and maybe a not the prettiest, if it resides outside the case put it may even be possible to run such a cooling solution passively.
reply
For an application like this, where I don't want to be opening it up again, I'd consider a graphene pad instead of thermal compound.
Depending on the method of conduction, it may be worth trying some sort of thermal insulation technique between the LED and the light processing part of the projector. This is assuming that it is the LED COB module that is a significant source of heat in the optical chamber. If it is the LCD itself that needs cooling then you may be able to reduce the fan speed if the airflow dynamics were improved.
I you don't mind a a more significant modification and a 'frankenstein' solution you could try a CPU cooler or maybe even a GPU cooler on the LED COB module. I may be a bulky solution, and maybe a not the prettiest, if it resides outside the case put it may even be possible to run such a cooling solution passively.
reply
nickbrockelman
6: 53 using a cutoff wheel like the one you have in your high-speed grinder is probably one of the worst ways to cut aluminum. Unless you get a cut off wheel specifically designed to cut aluminum all is really going to do is make a nasty cut off line and eventually separate your parts. If I were you and I were cutting aluminum like that I would switch to either a hacksaw or use a cutoff wheel that is designed for non-ferrous metals. Frankly aluminum is soft enough that with the right blade in a circular saw you can use it as well. And if you are going to grind on aluminum you definitely for sure need a non-ferrous metal grinding wheel or you're just going to gum up your ferrous metal wheels and make them nearly unusable
reply
6: 53 using a cutoff wheel like the one you have in your high-speed grinder is probably one of the worst ways to cut aluminum. Unless you get a cut off wheel specifically designed to cut aluminum all is really going to do is make a nasty cut off line and eventually separate your parts. If I were you and I were cutting aluminum like that I would switch to either a hacksaw or use a cutoff wheel that is designed for non-ferrous metals. Frankly aluminum is soft enough that with the right blade in a circular saw you can use it as well. And if you are going to grind on aluminum you definitely for sure need a non-ferrous metal grinding wheel or you're just going to gum up your ferrous metal wheels and make them nearly unusable
reply
donaldcarter1299
Last 3 cheaper sub 300$ projectors i got all had yellow spots after a few months of use. Even with loud fans. The heat on the lcd screen, just slowly killed em. And replacing the burned screen was as much as replacing the whole unit.
Its like they made them to last just long enough to avoid returns.
Personally i would rather a larger projector like my old one with a replaceable bulb the size of a small suitcase. But built with newer led/laser and fans and lower price point. It doesn't need to fit into a tiny box. Allowing for more room for fans and airflow etc. But finding a quality fair builds of anything is nuts nowadays. Quantity over Quality
reply
Last 3 cheaper sub 300$ projectors i got all had yellow spots after a few months of use. Even with loud fans. The heat on the lcd screen, just slowly killed em. And replacing the burned screen was as much as replacing the whole unit.
Its like they made them to last just long enough to avoid returns.
Personally i would rather a larger projector like my old one with a replaceable bulb the size of a small suitcase. But built with newer led/laser and fans and lower price point. It doesn't need to fit into a tiny box. Allowing for more room for fans and airflow etc. But finding a quality fair builds of anything is nuts nowadays. Quantity over Quality
reply
debrainwasher
To run LED-projectors hotter, than intended, is not a good idea. Since LED's are components, obeying the laws of physics (and chemistry, the so called RGT-Regel (Reaktionsgeschwindigkeit-Temperatur, also termed as van-’t-Hoff’s-rule (a subset off the Arrhenius equations) accounts as well. That means, linear degradation of the LED and phosphorus will increase rapidly upon some Kelvin increase of temperature. As a rule of thumb, mean lifetime of power LED's are halved by 5K increase of temperature. The perfect way to silence projector light sources is to use water-cooling. The secondary heat exchanger can be placed everywhere.
