One more -

are you calling a top and bottom pair, one coil?

Thanks, Penno

the coils are conencted in pairs, top with the bottom but when i am talking about coils I am refering to each individual one.
The current setup works with the driving coils in attraction to the magnets. I have originally tried in repulsion but this one works better.

@romero uk .The device that you have built could be far more important than you are aware .One key question is , how complex are the electronics . If you are certain of your measurement techniques , then you have a proven COP of nearly 2 . Even if this can not be looped , look at the significance in small scale wind power . Imagine DOUBLING the output of a small scale wind plant . Or a solar array . Applications are suitable for both suburbia and third world villages . In a solar set up ,you can double the energy the solar panel put into the battery . And Double it again as it leaves the battery to feed the load ! Hows that for compound interest ! If the tech is scaleable , you could run machines in series , so that the energy increases in an exponential manner . Given enough info , I would definitely replicate this one .So back to the electronics . How about a schematic . Or why don't you sell detailed plans , and ultimately Kits?If this is real why are we wasting time on less attainable technologies?

This is the parts arrangement in my Muller setup.
I had few questions from the 'replicators' here about voltage on the battery going down when the load is on.
The reason for that is that from the output bridge rectifier I am using one diode to send power back to the battery to keep it charged.When the 20watt bulb is connected the power going back to the battery is reduced resulting that slight voltage drop.
This setup was built for about 25watt load, if the load is increased then all setup must be changed.
The gap from the coils to the rotor must be changed and some other things ...
If anyone is trying to replicate this please remember to do the testing with the load on and compensate the drag with the magnets on top of the coilsI hope this helps.

All the best,
RomeroUK

Further to my last post . It looks like you have 2 dedicated driver coils . The inputs to those coils are triggered by hall effect sensors , So the input to the device is pulsed .The big question is , does that ,or does it not , cause problems in measuring the input current .In other words , can the multimeter that you use to measure input current give a true reading on a pulsed DC Waveform.

@ROMEROuk

i remember that all or most "overunity, the guys have find because wrong "metering" !!

Think about that ALL mesurements  from NON-DC  as also NONSINUS-odial-WAVES  are with wrong / failed results.

As "unprofessional, you have only ONE WAY to see and result.
To have more output that the Input need:

Use 2 packs 12v 2,2Amp Lead-gel batteries
or 2 packs with 10pcs 1,2 AA NiMh AA batiers.

One to drive the INPUT  another to charge
if possible also lightning your bulb.

Use an switch to swap the 2 battery packs
In to out,  so  if that is running and lightning some
days.. THAN (only than) your circuit is UNEFULLY

Gustav Pese

Answers to your questions:
Only 2 coils (pairs) are driving.
All other coils are connected to bridge rectifiers, each set of coils goes to a rectifier then all are connected in parallel not in series as someone said before.
Each coil has 300 turns of 0.8mm multistrand, one set of coils is 600 turns in total.
The core is ferrite 6mm diameter 15mm long.
Lately I had some messages saying that this is a fake, some kind of trick I do... what can I say?
...wait until I will have it self running then who knows what else will come.I finished this project about one month ago and I had it tested then I decided to post the video.
I had it working for days without any recharge to the battery.
So, my friends, if you think that this is a fake please ignore my posts.

All the best,
RomeroUK

@gyulasun
I am going to buy that regulator from Maplins tommorow.

Thank you,
RomeroUK