Introduction: Solar Spinner - Magbot (200 Rpm With No Motor)

Picture of Solar Spinner - Magbot    (200 Rpm With No Motor)
Cool Solar Spinner from Junk

Total Spend: US$ 0.75 that's right, 75 cents! everything else was recovered from old electronic junk

This solar powered ornament spins at up to 200 rpm in direct sunlight and was made in about 5hrs.

2 spinners in early morning sun. This instructable is for the rear one.

1. Strip of circuit board
This was pulled out of an old stereo, and was used to mount a series of small buttons along the face, I would think that any low profile electronics would have something similar such as CD/DVD players

2. Inductor coil
I recovered this from an old 5.25" floppy drive. There were 6 on the circuit board. I think 3.5" drives may have something similar you could use. Failing that, buy a "Major Henry Coil" from

3. 2 x decent computer fans
When I say decent I mean ones that will have smooth bearings in them. Cheaper fans don't have bearings and are no use for this project.

4. Neodymium Magnet from old hard drive
Recovered from an old hard drive, this magnet is super powerful.

5. Solar powered Led keyring torch off e-bay
These are really quite cool. I bought a couple of these off e-bay for 75c each including delivery from China. I used the solar panel, a LED, a diode and part of the case.

6. 3904 and 3906 transistors
These are pretty common transistors and can be found in a lot of electronic circuits. Recover these from old circuit boards if you can, or if you have to buy them they are very cheap.

7. 3300-4700uF Electrolytic Capacitor
Very easy to find but if you can't, make up around 4000uF by putting smaller cap's in parallel. I ended up using 3 1000uF caps instead of the 4700uF pictured.

8. 1000uF Electrolytic Capacitor
Same as above

9. 1n914 or 1n4148 Switching Diode
Easy to find, they look like a little glass bead that's red inside and has a black band up one end. I found one inside the keyring

10. 2 x 100k Resistors (Brown, Black, Yellow, Gold markings)
I was lucky enough to be able to pull these off the circuit board above.

11. Sheet metal for a base.
Back panel from an old stereo in this case.

Step 1: Keyring Destruction

Picture of Keyring Destruction

I recently puchased 2 "Solar Power rechargeble LED Flashlight Torche Key chain"
from e-bay for AU$1.00 each including free postage. That's about 75 cents US

After about a week they arrived from China and I opened the packaging expecting
to have bought a couple of pieces of garbage. But was pleasantly surprised to
find a couple of quality items.

The LED's were quite a bit brighter than I ever expected and after pulling one
of the keyrings apart I found that it was a REAL solar cell. (I've picked up a
cheap solar powered calculator before that after pulling it apart, found that
the "solar cell" was just some painted plastic)

Desoldering was extremely easy as everything (even through hole components) was
soldered only to the surface. i used a hairdryer to soften the glue holding the
solar panel to the circuit board and was rewarded with..

3 bright white LEDS
1 germanium diode which I'll use in the circuit
1 2032 Lithium Ion 3.6v rechargable battery with handy tabs for resoldering to
something else
1 nice momentary microswitch
1 low light solar panel giving 4.5V and 10.0uA at 200 lux

a quick google search on the markings and model number ST-3722-9
revealed this information.

Step 2: Computer Fan Destruction

Picture of Computer Fan Destruction

Removing the parts we need from the computer fans

Peel back the sticker on the fan, and there should be a small split ring keeper. This can be popped off with a fine screwdriver or knife. (keep it)

The fan should slip out of the housing now revealing a bearing, washers and the electronics. If there's no bearing, put this fan back together, and find another fan to destroy :)

Hacking into the fan bit with the dremel, I was able to cut out the axel which will be used later.

Step 3: Making the Stand

Picture of Making the Stand

Using a dremel, I marked out and cut the stand from the sheet metal.

I cut a small indent along the base where the folds go to make it easier to fold.

I then removed all the components from the circuit board strip. Instead of using solder wick or those solder sucker things, I just heat the solder and flick it against the table edge. Most of the solder comes off. Luckily it's a glass table and the solder just scrapes off it with a finger nail.

Step 4: Building the Circuit

Picture of Building the Circuit

I originally found this circuit whilst browsing through the catalog at I have reproduced the circuit here with their permission. Please visit their website to get any parts you might need to create this.

Especially their Major Henry Coil, which they had specifically designed for this type of application.

Now, back to the build.

The layout of the circuit on the board is extremely important here, I needed to put as much of the circuit up one end as I could, and the coil, by itself down the other. This is so I could balance it all out around the axel when I went to mount it on the stand. I wasn't able to do too much about it in this case because of where the holes in the circuit board already were.

I tried to use the existing holes and the pre-existing printed circuit where I could. It actually worked extremely well.

If the copper traces were not going where I wanted, I carefully cut through them with a cutting head and a deft touch using the dremel.

Where I needed to join pre-existing traces together, I carefully scraped back the protective layer revealing the copper beneath and then joined them together using shielded copper wire salvaged from a second induction coil.

Any fine wire from motor windings or inductors will do. Just carefully tin the ends of the wire with solder. To do this, I put a bead of solder on the tip of my soldering iron and slowly pushed the wire into it. The outer shield melted away leaving nicely tinned copper.

Ok, finished and WOW, it worked, first time! Without mounting near the magnet, if this circuit is correct, the LED will faintly start to rapidly flash (something to do with "Back EMF" and the earth's magnetic field reacting with the coil) To be honest I was very surprised, as I didn't breadboard the circuit first and I was using untested components such as the diode from the keyring (I assumed it was a switching or signal diode like a 1n914 or 1n4148 but wasn't sure).

For a full description of the circuit and how it works, download the sunswinger kit pdf from the website.

