I have never seen a fan built for any residential, or commercial application for that matter that would not start on the lowest setting. If the motor has lost that much of it's available output power, something is drastically wrong, and will most likely catch fire or trip on thermal overload.
The only load the fan must overcome is the initial friction of the bearings. Reverse flow isn't something one worries about in residential applications.
I'm sure that the pull chain attic fan was a regular ceiling fan motor mounted in a frame for installation between joists. It was NOT what you describe because it would have cost more if done correctly.
I'm willing to bet that the probelm is that many of these consumer items are never tested in their environment. I'm sure that the goal was to produce a fan that would sell and thus make money. In fact, I've had experience with products designed for specialized markets that were poorly tested prior to release and thus design flaws weren't discovered until they were in the field, when real people used them for real, and costs to correct the flaws is extremely high compared to the design phase.
The rationale for the first position of a 3 position switch or pull of chain is that it gets the motor running at maximum torque. As the motor ages, this is needed. Starting the motor in the slowest position, lowest torque, may stall it and cause higher currents,etc. So starting at high speed is sensible and the user can then pull twice or turn switch to low position after the fan gets moving.
I'm chuckling at the number of people who are confusing a ceiling fan with a whole house atic fan... The two are no where close to each other than the word FAN.
I do have to admit, I've never seen a whole house fan with a pull cord, let alone 3 speeds. Every one I've ever seen was wired down to a wall switch with a 30 minute mechanical timer with a hold position.
Thinking about putting one of these in myself actually, for the exact same reason, they move a ton of air, probably in the 10-20,000 CFM range.
Since I happen to have the technology available, I was planning on using a small 230v VFD and running it variable speed vs. on/off or 3 step control.
@Beth -- By far, I have found the Number One important concept in System Engineering is to be mindful of the Environment... Not the environment as in Environmentalism, like clean air and water, but the System Environment. There are several views on design optimization:
1) Use the Best Ingredients (Input)
2) Use the Best Practices (Process)
3) Insure the Best Quality (Output)
All three of these are great, but taken separately, or even as a group, if the list does not include
4) Be mindful of the effects of things external to the system (Environment)
then we get things like the inflow of house air causing a current which entrains the pull cord which subsequently gets caught in the blades.
Testing a finished or prototype product in its actual usage Environment should identify a host of potential problems.
This comment is already too long, but it brings to mind the valuable lesson from Langley vs. the Wright Brothers. On paper, Sam Langley was an excellent engineer that could design the best powered flight machine using the best components, the best technology, and the most powerful engines. It took the Wright Brothers and their wind tunnels to simulate the flowing air environment of thousands of wings, kites, and control systems before they started testing on the beach -- a place that provided wind and a soft landing for anticipated crashes. Langley launched his machines from an aircraft carrier in the Potomac River... and continuously fished them out from the bottom.
William, your reference to the system engineering angle is well noted. What would your suggestions be to your students, given this example, in terms of helping avoid what looks to be an all-too common scenario?
I feel blessed to have two small, inexpensive ceiling fans with globe lights that have worked nearly perfectly for 12 years--they are going constantly in winter and summer--and many years before I inherited them when I bought the house. They are small, painted white (which does not fit the rest of the decor) and I at first considered replacing them but never got around to it. Now I'm afraid to. Everyone I know with great big more expensive ceiling fans by well-known national brands complains about some fault or other. The only problem we have is that the one whose light is switched on and off most tends to occasionally wobble slightly and vibrate, making a very annoying noise. I fixed that by inserting a large, thick rubber band around the neck of the globe which makes a thin cushion, before it's inserted into the housing and screws screwed in to secure it. I change it every 4 months or so.
I have a fan that works exactly like this. I was very surprised at first, but silly me, I figured there must be some logical reason it was designed that way. What once seemed over my head was probably just a dumb design.
Classic case of solely using Analytical Problem Solving with no regard to Systems Engineering. In isolation, I'm sure the fan, the louvers, and the switch were perfectly compatible on paper and during assembly. But actually measuring the performance of the system as it would be used in the field was not addressed. To have it fail on first use is pretty silly. Even with modest Systems testing it should have been easily discovered in the lab. I'll definitely be using this story as an example in my Systems Analysis course...
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.