Bedini Brushless Fan Conversion

Nov 26 2016, Update
May 11, Updates
April 23, Updates

10 March 2013 at 13:22

This is the second time I replicate the Bedini brushless fan motor-generator
but the first time I used a rather small DC fan and was very hard to work on
with my thick fingers, this design was claimed to be a self runner but I was never
able to get my replication to self run.

A couple of weeks ago I was given 2 scrap G4 Mac computers and salvaged the nice big DC fans
in them along with a bunch of other components.

The fans are easily dismantled and were just a joy to work with, so much so that I connected each coil to
pc mount dual connectors which were installed directly to the back of the fan bearing mount.

I started the build yesterday but had to cut it short due to prior engagements, the entire build
took a total of about 3 hours and it started up on the first try.

Here's the schematic:

Replication #2:

Back View Side View Batteries

Charge battery voltage Cap voltage with load Cap voltage without load

A quick video:

Battery swapping is performed when either the Primary battery drops to 6.2vdc or the Charging battery reaches 7vdc.

The 2 batteries, Bat-1 and Bat-2 are brand new and have been charged at the same time by the same conventional charger,
battery specs are as follows, 6 volts at 4.5 amp hours each.

1st test run:

Primary Bat-1 starting voltage no load = 6.65vdc
Primary Bat-1 starting voltage loaded = 6.39vdc
Primary Bat-1 amp draw = 200ma

Charge Bat-2 starting voltage = 6.66vdc

Primary Bat-1 end voltage = 6.19vdc
Charge Bat-2 end voltage = 6.95vdc
Run time = 7:58:20

Input Joules = 38,458
Output Joules = 37,138

Terminal voltages are gonna be taking up too much time so minimum voltage will now be 6.3vdc and max 7vdc, also I will be using a computerized battery analyzer (CBAIII) to monitor primary discharge voltage and time duration.

New Setup:

2nd test run:

(NOTE: there is a major difference in voltage readings between my DDM and the CBA, DDM reads 6.41 volts and the CBA reads 6.36 volts on the same battery ???)

(SECOND NOTE: I've noticed since the start that every once in awhile during testing the fans rpm's drop for a second or less and amp draw also drops. I've checked all connections and even man-handled the unit while in operation and no drop in voltage so I decided to check to see if there is a voltage drop on the cap but nothing ???)

(THIRD NOTE: 48 minutes into this test run the fan suddenly dropped to half its usual rpm and the amp draw dropped to under 100ma and this time it stayed there so I fooled around with any and all soldering points, wires and found when I played with 1 of the connectors I installed on the back of the fan it sped up to its normal speed. There is no faulty connection on the outside so it must be the coil connection on the inside, I'll finish the second test run then I'll take the fan apart and check the coil connections.)

Primary Bat-1 starting voltage = 6.36vdc
Primary Bat-1 amp draw = 200ma

Charge Bat-2 starting voltage = 6.34vdc
Charge Bat-2 charging amps = 3.7ma

Primary Bat-1 end voltage = 6.2vdc
Charge Bat-2 end voltage = 6.35
Run time = 109 minutes

Discharge Output Joules = 4,159
Charge Input Joules = 8,293

(NOTE: I took the fan apart and found one of the coil connections had a cold solder joint and repaired it and added a little WD40 to the bearings.)

3rd test run:

While playing with the rpm's of the fan blades I noticed that a much higher voltage is produced on the cap at lower rpm's, so I am adding a 1k variable resistor between the negative rail and pin 8 coil for better tuning I hope.

I've switched to 12 volts cause 6 volts was not showing any measurable charging and when I adding the 1k pot it really lowered my primary amp draw, my analog amp meter read 200ma but the DDM read 225ma now I can't even use my analog meter cause its too low but reads 62ma on my DDM, also the charging amperage went from 1.74ma to 19.45ma at a much higher cap voltage.

My testing method sucks !!
Lets make is simple, I will now swap the batteries when the primary battery's voltage drops to 6.10 volts and I'll keep swapping them till either there is not enough charge to surpass 6.1 volts which would mean that this is not a self runner.

UPDATE March 13, 2013
After 465 minutes run time neither batteries will hold a primary voltage above 6.1 volts therefore self running does not seem possible with this design.

UPDATE April 23, 2013

I've taken it apart once again because I didn't want to scrap this nice little fan so I was turning it into a Bedini type battery charger when I noticed the fan speeding up then slowing down when charging batteries.
It seems one of the coil connections had a cold solder again so this time I soldered everything with flux and an extra hot iron checking all connections.
It now purrs like a kitten and I'll have to reinstall the cap dump circuit that I removed, once that done I will restart testing again.

Test 1:
After installing the cap dump circuit testing was started again with two 6 volt batteries, one charged to 6.69vdc (primary) and the other was left discharged at 6.09vdc(charging battery). 48 hours into testing the two batteries fell to under 6 volts, therefore self running was not achieved.

Test 2:
Now testing with 12vdc primary.

Unfortunately after approximately 53 hours both batteries no longer had enough charge to sustain my baseline voltage which was 12 volts dc. For awhile it did look like the higher operating voltage would make a difference but that quickly disappeared and so did all hopes of a self runner.

After four different series of tests, this design did not self run, great care was taken to use exactly the same components as Mr Bedini did but our results differ greatly !

One thing I can say though, is that this is one of the best battery chargers I've ever built and the addition of a switch to toggle from conventional charging to radient allows you to desulfate your batteries along with a radient charge.

UPDATE May 11, 2013
I'm making another bedini fan but this time with 7 strand litzed wire, wound 80 times on each coil which will give me a total of 560 turns per coil once all strands are connected start to end.

UPDATE Nov 26, 2016
Rick Friedrich, one of the people that has worked with John Bedini recently published a video about the fan kit they sell and disclosed a possible way to get over-unity with this device.
Heres his video explaining how its done.

Cheers Hitman