reply
To run LED-projectors hotter, than intended, is not a good idea. Since LED's are components, obeying the laws of physics (and chemistry, the so called RGT-Regel (Reaktionsgeschwindigkeit-Temperatur, also termed as van-’t-Hoff’s-rule (a subset off the Arrhenius equations) accounts as well. That means, linear degradation of the LED and phosphorus will increase rapidly upon some Kelvin increase of temperature. As a rule of thumb, mean lifetime of power LED's are halved by 5K increase of temperature. The perfect way to silence projector light sources is to use water-cooling. The secondary heat exchanger can be placed everywhere.
reply
rpals5412
I REALLY don't recommend doing this! (no offense) But;
1: I owned a same price range wimius projector, also loud, and after 2 years, it eventually burned the LCD panel, and has a brown spot now. (no mod)
2: Your temperature testing is not realistic, they best day scenario tests. In real life, people mount a projector under an isolated ceiling. one warm summer day, projector is on for hours and heat is accumulating, the surrounding air will be warm, and the internal heat in your projecter will reach unhealthy temperature for sure.
If you must mod for less noise, go with larger heat sink and more expensive quiet fans
reply
I REALLY don't recommend doing this! (no offense) But;
1: I owned a same price range wimius projector, also loud, and after 2 years, it eventually burned the LCD panel, and has a brown spot now. (no mod)
2: Your temperature testing is not realistic, they best day scenario tests. In real life, people mount a projector under an isolated ceiling. one warm summer day, projector is on for hours and heat is accumulating, the surrounding air will be warm, and the internal heat in your projecter will reach unhealthy temperature for sure.
If you must mod for less noise, go with larger heat sink and more expensive quiet fans
reply
freaky2xd
I remember doing something similar a couple of years ago with a projector. Completely different projector, exact same cooling system. I also replaced the 80mm fan with a slightly weaker but more quiet one and then noticed that the other fan that cools the inside of the projector is way louder. However, it is a very important fan, if airflow is too weak the mirrors will make the display heat up more and more until it gets burn marks, don't ask me how I know xD
One would have to design a completely new case for the components with a more optimized airflow path that allows the use of high quality fans like from Noctua.
reply
I remember doing something similar a couple of years ago with a projector. Completely different projector, exact same cooling system. I also replaced the 80mm fan with a slightly weaker but more quiet one and then noticed that the other fan that cools the inside of the projector is way louder. However, it is a very important fan, if airflow is too weak the mirrors will make the display heat up more and more until it gets burn marks, don't ask me how I know xD
One would have to design a completely new case for the components with a more optimized airflow path that allows the use of high quality fans like from Noctua.
reply
TankR
Knock a hole in the wall, set it back in its own enclosure with baffled air flow, encapsulate it away from the viewing room. At my exs apartment, there was a closet in the side room directly behind the couch. I really wanted to pop a round hole in the wall, plop a projector on the closet shelf shining through the hole, and use a picture frame with an image of like, an old steam train (and projector shining from the light, or a camera, or. uh projector (film of course, shining through the obvious bit) to mask the hole. Of course the landlord (read soulless property management company) would have hated it.
reply
Knock a hole in the wall, set it back in its own enclosure with baffled air flow, encapsulate it away from the viewing room. At my exs apartment, there was a closet in the side room directly behind the couch. I really wanted to pop a round hole in the wall, plop a projector on the closet shelf shining through the hole, and use a picture frame with an image of like, an old steam train (and projector shining from the light, or a camera, or. uh projector (film of course, shining through the obvious bit) to mask the hole. Of course the landlord (read soulless property management company) would have hated it.
reply
user-dj1hy6zc6q
Dry thermal paste is purposely dry. As you saw, you gained little performance difference by replacing it, but in time you will see that longevity suffers. Wet thermal paste performs slightly better, but pumps out over time (thermal cycles cause thermal expansion, contraction, and migration of paste, and voids will appear. Then the thermal performance will nosedive.
Manufacturers use dry paste on purpose for longer longevity and to reduce warranty claims and complaints.