Step 5: Creating the Bearing

Picture of Creating the Bearing

So I've already freed an axel from the Fan, now I need to mount it to the circuit board.

The positioning of the axel on the circuit board is very important and will affect how well the spinner will perform. It has to be as close as possible to the center of gravity of the board AND still over half way from the coil. This is so the capacitors up the other end aren't attracted by the magnet as it spins around.

Luckily the 3 holes for the Led's in the keyring were exactly the same diameter as the bearings from the fans. After cutting the plastic down to size and fitting the bearings I mounted them onto the stand with hot glue. Positioning of the bearing on the stand is also important.

Hard Drive Magnets have the North and South poles on the same side, so we want the coil positioned so it is only over one end of the HD magnet. If everything works, the contraption should start moving immediately. Now we need to do some fine adjustment. With the coil over the magnet, take note of weather the coil is attracted or repelled when it "kicks", if it is repelled, the magnet needs to be turned around so the other end is up. There are 2 reasons, the machine will work better, and when magnets are repelled from one another they are slighly weakened. (and when they are attracted, they are strengthened)

To finely tune the balance of the circuit around the axel, I had to put some BIG dollops of solder up the capacitor end on the back. This served 2 purposes, it allowed the swing to travel further from eachg kick, minimising the number of kicks needed to start spinning, and also allowed more time between kicks for the solar panel to recharge the capacitors.

Step 6: Final Thoughts.

Picture of Final Thoughts.

This was a resounding success. Built in 1 afternoon, instead of the week and a half it took me to do the first one. The base is stronger, the bearings better, etc

I really enjoyed trying to use the preexisting circuit board traces for this circuit. It made for some interesting design decisions.

I liked how most of the electronics came from the keyring and the circuit board strip. I have a suspicion that the transistors in the fan electronics could have been used here as well, but it's hard to know what SMD components are, as there isn't an industry wide marking scheme. It's very hard to salvage and even know what SMD components are on a board.

I'm not to happy with the strength of this one. The first one I built is much more powerful. To address this, I think I will increase the total storage capacitance by either adding another cap, or replacing the ones that are there with caps of a larger value. I will also add another solar panel from the other keyring, to double the amount of charge that can be stored in the caps between swings.

This solar panel is very low current compared to others that are available, replacing it with a poly-crystalline solar panel should increase the available current.


dzhang (author)2009-06-12

I couldn't find any inductors like that in a 5.25 floppy or in a 3.5 floppy that i took apart. I really don't want to buy an inductor cause i have all the other parts, so could you tell me how big your inductors were and how many ohms?

zozzles (author)dzhang2009-07-08

The "Major Henry Coil" from is an extremely good piece of equipment, with a rather low price, The catalog states a "measured inductance of 1.03 millihenries" but to be honest, the strength of the magnet is just as important. Besides which, for this circuit that part of the design is somewhat overkill: It ought to work with parts of less quality. In a way, getting it to spin over the top the first time is the true test of things. After that, you've got a positive feedback loop going that is only limited by the bearing - and get this - the air resistance of the rotating arm! The triggering voltage to dump the current is usually at a hair's level from being high enough (almost twitchy, in fact), and you probably have more of a hassle coming up with a really decent bearing (i.e. low friction) than worrying about how incredibly spiffy the coil is: when the circuit DOES dump the current, it is going to yank the magnet toward it or the other way around. It is obvious that you've got the timing right. As noted, once you've got the (magnetic) polarities right, it works great. Although I wish that some of the photographs had been taken with slightly more light, this is an incredibly clever Instructable (and re-engineering of the original SolarSwinger) and I'm extremely grateful that you've shared it with us. I have no personal interest with Solarbotics whatsoever, but I've had extremely good luck with their service since 1997 or so.

FreeBaseBuzz (author)dzhang2009-06-14

I don't think the internal resistance of the inductor is really important for this circuit. At a guess I'd say the coil is about 1 inch across and around 300-400 turns. You could wind your own... Hey, send me a personal e-mail to at I may be able to send you a spare one I have in the post.

maurice1993 (author)2009-05-27

I'm sorry, I already didn't read your 'ible, but, looking by the pictures, this could be used has a ventilador, isn't?

FreeBaseBuzz (author)maurice19932009-05-27

possibly.. I don't think it'd be too efficient though. it wouldn't spin anywhere near as fast with the drag caused by trying to push the air.

maurice1993 (author)FreeBaseBuzz2009-05-31

yup, but maybe, if I don't twist thw suport so much, just a little, the drag it'll not to be a spin problem, perhaps the air will not be pushed very well, but for a air recicler, I guess the air pushed it'll be not a problem... but I guess I never will had the answer if I don't make one for me...

Bongmaster (author)2009-05-24

kool stuff :) I got some of those exact same torches :) one for keeps and 2 for the parts :P needed the panels for a solar fled engine ;) and those were the best and cheapest i found..

FreeBaseBuzz (author)Bongmaster2009-05-24

Yeah, they are pretty good for the price, but I think you will find that they probably don't provide enough current to be useful for solar engines for bots. Your FLED engine will work, but it'll be a bit of time between triggers. These provide 10uA and you really need something in the mA range. I think these panels would only be good really for this kind of engine or building pummers. If you are doing motorised bot's, spend the money and get the solar panels from, you won't be disappointed.

Bongmaster (author)FreeBaseBuzz2009-05-27

well 2 of them panels seems to do ok on the one i made. but i have a few test solar engines in the make. seeing wat comes out best :3

Tall_Tinkerer (author)2009-05-26

Great instructable,very well written and illustrated.

About This Instructable




More by FreeBaseBuzz:Solar Spinner - Magbot    (200 rpm with no motor)
Add instructable to: