The problem

Light leakage has always been something we've had to deal with in our projectors. There have been various ways of "coping" with the problem, but none have completely eliminated it. Here are the most common solutions:

1: Fan size -- Large fans can adequately cool a box at lower speeds, reducing noise, but increasing light leakage. Small fans need to spin faster, increasing noise, but decreasing light leakage.

2: Indirect cooling path -- The idea is to place some obstacle in cooling path (metal trim, air vent, etc) so that the bulb is not directly visible through the vent hole. So all of the light that leaves the case is indirect, and therefore, less intense.

3: Air filter -- This one is probably the least effective of the bunch. With an air filter, the amount of light blockage is directly proportional to the reduction in airflow.

All of the above have one glaring issue (pun intended): None of them eliminate light leakage. That's where my idea comes into play...

The Solution

The solution is to use a revolving door! Revolving doors are always sealed, so light can't escape. But if we revolve the door, air is free come and go, which is exactly what we want. A picture is worth a thousand words, so I've made a diagram to help you guys see what I'm talking about. This is a top-down view. Try to imagine a revolving door, when looking at this diagram.

Click to view attachment

As you can see, we start off with a normal projector with an intake and exhaust fan (they need to be next to each-other). The revolving door apparatus is retrofitted to the outside of the case. In real life, the apparatus would be covered on the top and bottom, so that it's completely sealed. The blade assembly is a single piece of plastic, or some other light material. The blade is freely-rotating in my diagram, but it can be powered, if need be.

Here's how it works: Fan A creates pressure, forcing the blades to rotate in the clockwise direction, and depositing air outside the case. Fan B reduces pressure, "sucking" the blades in the clockwise direction, and in the process, takes air from the outside. The blades assembly itself, might need to be powered, so as not to impede airflow, but this is completely possible.

Benefits to the revolving door setup:
(1) Essentially eliminates light leakage.
(2) If powered, it the revolving door won't impede cooling performance at all.
(3) We can use even bigger/slower fans, for the quietest possible operation (again, without worrying about light leakage).

Disadvantages
(1) This will increase the size of the projector, considerably.
(2) A bit complex to build.

You might have a problem sucking hot air back into the box if the inlet and outlet are too close together.



You could use 2 revolving gates. One on the inlet and one on the outlet. That way you could seperate them and avoid the problem.

Yeah, I thought about that. Certainly, there will be some recycling of air if the fans are this close --the question is how much?

I also have to consider what would happen if I removed either fan. If I removed the exhaust fan, for example, then a low-pressure zone would form where the exhaust fan used to be. Because the inlet fan consumes all of the air within each section, and then passes the "vacuumed-out" section to the left. The pressure won't be equalized until the vacuumed-out section passes back outside. In the meantime, the higher-pressure air outside, will want to push the gate backwards. I think the gate would work under these circumstances (if not, we could always "force" it in the right direction), but it's not as nice.

I think we could split the difference between these problems by using a larger-diameter gate. A larger gate would allow us to put some distance between the fans, whilst avoiding pressure anomalies.

What do you think?

There has been alot of thought put into the standard air flow design used by most people. I would hesitate to modify the basic set up in order to eliminate light leakage. Failure to ensure adequate air flow over the lcd, fresnels, and ballast could seriously affect video quality and maintenance costs. Having said that many people have successfully struck out on their own. If this is your first build however understanding why the most common airflow model works might save you a lot of wasted time and money.

It's generally good advice to keep your first projector as simple as possible. Good luck with what ever path you take...... JP

I think some clarification is in order. I'm not building a projector --I've already built a perfectly good 1920x1200 projector. In fact, I think it's worth emphasizing that anyone looking for build advice should stick with the "tried and true" standard design. My goal with this thread is to explore, debate, and experiment with new ideas. If, after that point, we're able to reach the consensus that this new idea is better than the current standard, then the new idea should become the new standard.

