Homemade Ducted Fan Motors from PC Case Fans

[LBMiller5]

Hello everyone! This is my first post, and I would like to introduce myself to the GoBrushless community. My name is Lucien Miller, and I live in San Marcos, CA about 30 miles north of San Diego. A friend of mine turned me on to this forum a couple weeks ago, and I have been checking it out every day since. I used to be very heavily into R/C several years ago when I lived in Florida. When I moved to California 3-1/2 years ago, I sold all my R/C equipment.

When my friend told me about all the fun he was having building CD-ROM moders and flying these little Foamy Planes in his back yard, I just had to get back into the hobby.

Well enough about me, let's get to the topic at hand. I maintain a lot of computers at work so I have access to a lot of dead CD-ROM drives, cooling fans, and other PC Parts. I have already accumulated quite a pile of used motors from old drives. Here is a photo of my collection so far.

http://innov8tivedesigns.com/rcgroups/Stators.jpg

At the bottom center of the photo are 2 notable items. At the very bottom is an interesting Double Stack Rotor Can. This came out of a Toshiba 6702-B CD-ROM Drive. Many of you that have torn apart your own CD Drives have probably noticed that some of the CD Spindles at the end of the motor shaft have a hollow ring full of little ball bearings. This is a balancing assembly that evens out an out-of-balance CD Disc. The Toshiba CD Drive had the balancer up inside the can, so the can was twice as tall as a standard one. I plan on using this one to make a double stack motor.

Above the Double Stack Can is my first CD-ROM Motor in progress. My Brother-In-Law has a neat little Unimat Lathe that I borrowed to make my own bearing tube. It has a GWS gearbox shaft and a pair of 7x3x3 ball bearings. I still need to put in a set of NdFeB Magnets and re-wind the stator, but mechanically, all the machine work is done. Here is a closer look at it:

http://innov8tivedesigns.com/rcgroups/FirstMotor.jpg

Another item that I have accumulated a lot of is the fans used to cool power supplys and CPU's in computers. The fans come in 50mm, 60mm, 70mm, 80mm, 92mm and 120mm diameters. Last week I popped apart an 80mm fan to see if there was anything useable inside. This particular one had a pair of 8x3x4mm ball bearings supporting the shaft, and the fan assembly had a 3mm metal shaft. The fan itself has a diameter of about 73mm. There was one of those 2-pole rubber magnets wound around the inside of the fan and what appeared to be a pressed-in soft iron flux ring. I slid an Xacto knife in between the magnet and the metal ring and the magnet popped right out. Here is a photo of what the fan looked like at this point:

http://innov8tivedesigns.com/rcgroups/Fan1small.jpg

I got out my dial calipers, and the metal ring measured 29mm ID. I took one of the stators that I had removed from one of the CD drives and slid it on the shaft. It looked like there was a little over 2mm of clearance between the metal ring and the end of the stator pole. I had one 10x4x2mm NdFeB magnet so I slid it into the gap and voila! Perfect fit! There is a little lip at the bottom of the flux ring, and the sides of the flux ring are 10mm tall. the bottom lip also helps to square up the magnet so the positioning is perfect. Here is the fan with 1 magnet and a stator.

http://innov8tivedesigns.com/rcgroups/FanMag.jpg

I then started looking at the various size magnets that are available and made some drawings in my CAD program to check magnet placement and air gap spacings etc. With this rotor if I use a double stack of the GoBrushless 22.7mm 9-pole stators in conjunction with twelve 10x5x2.5mm magnets, I end up with a 0.43mm airgap between the magnet and the pole tip. This is a little on the big side. However, if I use the 10x5x2.75mm magnets that are available at www.engconcepts.net , I only have a 0.18mm airgap! This looks very promising! Here is a copy of the CAD drawing showing the placement with 2.5 and 2.75mm magnets.

http://innov8tivedesigns.com/rcgroups/227df29.jpg

Then I looked into the possibility of using the 25mm stators, both in 9 pole and 12 pole LRK configurations. Using 10x5x1.5 magnets I get a stator to magnet air gap of 0.28mm. Not bad at all! Here are 2 drawings showing the layout of the 25mm Stators. The first one is a 9-pole stator with 12 magnets and a standard ABCABCABC wind. The second one is a 12-pole stator with 14 magnets in an LRK configuration with AabBCcaABbcC winding on the rotor.

