@gadgetmall,

Thank you my friend. Today I got the SK3606, they are a beauty. I immediately removed my diodes and replaced with the SKs. Right out of the batch I notice the perfomance is totally changed. I can run this with the fan very slow and I even think the sound is the same as you have. It runs with a strange fluctuation of 2ma to 10ma per turn. Battery voltage did not even blink from the resting voltage.

Definitely those diodes make a difference. Hopefully I will be able t tune it to the point of not only keeping the voltage stable but also going up. Lets see.

@all

Today before leaving to work, I replaced my transistor NPN-3904 with a TIP32A and tried different quick configurations. Before I did that voltage of the battery was about 0.01v down again but as soon I left the system disconnected while I changed the parts the battery came back to the same voltage of 4 days ago.

I tried a few tunnings and had to leave to work and asked my wife to keep an eye on the fan to see if it would run away too fast or stop. Later she called, about 3 hours later, and said the fan had stopped. So quickly I figure the pot resistance setting was too high, so I asked her to turn the nob. Oh boy, what a mistake. When I came back home from work the fan was simply running too fast and drained the battery about 0.02v above the normal 0.01v totaling a 0.03v down. 

BUT, i got the new parts from gadgetmall and disconnected everything. Guess what? battery in 5 minutes was back to the same voltage again and I put it to run with the new SK diodes. Voltage has not dropped even a tinny little bit with those guys!!!!!! AMAZING.

Off course now that I am here I tuned to my best and it was easy to tune it too.  Another thing different with the SK's and the new transistor is that the voltage on the CAP (which I had to install back) rises very, very fast to 100+volts, may be 20 seconds it will be there, which before with the other diodes and the PNP3904 was not happening that fast.

Fausto.

@ All

I also bought a 48 volt fan today but it's only a 2 pole bifilar (see pic), anyone know if and how to go about wiring this fan?  The last 48 v fan I had was wrecked when I brought it to a shop to press out the shaft 

Regards,
Paul

Oh boy what a nice widdings. Goat I think you are in trouble with this one here. What I can advice is, see if you can somehow find 4 wires ends on those metal posts. Measure the resistance in all sort of combinations and see if you can infer which one is which. I kind of did that on my fan and I noticed that two widdings are really one, as Fusionchip is saying.

From those two widdings you will have 2 posts and the other 2 widdings would be another 2 posts (or wire ends) and one widding will be infinite resistance to the other widdings, forcing you to see that they are really two sets of two widdings in the motor.

In nievesoliveras picture is what we see logically but I dont think it is really that simple. The SSG motor needs only two widdings really, one is the trigger and the other is the power (as nievesoliveras  shows) but with this fans they are somehow organized in two sets of trigger and two sets of power coils. Makes any sense?

Fausto.

@gadgetmall
I already built a bedini from a 12 V computer fan so I know how to do it now.
But this 48 V motor I bought is wired different it definitely has only one strand going from coil to coil.
Also one end goes to a pin but the other end goes directly into the circuit board. 
Totally different setup from the 12V computer motor I modified.
Anyone else buying these motors should try peeking into the crack of the motor to make sure it is bifilar wound.
Because some of them are wound with only one strand. 

@gadgetmall did any of your 12 volt versions clearly show more output then input?
Mine definitely runs the battery down as it runs.
I tweeked the pot. for hours trying different settings and always seen the battery voltage keep  going down.
I bought a few larger 12V fans to try building another one, maybe a larger fan will put out more power.
I looked through the crack in the rotor of them and can see they have bifilar windings on them.
Princess auto only carries the one type of 48 volt  fan and  it's not bifilar wound.

Well i didn't want to say anything about another one that selfruns BUT yes the first one i did in the video with the caps on the input with a 9 volt battery will run itself .thats all it does it dost drop and it dosnt gain . it was a 12 volt fan . Its resistance was the highest of 8 12volt  fans i modded . 91 ohms per coil.. . No i am not kidding . the other fans had around 30 to 40 ohms resistance and shorted out when i tried all of them . Weird ? I gave it away for Christmas . You have got to get some of these 48 volt fans shipped to you the link is a few pages back . they were only 1.79 each .i bet everybody and there father has bought all of them . I got 20 comming and i will probably sell My Completed units including the battery for around 1000.00  or you can make it but you wont get the diodes . i sold 16 of them in like 3 days on the net and i have about 20 more and i am using them in the completed units .. Fasto got the last one
Ps Can you Snap a Picture Of the Stator you have in question and post it here . ?

Thanks gadgetmall

There were 2 wires gold and red so I had to solder extension wires and made them yellow for the gold colored wires and green for the red wires (see pic below).

Notice that the wires have been soldered together to extend them, in the picture that I posted previously you could see that there were 2 short wires and 2 longer wires, now you can see in the picture that there is no shrink tube on the 2 shorter ones because they are under the back casing.

So tomorrow I will hook up the 2 yellow wires onto the primary side and the red on the secondary of your circuit if I can get a hold of the components that I'm missing.

Ok . glad you got it apart . check the continuity on the two sets of wire . solder for the 4th wire so you have 4 post and two circuits . should work fine.It appears to have two windings because it has two colors of wire therefor its bifilar.
Al

Edit:  Here's the specs on the fan, Hartiberlin this fan comes from your neck of the woods 

EBM W2G110-AM41-92
48V -(36-56V) DC
5,9W 3100U/min
= rot -shwarz
Made in Germany 12/93

Thanks,
Paul

Thanks gadgetmall

There were 2 wires gold and red so I had to solder extension wires and made them yellow for the gold colored wires and green for the red wires (see pic below).

Notice that the wires have been soldered together to extend them, in the picture that I posted previously you could see that there were 2 short wires and 2 longer wires, now you can see in the picture that there is no shrink tube on the 2 shorter ones because they are under the back casing.

So tomorrow I will hook up the 2 yellow wires onto the primary side and the red on the secondary of your circuit if I can get a hold of the components that I'm missing.

Edit:  Here's the specs on the fan, Hartiberlin this fan comes from your neck of the woods 

EBM W2G110-AM41-92
48V -(36-56V) DC
5,9W 3100U/min
= rot -shwarz
Made in Germany 12/93

Thanks,
Paul

Ok paul . remember if it dosn't spin reverse the trigger wires !

Hi Fausto,
just try also to put a neon bulb parallel with your transistor.
When the BackEMF happens the spike should also light up the neon
and thus the spike will go back to the battery via the neon bulb.

( Your circuit might have to be modified a bit.)

This ( transistor plus neon bulb in parallel)  is a trick to simulate a mechanical switch,
where you also have a plasma sparking when you open the switch ( transistor)
and through the plasma goes the BackEMF
current spike back to the battery and will recharge it.

Good luck and happy experiments.

Regards, Stefan.

@ Gadgetmall:

I just ordered some of those fans from the link you posted.  I am going to give this a try.  They will probably be here in about a week.

Bill

@Hoppy,

oh boy, I am Sooooooooo happy that I am not craizy here and gadgetmall is becoming more and more correct (which I never doubted anyway but it is good to have external evidence).

I am having all sort of tuning points that behave in vast consumption levels. Some will maintain even my SLAs on a long time without dropping voltage. Some will keep the Wet cell battery virtually unchanged and some will simply drain it all.

It is indeed a matter of a sweet point. I also think as you Hoppy that it is the magnetic effect somehow influencing this. I am playing a lot with different positions of a magnet on top of the fan too.

Different transistors will also in a major way change the radiant content. My 3904 is very good but difficult to tune. The TIP32A is very easy to tune but consumes a little bit more. The MJL21194 (if I am correct) and the 3055 are the worse.

I am going to open another 48v fan tonight, take pictures and try a second circuit tonight.

Fausto.