Author Options:

Ceiling fan generator mod to the max Answered

I stubled upon several mods to convert a standard ceiling fan into a more or less usefull generator.
So if you are looking to go this route then I might have some nice improvements that can be implemented.

People like these mods for some weird reason, despite the fact that it requires quite a bit of extra work to make them weather proof.
However, when it comes to the fundamentals then to me it looks like some folks out there are missing out.
On the available power that is...

Always the first step for a mod like this is to replace the induction ring with a lot of magnets.
Second step usually is to remove a lot of the coils, especially the inner ring.
Now, these two stator designs are common for fans with two speeds.
Those with three or even reverse might have a different configuration!
Lets start on the magnet part:

The recommended way of placing the magnets is by creating an air gap as small as possible - makes sense.
But then it is always the same amount of magnets as there is coils - and the spacing is also the same as for the coils.
In the general generaotr design world this configuration is prefered as it allows for the best performance.
If you dare to go a bit further and cosider how the magnets react to the stator configuration then you might want to consider a different option.
You see, these two sets of coils for two different speeds mean just one thing:
A different amount of poles is created, with the outer ring having more poles than the inner ring of coils.
The core is split around the coils, not just to allow the windings to be made but also to provide independent paths for the magnetic field - resulting in the two pole configurations.
Amounts differ by diameter, power level, manufacturer and so on.
What is always the same is that the inner ring has less coils and that the outer segments of the poles created have even spacings.
In the normal mods you see posted these gaps in the core for the outer ring are closed by inserting lamitaed pieces from some old transformer.
And you end up with ONE usable coil configuration and ONE power output.
The slightly advanced mod uses the inner coil to add some load depending on the speed to prevent spinning out of control in high winds.
If you try a normal DC motor with permanent magnets than you will notice the strong binding forces, it is like the rotor sticks in certain places.
The better ones use and uneven configuration to reduce this binding effect ;)
In my mod the magnets are selected in size to almost be the same length as two stator poles next to each other.
This allows for the best induction while still allowing "to experiment".

Bringing the magnets and the coils into play...
As said an exact match of the number of magnets to either coil ring is not ideal.
The prefered option is to go somewhere in between.
For example:
Outer ring has 18 coils then the inner ring will have 9 coils - exactly half.
360° divided by 15 make a nice 24 degress per magnet.
But with 12 magnets you get an even 30°, which is far easier to deal with.
16 magnets at 22.5° is another option.
So, what does that exactly do for us?
The bad thing is we get slightly less performance if you only see the standard mod with one coil ring.
The good thing we get far lower binding forces and through that the thing will even spin in very light winds.
Adding both coil rings with a suitable rectifier however results in a pulsing output of two sine waves.
With just the rectifier we get a ripple that is easier to deal with through a capacitor.
The extra power available is in the range of about 40% and make more than up for the "reduced" amount of magnets.

Going the extra mile once more ;)
Having created a much fancier ceiling fan mod now you might wonder if there is not a way to get even more out of it.
And there is.
For example by utilising a gear system or belt to get a far higher rotational speed on the generator than what the blades would provide, prefably then with quite big blades too and an automatic break for high wind conditions.
With the reduced binding forces the generator will be happy to spin at quite high speeds in low winds.
Downside is that you will need to build a far more sturdy bearing housing.
In return though you get more stability and durability.
You can do the math yourself based on the number of poles per ring and magnets to get the output frequency based on the RPM's.
Perfect would now be to use a switch mode power supply configuration to directly transform the provided output into a stable DC per ring.
And yes, it is possible to use mechanical systems to provide a fixed output speed from the blades to the generator - but way to complex and lossy!
Lets do some lame math with no regards to realities:
If the original fan would spin at 100 RPM at full speed than we could say our generator should provide the mains voltage at about 100 RPM.
Keep in mind we utilise both coil rings and not just the high speed one!
Geared and with the blades spinning at 100 RPM we might get as much as 1000V from this little generator....
And even with the lower amount of magnets we migh see frequencies above the 500Hz range.
The good thing now is that normal iron core transformers can still operate at these frequencies.
A bit lossy in the upper range but acceptable for the purpose.
Put simple: A 10 or 20:1 transformer per coil ring would provide us with a far more suitable output voltage and much higher amps.

If you made it to here than you certainly wonder about other magnet configurations.
Checking the stator configuration you will by now realise why I selected the magnet lenght accordingly.
The magnets "activate" one coil after the other.
The spacing between them means there is always some overlap where the magnets only cover one half of the stator for a coil.
This is ok because we don't really have to worry about the resulting messy output.
Ideally though you would want to have a magnet activate both coils, the inner and the outer at the same time.
What we did though was to make sure that at no time more than ONE magnet fully covers more than ONE coil!
It is the best option to cover both coil sets while minimising binding effects and increasing the avialable output.
To go the last step you would need to invest a lot of time re-winding all coils :(
You don't want to do this unless you have the means and no friends and family that might miss you for a few days....
I found a far simpler way to change the coil configuration, although it is not as good a re-winding.
So let's go full scale shall we?

Ceiling fan reconfiguration!
If you take the usual 18 to 9 configuration than one thing jumps to mind reight away: 3-phase power!
Cutting the wire that goes from coil to coil might not always be possible and if it is then you need to know how to handle it.
Magnet wire can be hard to solder.
Burning the coating off results in corroded copper that is even harder to solder.
If you are lucky though than a reall hot soldering irong will be able to melt the coating.
The flux from the solder will start to cover the wire from the cut and the solder will follow.
If not then using some fine sandpaper and time is the other option to remove the coating...
Ok, you seperated all coil and have two wire ends per coil?
I hope you did not cut off the ones going out to the actual connections to the outside world ;)
Properly solder each wire end and take your time to check it is really proper and not just a few spots.
Mark or number the coils on the rings!
For the inner ring we have 9 but need only 3, so we start at one connection to the outside world and check if this connection is on the outside or inside of the coil.
For this example I assume you picked the one that goes to the outside of the coil.
Connect the inside wire to the outside wire of coil number 3, assuming we start with 1 here ;)
From the inside wire of 3 you go to outside of 6 and the inside is you first new output connection.
Do the same with the remaining 6 coils and where needed add the required output wire.
It really helps to have wires with three different colors here, one color per new coil set.
Note which color corresponds to to the three coils used!!!
The outer ring with 18 coils is sightly different here.
You see, we want a "flowing" magnetic field that makes best use of the new coil configuration!
We can not simply bridge them in any way we feel like without considering how this might affect the electrical side of things.
As we now take the approach of a three phase system it makes sense to use a more suitable magnet configuration as well.
So before go to the outer ring of coils lets have a look of the best option for the magnets first:

The stator packs are evenly spaced in our example and will alow us to use 18 magnets.
This provides the best performance with the downside of a higher binding effect, but we need this configuration to get the best possible output.
As said at the start I selected magnets that are just shy of being the same length as the corresponding stator segments.
In a "free" setup these magnets would now be quite hard to place in a makeshift ring.
Even harder in the original casing.
A 3D printer certainly helps but some common sense too ;)
Wood is easy to work with and if you select the right stuff than making a suitable ring to hold your magnets and attach to the drive system metal parts is not too hard.
Bar or brick type magnets can be quite easy be utilised on a wood setup :)
The key is that you add Flux Capacitors - sorry couldn't help the reference to Marty....
What I mean is to add some magnetic material between the north pole of one magnet and the south pole of the other.
Lets say your magnets are 15mm long and have a spacing of 5mm.
Then a little plate of 12mm would be next to perfect.
This plate needs to connect the magnets on the backside, the side facing away from the coils.
Use a dremel tool or what you have to first create slots for the metal strips or bars, then the same for the magnets.
Glue in the metal first and once set add the magnet, making sure the always go north to south with their alignment.
Ok, and what does this do for us?
I hope you are not one of these persons who starts building while reading...
What we created now is a shortcut for the magnetic forces.
The field between the magnets is severly compromised in terms of being usable for the coils.
We do get a much soother run though...
I only did that to have some fun and check if you paid attention - sorry :(
What we really want is an effect similar to what you see on a loadspeaker magnet that is still in its metal shielding.
A ring magnet with one pole on the inside and one on the outside is used here.
The shielding provides a path for the magnetic field that is not going through the speaker coil - hence the little air gap for the coil.
If we do the same then our efficiency will be going up quite a bit.
Take two identical steel parts, like some butter knifes, and prefarbly a force gauge.
If you try to pull your magnet at a 90° angle from the blade you will get a certain reading for the required force to lift it off.
Most people now think that this would be the max a magnet can hold.
So take the other knife and place the magnet between them.
If you pull the knife off with the gauge now the reading will be higher than what you get from just the magnet ;)
Taking that to our model and keeping the field lines in mind we now know that we could even use slightly longer plates if our magnets happen to be a bit short :)
Just place them right behind each magnet !

Back to the outer ring of coils....
With 18 magnets we get an even system for both coil rings.
However we want to make sure that our output waves are syncronised and not at random order.
We need to combine two coils to be back on a 9 coil configuration as on the inner ring.
The other option is to provide two sets of outputs for outer ring, resulting in 3 3-phase outputs.
Both have their pros and cons....
But if you check the 18 magnet configuration ina ction over the coils it becomes clear that combining two coils the usual way is possible but also that our inner ring does not get a proper north south action from the magnets!
Only the outer coil ring works properly!
For the inner ring we never get only a north south combo, instead a lot of mixes.
Did I mention to read first? ;)
Of course we can only use 9 magnets in our configuration, but at least I did not traick you on their size....
You see, we need to account for the fact that the coils are not just evenly spaced but also that all configurations in terms of coils to stator pack are doubles or halfs.
Makes a lot more sense if you know how these asyncronous motors work :)
With 9 magnets we actually get both inner and outer ring coils activated properly.
Plus we now have the benefit that there are always twoouter coils in sync with each other.
Means apart from the same way you wired the inner ring you make this addition to the outer ring:
"One" outer coil is created by going from one coilinner connection to the outer connection of the second after this, skipping one coil.
The resulting output is again just 3 phases but with double the output voltage.
The key is to again take notes of how you connect and wire the coils - and the colors used for the output wires!
Let me give you an example for the correct order:
I we take the number 1 coil on the inner ring then coils number 1 and 18 would be next to it on the outer ring.
You want to combine 1 and 3, 2 and 4, 5 and 7,....
And you want the resulting three coil packs and wires colores to correspond to the inner coils in the same order!
That is true for the always same way of combining coils from the inner to outer connection - or the other way around but never mixed!

Ok, we have done the magnets and the coil configuration now properly, no jokes this time!
With two simple 3-phase rectifiers we get two DC outputs that can be combined or used seperately.
As we end up with roughly double the output voltage on one output but all coils are the same it makes sense to treat them independly.
For those who wonder why:
If you add a load than one coil system would take a higher loading of it.
Meaning while one coil set is stll fine the other will already start to overheat - if the load is too great.
So we use two rectifiers with some filtering.
In the basic form just a really big electrolytic capacitor of suitable voltage or a full LC filer system with multiple stages.
Either way we can now utilise some better DC-DC converters to get going.
Considering the equal max watss the coil rings can handle it make sense to include some current limiting.
A good converter will provide this option.
Both converters can now set to the desiered output or with some added protection diodes and adjusted properly to the same voltage combined for just one DC output.
Compared to the standard mod of removing coils and bridgning stator packs the resulting output power in overall Watt will now be about 40-60% higher - depending on the model and quality of parts.

Special words of wisdom:
Consider the orignal max speed of the fan when used as intendet - see this as a theoretical max output that equals your mains voltage.
Just ignore losses and such things - better to be safe than sorry.
It becomes clear that it quite possible that your output will be far higher than mains voltage and that you need use transformers for the two 3-phase systems so you can use standard DC-DC converters, which have a max input voltage of around 50V only.
This means your converter must be able to handle the higher amps!
The fan might have only used 100W or less than 500mA but at high speeds and a ratios of lets say 10 to 1 for the gearing high wind speeds might get it up to over 5 amps on the transformer outputs.
Please do the math first for your gear system in relation to the max wind speeds you want to use with your blades!
If in doubt use a converter that has some reserves to offer, especially if you aim to charge batteries as quickly as possible.
The most vital part however is to ensure that all previously cut wires are isulated properly!!!
Magnet wire of the standard kind is good for about 1000V max, so don't drive it higher!
Heat shrink with a hot glue liner is prefered but hard to apply in these thight spaces.
Since nothing moves consider using long enough wires for your connections so you have enough space to solder without affecting the heat shrink tubes.
Liquid insulation or rubber is the last option and should only be used to finalside the heat shrink security measures.
Best option once all is confirmed to be working fine would be to make a custom mold and to fully enclose the staotr pack and wires with casting resin or an insulating casting mix.
Make sure to keep the output wirese free at the their ends ;)

What if I don't want to build a complicated three phase rectifier and just use a single phase system as it was?
Firstly chances are your coils are already connected in a three phase configuration, just all in series.
But working out a suitable magnet configuration to suit this is much harder if you want to use both sets of coils.
In a series configuration like the original you also have to accept the losses from these connected coils.
The higher the overall resistance the lower the possible output ;)
Main problem however is to get the magnet working properly.
The standard 9 or here even 18 magnet configurations still works, especially with the added shielding from behind.
But the coils also produce a magnetic field, which grows with the load.
Means that an top of all you also have the coils working against the magnets and create even higher losses.
Explains why the simple folks prefer not use the inner coil set if they go with a single phase system.
So either accept the losses and just use the outer coils or do it fully and get far mor output.
And by the way: a 3-phase rectifier modlue is only a few cents more than a standard bridge rectifier ;)

Ok, and why do I bother to write all this?
People like to tinker but most don't really invent.
Following some simple instructions is easy, trying to work it yourself much harder.
The reward however is that you actually start to know what you are doing :)
And what works for a ceiling fan can be used for these ring style washing machine motors too ;)
We need to get back our roots.
Start thinking for ourself again, work things out instead of just looking them up.
If people would be aware that a simple ceiling fan could provide about 3 times the output power of its rated installation value instead of only just about half......
Super strong magnets allow real output even without re.winding all coils.
And what works here works for other things too.
We only learned to use magnets in a striaght way because we can not bend them.
But we can bed the magnetic field lines to our advantage!
The simple shielding used in this mod is nothing more than a shortcut to enhance the field strenght where it is is needed.
By a simple coil modification we basically bet two electrical generators for the price and size of one.
Apart from stating how easy it would be to place multiple stators and magnet rings into one generator the magnets itself also allow for even more output.
If you ever played with hook magnets or speaker magnets then you know how much stronger they are compared to just the magnet once they seperate after hours of fun for you.
Imagine you would replace the single bar magnet with two block magnets that are joined by a magnetic shunt like out simple shielding before.
If the magnet blocks now would have a slightly smaller footprint than your individual poles:
Imagine you create a hlaf ring shaped magnetic connection between the two blocks that also goes aruond the outer perimeter up to the outside of the magnets surface?
I mean the surface facing the stator poles?
Damn your imagination is good, yout it right away!
Of course we would then have a magnet that allpies its full strength focussed onto each pole of a coil!
And of course the resulting field would be far stronger than just using the magnet blocks itself and still significantly higher than just adding a shielding or connection between them.
The affect of the next coil coming is also drastically reduced, which in return also increases the efficiency.
In terms of numbers:
If a fixed neodymium magnet would provide us 100$ field strength as the base point with no shielding (just the magnet blocks alone);
A fully shielded and connected system, like in a hook magnet combined with a U-style magnet, would reach above 400% here.....