http://innov8tivedesigns.com/rcgroups/25df29.jpg

http://innov8tivedesigns.com/rcgroups/25df29lr.jpg

I am planning on purchasing several 22.7mm and 25mm stators, along with several diferent thicknesses of 10x5 NdFeB magnets and start experimenting. My goal is to find the optimum Pole Number-Wire Gauge-# of Turns-Magnet number and Delta vs Wye combo that gets the most thrust out of this 73mm fan. Is there anybody out there who has done anything like this before, or who would like to join me in some experimenting? If so we can pool our resources and get something that works.

Well I think that is enough for this post. Hopefully this will generate some interest.

Lucien

[Q]

I am interested in the outcome of your project, but not into EDF at this time.
I just want to tell you to be careful with these fans. Spin them up into a safe environment before regular use. They are not built for very high RPMs and can really explode. I have heard reports of people getting really hurt by a fan shedding a blade and hitting the person in the face.

[LBMiller5]

Thanks for your interest & concern Q. I have tested these fans before by chucking them up in a Dremel tool and spinning them to over 20,000 RPM to see how much thrust they would make and to see if they would stay together. I always perform the tests with the Dremel tool at arms length with the fan oriented so that pieces will not come flying my way in the event of a blade failure. I will most likely reinforce the back end of the fan hub with some Kevlar cord and thin CA glue to insure that the centrifugal forces do not seperate the plastic fan assembly from the iron flux ring. With the fan having a 29mm ID, even at 20,000 RPM the forces are rather low due to the small radius, so this will probably not be an issue.

During these tests, I did get a good amount of thrust. This is in spite of the fact that the fan is running in free air, without a shroud, with significant tip losses. I have also had no blade failures or even the hint that the fan was being overly stressed.

After I get something going, I will probably build a foamy stand-of scale HE163 or A-7 Corsair or some other plane with a fairly large inlet opening to simplify the ducting issues and get some real world flight testing.

If I find that the double stack CD-ROM typs motor does not have enough oomph to turn the 73mm fan, I will try again with a fan from a 70mm unit, which will probably have a 63-64mm diameter.

[Q]

Your tests sure sound promising. I need to go through my useless PC stuff and see what I can dig up.

If you need big motors, go for the A10 Warthog. I have seen that beast IRL and it's HUGE

[LBMiller5]

I went through a some boxes of parts I had laying around the shop and dismantled a few fans. I have noticed that just as in the CD-ROM drive motors, it is tough to get consistency from various manufacturers. I have collected 35 fans so far for experimenting, and I still have a dozen or so I have not parted out yet. Below is a photo showing the collection thus far:

http://innov8tivedesigns.com/rcgroups/35Fans.jpg

From the photo you can see the several 80mm fans, a bunch of 60mm fans, and a few 50mm and 40mm fans. The actual diameter of the fan runs about 93% of the shroud size. The 80mm fans are 73-74mm, the 60mm fans are 55-56mm, and the 50mm fans are 46-47 mm. I also noticed a lot of variation in the ID of the iron flux ring. On the 80mm fans, some were 29mm, some were 29.5mm, some were 30.5mm and a couple were 31.5mm ID. This can be be taken into account with different thickness magnets, but for my purposes, I will try to find groups of 4-5 fans that are exactly the same.

Another issue is the fan blade shape. From the photo you can see that some sweep forward, some sweep backward, and some are straight. I am not sure what difference this will make to the performance, but we will find out. Some have rounded edges, but most of them have squared off edges. All of the 80mm fans have 7 blades, On the 60mm fans, some are 7 blade, some are 9 blade and some are 11 blade.

I checked the pitch of the blades on several 80mm fans, and most of them have a blade angle of about 45 degrees at the hub, and transition to 25 to 30 degrees at the tip. Doing the math, this works out to a pitch of 4 to 4-1/2 inches. So basically these fans are 3" Dia. x 4.5" pitch, 7-blade propellers.

I am going to try to use the existing shafts that are already in the fans. They are long enough to hold 2 bearings, but I will need to machine some type of mount to hold the bearing tube.