Almost every time you see someone complain about dry thermal paste, it is a mistake to complain.
reply
Dry thermal paste is purposely dry. As you saw, you gained little performance difference by replacing it, but in time you will see that longevity suffers. Wet thermal paste performs slightly better, but pumps out over time (thermal cycles cause thermal expansion, contraction, and migration of paste, and voids will appear. Then the thermal performance will nosedive.
Manufacturers use dry paste on purpose for longer longevity and to reduce warranty claims and complaints.
Almost every time you see someone complain about dry thermal paste, it is a mistake to complain.
reply
AdricM
might be able to better like the sound with some simple baffles. directing much of the sound away. from you, something as simple as duct tape or PVA foam could be mounted externally to send the sound from your direction. or bounce it down to your floor and a carpet. or away from you and towards something sound absorbing. Another option is sometimes removing the grilles or making more minimal ones or ones that can also conduct more heat out. can allow more and quieter airflow, you can also drop fan speeds but putting a diode inline. the forward drop can drop a volt or two.
reply
might be able to better like the sound with some simple baffles. directing much of the sound away. from you, something as simple as duct tape or PVA foam could be mounted externally to send the sound from your direction. or bounce it down to your floor and a carpet. or away from you and towards something sound absorbing. Another option is sometimes removing the grilles or making more minimal ones or ones that can also conduct more heat out. can allow more and quieter airflow, you can also drop fan speeds but putting a diode inline. the forward drop can drop a volt or two.
reply
Del_UK
The 80mm fan. A static pressure version would be a better choice, as you cooling fins on a heat sink.
The centrifugal fan. I not sure of the size from video. With a slight help from a 3D printer, 1 (or 2) x 40mm air flow fans. Make/print a 90 degree angle housing and place 1 fan at either bottom or top section or 1 fan at bottom and 1 fan a the top. As you said in video, this was just for air flow in that area,
As others have mentioned, Noctua fans. I would also like to add Arctic Cooling and Thermalright into the mix.
Great video and problem solving.
reply
The 80mm fan. A static pressure version would be a better choice, as you cooling fins on a heat sink.
The centrifugal fan. I not sure of the size from video. With a slight help from a 3D printer, 1 (or 2) x 40mm air flow fans. Make/print a 90 degree angle housing and place 1 fan at either bottom or top section or 1 fan at bottom and 1 fan a the top. As you said in video, this was just for air flow in that area,
As others have mentioned, Noctua fans. I would also like to add Arctic Cooling and Thermalright into the mix.
Great video and problem solving.
reply
greatscott
Just for you information there are great 80mm beQuiet fans as well. As a general rule of thumb: 80mm, 120mm and 140mm are most easily sourced from a PC parts vendor instead of a general electronics parts vendor. I also used a cool 3D printed adapter once to get a good 80mm fan out of an old Intel stock cooler (basically just needed to cut a few pieces and glue the spokes into a replacement ring. Since I got the PC for free, it was the cheapest way to make my secondary Ender 3 more quiet. Its PSU was loud af.
reply
Just for you information there are great 80mm beQuiet fans as well. As a general rule of thumb: 80mm, 120mm and 140mm are most easily sourced from a PC parts vendor instead of a general electronics parts vendor. I also used a cool 3D printed adapter once to get a good 80mm fan out of an old Intel stock cooler (basically just needed to cut a few pieces and glue the spokes into a replacement ring. Since I got the PC for free, it was the cheapest way to make my secondary Ender 3 more quiet. Its PSU was loud af.
reply
phinok. m. 628
7: 43 Assuming the fan runs at a fixed speed. Cooler heat sink = less power transferred to the heat sink = worse thermal conduction = hotter LED.
So basically you made the thermal conductivity a little worse. Though one degree variation could also be within the margin of error. Either way, you want the heat sink to draw more heat from the LED, not less. Ideally you would have measured the temperature of the LEDs, themselves. Not easy to do of course. Maybe from the front.
reply
7: 43 Assuming the fan runs at a fixed speed. Cooler heat sink = less power transferred to the heat sink = worse thermal conduction = hotter LED.
So basically you made the thermal conductivity a little worse. Though one degree variation could also be within the margin of error. Either way, you want the heat sink to draw more heat from the LED, not less. Ideally you would have measured the temperature of the LEDs, themselves. Not easy to do of course. Maybe from the front.
reply
fmphotooffice5513
Nice design. I wonder if you might have gotten similar sound results and maintained the unit's lifetime by building a sound enclosure with its own cooling fans, etc. using inexpensive fire brick, vermiculite, portland cement, etc.
I have also encountered heat paste application problems in devices. I had an inductive cooker fail because the thermal grease was only touching about 30% of the surface. The rest was dripping off one side. I'm guessing assembly line errors.
reply
Nice design. I wonder if you might have gotten similar sound results and maintained the unit's lifetime by building a sound enclosure with its own cooling fans, etc. using inexpensive fire brick, vermiculite, portland cement, etc.
I have also encountered heat paste application problems in devices. I had an inductive cooker fail because the thermal grease was only touching about 30% of the surface. The rest was dripping off one side. I'm guessing assembly line errors.
reply
greatscott
7: 32 Those heatsinks look to not be milled. The dry thermal paste is nano clay based. It is dry because of the nano clay which transfers heat thru itself even when there is a small gap, kind of like a very thin thermal pad. You may ant to check the temps again after a few days, if there was a gap the new thermal paste would have dried up and gotten much worse at transferring the heat. Check out shin etsu, it is a nano clay based thermal paste.
reply
7: 32 Those heatsinks look to not be milled. The dry thermal paste is nano clay based. It is dry because of the nano clay which transfers heat thru itself even when there is a small gap, kind of like a very thin thermal pad. You may ant to check the temps again after a few days, if there was a gap the new thermal paste would have dried up and gotten much worse at transferring the heat. Check out shin etsu, it is a nano clay based thermal paste.
reply
kornrat8284
I modified many stage lights and I have had great experience with 12v fan speed controllers from aliexpress. I use the black PCB ones that cost around 2 per piece including shipping if you buy multiple ones. With them the fan speed is dependent on the temperature of the heat sink and I don’t understand why such a cheap controller is not standard practice for LED cooling.
(Maybe those are a great aliexpress item for the next edition )
reply
I modified many stage lights and I have had great experience with 12v fan speed controllers from aliexpress. I use the black PCB ones that cost around 2 per piece including shipping if you buy multiple ones. With them the fan speed is dependent on the temperature of the heat sink and I don’t understand why such a cheap controller is not standard practice for LED cooling.
(Maybe those are a great aliexpress item for the next edition )
reply
el3dprint
To add, you can also use snoozing-chambers around the air output to silence it a lot probably.
And also, keep in mind that it is now 'the cool season' around here. I would certainly think about a temperature-regulated fanspeed which can probably be made even with some discrete components like one or 2 thermistors, one transistor, one resistor and a pot to adjust it to go slower basically and start running faster if it heats up.
reply
To add, you can also use snoozing-chambers around the air output to silence it a lot probably.
And also, keep in mind that it is now 'the cool season' around here. I would certainly think about a temperature-regulated fanspeed which can probably be made even with some discrete components like one or 2 thermistors, one transistor, one resistor and a pot to adjust it to go slower basically and start running faster if it heats up.
reply
florihae
So sad you only used a generic fan and not some real nice Noctua or Arctic ones: (
I am sure you could have reached quite a bit better results with them.
Maybe in another video
And then you could also include either a manual fan speed regulator, or even better a temperature control. Measure the LED temp and set the fan speed to it!
Or you fulfill our dreams and watercool it
reply
So sad you only used a generic fan and not some real nice Noctua or Arctic ones: (
I am sure you could have reached quite a bit better results with them.
Maybe in another video
And then you could also include either a manual fan speed regulator, or even better a temperature control. Measure the LED temp and set the fan speed to it!
Or you fulfill our dreams and watercool it
reply
Add a review, comment
Other channel videos