Having said that, you definitely made some good criticisms, which I wholeheartedly encourage. That's exactly the sort of stuff I'm looking for. I'm trying to get as much feedback on this as possible before going forward. That way, when I get around to building an actual prototype, I'll be able to do it smarter, with less wasted time, etc... Additionally, I'll be able to test out your ideas, my ideas (and, hopefully, other people's ideas) and get a better understanding of what works, and what doesn't. Then I'll post back with the results, get more feedback, and try again. Eventually, I'll either get it right, or concede failure.

Click to view attachment

The idea is great, I was just thinking to reduce recycling of the air you could use ducting. Since its a sealed climate control apparatus, the inlet and outlet can be anywhere. The Fans need be where they need to be however the ducting can make the air flow go where ever you want. You can still direct the airflow over the face of the lcd, and if you so choose the light engine then out. the chamber for the door could have 4 tubes ( think exaust crossover ) off the main cylinder.

now that I think of it ( after drawing this out) a powered "door" would basically BE the fan, the added exhaust fan may be needed to pressurize the exhaust side to make sure it blows out instead of riding around the "door".
With the "door" being powered it basically turns into a turbo impeller.


edit: another thing, if its not powered I was figuring the intake fan should be opposite of the exhaust fan, that way they both spin the blade with positive pressure.

I like the idea, and the theory is good. There are a lot of ways that this could be made to get acceptable performance. Make sure that the cold air intake is below the warm air exhaust, for example, to let convection help you out. An advantage is that the revolving door itself will also provide a muffler for the fan noise.

A few problems, as I see them.

1. There is inadequate separation of the hot and cold sides insid the projector case. This could be assisted with an additional wall betwen the fans, with another curved section touching the "doors"

2. We rarely need the cold air intake so close to the exhaust. This MIGHT work if the wall separating the fans is an extension of the collimator fresnel frame. That would have the "revolving door" on the top or side of the projector, though with this, you could not use the "convection" trick that I was talking about. That would reuire considerably more plumbing, increased complexity and other bad things.

3. The door itself would require power, and a fair amount.

Now, let's do a bit of math...

Let's say, for the sake of argumnt, that your "revolving door" is 160mm diameter, by 100mm width. You have 16 blades, separated by 22.5 degrees.

So we have a cylinder of air that we're dealing with. 160mm circle by 100mm shaft gives a volume of 2,186,189mm^3, or a little over 2L Each "chamber" cormed when 2 blades are touching the curved surface is just over 1/8L

A standard 80mm fan moves about 25 CFM, or about 708L. In order to match that airflow, our 2L chamber would need to rotate about 354 times, but I'd round that to about 400 times per minute.

400 RPM is doable, but I really wouldn't want to get my fingers in the way of those blades! Some sort of a safety cage would be mandatory for this setup.

You could also try using a side blower or squirel cage fan such as these:
Click to view attachmentClick to view attachment
My experience is that they will reduce light leakage but are often less efficient than standard venting fans

"Squirrel cage" fans are actually more efficient, but are sometimes louder. They also reuire a different mounting solution, though when we're designing things from the ground up, that's less of a problem. These still have some light leakage, as there's no "seal" to stop it. It is, however, considerably less than regular fans.

Why not just power the revolving door and use it as the fan. It is basically a type of squirrel cage fan

Because the powered revolving door moves just as much air in ans it moves out, and you would need it to be MUCH faster in order to compensate for losses. It's also a very heavy mechanism, which need more power to rotate, particularly at high speeds.

The reason that "squirrel cage" fans work is becasue the intake is at the end, and the output is perpendicular to the axis. A revolving door won't move air by itself, at least not much -- this is why it works great for buildings that don't want to lose all of their heat (or AC) when people go in and out. Literally, an open hole would flow more air than the revolving door would, when used as a fan this way.

So we need something else to generate a high pressure area on the intake (lamp chamber) side, and a low pressure area on the output (outside) before any air will flow. That means something else as a fan first.