Once I get the fans sorted out I will build a set of interchangable Stator and Rotor assemblies. I can set up 4 fans, one with 8 magnets, one with 10 magnets, one with 12 magnets, and one with 14 magnets. Then I can make 6 Stators with different winds of different gauge wire. This way I could try 24 different combinations from 6 stators and 4 rotors. I could also bring out all 6 wire ends about 2 inches and attatch them to a terminal block. This way I could connect the wires outside the motor to test both Delta and Wye configurations for each stator.

Wow!! 48 different configurations from 4 rotors and 6 stators. I just need to build all the bearing holders the same, and hold the shafts in with C-Clips so they will all be removable and interchangable.

I would like to try both 9 pole and 12 pole stators in the experiment, with both standard wind and LRK wind on the 12 pole stators. I will probably end up making more than 6 to try all the combinations, possibly 10 or 12. (Perhaps I can talk ScubaSteve into a quantity discount on stators in the name of Science!)

I have access to my Brother-In-Law's Unimat lathe to make bearing tubes. This is a great little tool that is perfect for this purpose. It is about 18 inches long from one end to the other. Here is a picture of the lathe:

http://innov8tivedesigns.com/rcgroups/Lathe.jpg


Here is a shot of me turning down a piece of 3/8" aluminum tubing to fit the inside of the stator:

http://innov8tivedesigns.com/rcgroups/Lathecut.jpg

Well I think I know what I will be doing over my 4-Day New Years Weekend! I will machine some bearing tubes while I wait for the magnets and stators to come in.

After I get the parts done, I will build a test stand with a digital scale and get some RPM and thrust measurements. I will use a lab grade power supply with digital voltage and current displays to drive my speed control. This way, I can get exactly the same voltage for each motor, and accurately measure power in on each test.

By using a power supply, I can set it for 7.40 volts and 11.10 volts to simulate 2 and 3 cell LI-PO battery packs. I can also set it to 7.2, 8.4, 9.6 and 10.8 volts to simulate 6, 7, 8, and 9 cell Nicad or NiMH battery packs. This will provide consistent data so I can compare apples to apples.

If you have not figured it out by now, I have a degree in Electrical Engineering with a strong background in Mechanical Engineering as well. You probably got that already by my test set-up description. :wink:

I will be generating a BUNCH of data from this experiment. After I get all the data compiled, I will check with ScubaSteve and see if he wants to put it into the performance database, or if it would warrant a seperate database dedicated to ducted fan designs. We will see.

Well enough for now. I will post some more later.

Lucien

[Q]

Hi Lucien,
Very interesting "stuff" you got there.
What's the trick to take the fans apart? Many people try to use the fans as a cooling fan on bigger brushless motors, but I never managed to take them appart in a half decent way.

I'm planning to build a test stand using an 8 channel 12 bit data capture card. This card can measure voltages over a range of 0 to 4V so I need a volt, amp, rpm and weight sensor that gives me a range from 0 to 4V on the output side. I know the basics of electronics, but not enough to design a layout myself. Any help would be appreciated.

[LBMiller5]

Q,

Taking apart the fans is pretty easy once you know how.

The sleeve-bearing fans usually have a nylon clip holding things together with a lot of slop. They rely on the magnetic flux field to center the magnets around the stator.

The ball-bearing fans have a metal clip holding everything together with a spring of some kind at the front end holding it all in place.

The procedure for disassembly is the same for both types.

1. Remove the paper or plastic sticker that is on the back side or the fan.

2. Most fans have a small bearing cap about 1/2" diameter under the sticker. This cap can be removed by lifting up an edge with an Xacto knife or small lewelers screwdriver. If there is no cap, go to step 3.

3. If you look down into the hole, you will see the back end of the shaft and some type of retaining clip, either plastic or metal depending on the fan. The plastic ones are pretty easy. If you don't need to save the clip, you can cut it out with an Xacto knife. If you look closely you will see a slit in the plastic clip. Just cut the clip from the outside towards the center opposite the slit. The clip will break in half and you are done. If you want to save the clip, carefully lift up one edge with the tip of a small screwdriver and work it off the shaft. This can be like trying to catch a wet fish as it spins and slide around on the shaft, but it can be done.