Adding witchcraft to the mix ;)
Although I know better I just assume some of you have now a working double-three-phase-ceiling-fan-generator.
And that would mean you also have some fans to spare from your long experiments.
Modern ignition coils seem to have nothing in common with our ceiling fan or resulting generator.
So why do I try to use them anyway?
For the ignition only one polarity is prefered so the spark works and travels as intendet.
Means the "wasted" energy from the othe half of the pulse seems to be lost.
The electronics do a lot here but magnets too ;)
The core of the coil has magents at either end, turning it into one long magnet that still has the right properties to act as high voltage transformer system with the coils.
The coil appear to be pre-loaded and with the ignition pulse it has to overcome the magnetic field pre-set by the magnets.
And when the electrical impulse is off the same magnets also accelerated and increase the resulting fall back impulse - which provides the spark.
Unless you have a suitable laser cutter or simlar cutting tech available somehow it will be hard to modify the metal plates of the stator.
But if you could...
Imagine you could add magnet inside the plates that are inside a coil.
The same pre-loading would happen.
Does not really help in terms of adding outpur as our rectifier would suffer badly here.
It does give ideas though...
Shielding works fine for the magnets, same for field shaping.
Electromagnets use the same techniques...
So why not use some leftlever transformer cores to add more "shortcuts" for the coils?
Strips of transformer core sheets added either side of the coils increase their field strenght and result in better output!
Three packs either side of the stator pack are usually no problem.
Now take your leftovers and do a standard mod.
Compare the max output on the same windmill with what you get from my mod(s).
The only real magic I used here is that I actually bothered to combine multiple and already used methods to drastically increase the available output of an otherwise utterly useless generator mod ;)

If you take the above mods serious and to the their extreme than it is imperative to make sure you have safety measures in place!
Assume the lowest rating for the magnet wire and if in doubt stick with a max output voltage of 800V.
These mods are potentially lethal if you don't follow what is common sense to everyone dealing with high voltages for a living!
Most people will start without any gearing or belts and use the wind directly.
Even here it is easy to get far higher RPM than what the thing ever did under your ceiling.
Without some fixes you will need transformers to reduce the output voltage accordingly.
Only other option is to limit the max speed to what your DC-DC converter can handle.
Making mistakes with mangets can cost you a lot of time and work, make sure to mark their poles somehow to prevent putting them in wrong.
If in doubt then double check!
Always keep in mind what the magnet wires and your connections can handle!
You don't want any arcs or overheating.
Some added electronics to monitor wind speed, rpm's, load and temperature of the coils can turn out vital once you upscale.
Before letting your new generator do its thing make sure you tested all to the max!
Use a drill or so to speed it up and check the limit regulation for the converters.
Measure the actual volts and amps going through your coil sets at assumed max speed and max load.
Monitor the coil temp while doing so to ensure nothing is out of limit!

You are kidding me here right?
A scrap ceiling fan shall provide more output as a wind generator than what was used to spin it as a fan?
And of course I need not one but two 3-phse transformers...
Pretty clear it is all a fake because nobody could replicate any of it unless limited to what the converters can handle...
Didn't I say to think outside normal restraints already?
A single phase transformer uses two coils in the most basic configuration.
For example one side for 240V and te other for 12V.
But some of them are more efficient than other ;)
A 3-phase transformer uses 6 coils, two for each phase.
And there are plenty of standrad transformer cores out there that would allow us to use this configuration.
The worst being the MOT, or microwave oven transformer.
Very lossy for a reason but good as an example as these have three core stems ;)
Now that you see that you will that a lot more transformers actually allow you to replace the two coils with 6 ;)
Ok, but why not use a rectifier first and not use a transformer or two at all?
The resulting output voltage will without a gear REDUCTION be much higher than what a cheap DC-DC converter can handle.
And at such speeds the effiency would be very bad too.
You would need huge capacitors of good quality to deal with the now more impulse like output.
And considering the primary side of the transformer does not require anything thicker than the wire on the coils of the fan...
Not hard at all to find some suitable tansformers to salvage - or to use some nice ring transformers ;)
No kidding around, just facts and possible options you might want to explore.

Does that now mean I get free energy?
Sure, if you mean you get the free nergy from the energy of the wind at no cost.
No if you think a ceiling fan could ever power your house.
Internal resistance, size and wire/connection properties set our limits.
Not to mention that they are designed to be dirt cheap.
If you are in a windy region and assume a realistic 300W minimum output from a 100W fan then adding more stage multiplies this.
These fancy upright windmills are not just powerful but also would allow to use one modded fan either end.
If big enough and with enough wind force throughout the year you could just add a second or third stage to ech end.
With 3 on both ends the resulting output would then be suddenly 1.8kW per windmill....
And all from scrap parts with only the costs for the magnets...
No wind? Then use water....
None of it? Get some greyhounds and build a big hamster wheel :)
You get the general idea I hope...


The forums are retiring in 2021 and are now closed for new topics and comments.

Reply 2 years ago

I did not say there are easier ways or motors that offer a better output right away ;)
Keep in mind though that most DC motors will require brushes.
My aim here was to show that there are quite a few options to increase the efficiency of such simple generators.
So, not really a thing to give you the best or most powerful generator from the scrap pile.
A great thing though if you want to learn more about how modern generators "cheat" to get better performance. ;)

I mainly wrote all this nonsense up because down here a lot of people still like the "off grid" idea.
They have a hut for hunting, a little beach box or small rural property with no mains supplies for water, gas or electricity.
A real DC motor is an easy choice that does not leave you with much to tinker.
But it also brings the problem of brushes - not always an option with bad weather and dust.
Since you mentioned the washing machines...
A lot of what I collected and tried for ceiling fans actually came from my early mods on these big ring type motors.
Only real difference is the size and amount of poles.
They do offer a far better output without any need to rewire the coils for this but at least around here they are now really hard to get from the scrap.
Copper prices made them a value target for the scrap dealers and you need to "register your interest" if you want to get one for yourself LOL

I think what I am trying to say is that if we torture our brains more often instead of using it to watch videos we might get more ideas ourself ;)
Nobody would expect that a home made and quite tiny generator would be able to provide 1 or 2kW with ease.
People prefer to pay a lot of money to buy one ready to install because it said on the label it has 2kW.
Over the years I refitted 4 old water driven generators with modified waching machine parts.
Mostly because the old ones were getting impossible to fix once age became an issue.
But one just for fun and to get more power.
Why might this be important?
You see a normal turbine generator has a fixed DC motor attached, I only found one that used an AC generator.
And all of them around here are limited to what our standard solar systems use - a max of 60V.
Ever tried to get 60V of DC from the generator to the battery shed that is 100 or more meters away? ;)
Not much left to work with :(
In the fun project we used 3 stator rings, shielded against each other and around the individual coils.
Magnets came from china and were only financial setback at around 450 bucks all up.
The original setup was like using a big V8 to power your chainsaw LOL
The turbine never saw any real load even with the generator at a dead short.
But then again it was only rated for a max output of 200W...
Under the same conditions our three stage waching machine setup produced just over 2kW!
And still the turbine would have been able to support two or three more stages.
Similar commercial solutions would have meant to replace the turbine for a much bigger model - not always possible and cetainly not feasable considering it would have cost us a bit over $3000.
So instead of providing a fixed 60V DC and accept the massive transmission losses we had three AC supplies is total sync.
At close to 400V on the output side the transmission lossed can be neglected if the transformer and rectifier with DC-DC converter is in the battery shed.
Even a standard mod of these three waching machine motors would have provided about 800W of power already, so a win-win on all ends and still far cheaper than any commercially available solution.