3a. If you have a metal clip and ball bearings the removal can be a little tougher. I have seen 4 different types of retaining clips in the 50+ fans I have dismantled. The first is a thin spring steel washer that is missing about a 30 degree segment. These can be pressed off the shaft easily by pushing them off the shaft sideways. The next type is a spring ring clip with a couple small cut-outs at one end. If you have some very small Spring Ring Pliers you can open them right up. If not you can use 2 Xacto Knifes or 2 tiny jewelers screwdrivers to spread open the clip to remove it. Just be careful you do not ding up the metal bearing shields!

The last 2 clips are both made from very thin spring steel wire. One had the wire bent into a 3-sided star shape with the clip engaging the groove in the shaft in 3 places. This one is real easy to pop off. The last clip is a real bear. It is a ring that is just 1 coil of a spring that completely fills the grove, and only has about .001" of gap at the ends. The technique I use for these is to press the ring at the gap with the tip of a small jewelers screwdriver. This causes the ring to open up a little on the opposite side. Then I take a second screwdriver and lift the ring up and over the end of the shaft. Now that the clip is removed the fan can be pulled out.

4. Hold the fan shroud in one hand and push the end of the shaft with a toothpick, or a small jewelers screwdriver and slide the fan out. There will probably be a small O-ring at the base of the shaft that can be removed if necessary.

5. The magnet in these fans is a rubber ring that varies between 1/16 to 1/8 inch thick. In the majority of the ones I have taken apart, there is no glue holding the magnet in. It is a snug press fit, plus the magnatism holding things together. Look closely and find the seam in the magnet. On some there is a little dot of paint at the seam, and on others you just have to look for it. Slide a small jewelers screwdriver between the back of the magnet and metal ring. Pry the magnet up slowly, being careful to not bend the metal ring. Once you get one end popped out, the rest of the magnet just about falls out. Occasionally I get one that has a couple small dots of glue on the magnet. For these, juse slide an Xacto knife between the magnet and the ring and it will seperate.

6. If you have a ball bearing fan, you can carefully push them out and re-use them if you want. Every single 80mm fan I have taken apart used 8mm OD x 3mm ID x 4mm Thick bearings. This must be the most common size made in China, so everyone uses them. These are bigger than the 6x3x3 bearings commonly used to build GoBrushless motors so you would need to build your own bearing holders to use them. They can be used with the larger 25mm stators that have a 9.3mm center hole, but they are too big to work with the 22.7mm stators which only have a 7.9mm center hole.

Well, know you all know how to take apart the fans now! Hopefully this thread will generate some interest and cause some real growth in electric ducted fan aircraft. If anyone would like some photos of the process. Let me know and I will edit this post to add some.

Q, it just so happens that most of my Electrical Engineering experience is in analog signal conversion so I may be of some help to you. I will have to knock the rust off my slide rule and dig into some data books, but it sounds like fun. Pop me an email at lbmiller5@cox.net and let me know the particulars of what you want to do.

Lucien

[Q]

Thanks for all the info Lucien. You must be a fast typist.
I'll type up an e-mail tomorrow. I'm using a different computer today and don't have my e-mail access.

[LBMiller5]

I went to the AMA show in Ontario, CA this past weekend. It was very cool! There was a lot of stuff for electric aircraft. They even had a special room set up for flying foamies & mini helicopters. I picked up a bunch of stuff to get going on several projects. One of the items I got was a GWS A-10 kit to use as a test bed. I bought the slope soaring version that comes without the fan units, since I will be making my own ducted fan motors from a pair of 60mm CPU fans and a couple 20mm CD-ROM stators. It should be a fun little plane!

Back to the Ducted Fan Project.

I received my magnets from engconcepts.net last week. I still need to get the stators to complete my parts package. I should start building a test stand later this week while I wait for the rest of the parts.

Once I start getting some data, I will post it here, so keep watching!

Lucien

[juki48]

any new progress reports on the duct fans? pics of complete units? Im pretty interested to see how it works out.

[LBMiller5]

No not yet. I still need to get some stators from Steve. I will probably get started next week.

Lucien