Hey Deep

Put the diode in parallel on each coil..  Say if you can see the Pos input to the coil, the cathode(silver bar end of diode) of the diode to that side of the coil. the diode wont conduct till the drive is disconnected.

 This will allow current to continue in the coil for a bit beyond the time the drive current is disconnected.

More rotation benefit without using more input. the inductor acts as a flywheel.

Ill draw it up after work if you dont get it by then.

=]

Mags

Hi Mags, thanks, sometimes i have a really hard time seeing the bleeding obvious !

I still don't think i get it so here is a picture of what i tried, what am i doing wrong ?

WOW !

Thanks Mags what a simple but effective little trick, who invented that ?

I've only done it to one coil and i've got 1432 RPM instead of 1417 RPM with the same current draw.

Luckily i have exactly three more 1N4001's ! I will stik them all on then post results.

Thanks Gyula too

**EDIT ADD**

OK the first one i tried was on the last (4th) coil in the series and as i say above it was good.

The second one i tried, in the 3rd coil in the series, i got less RPM for the same current draw.

I added a third diode to the second coil in the series and things slowed down dramatically and my current draw doubled.

I have checked all the diodes and re-checked the coils and all are functioning.

I have also tried them one at a time and the only one to give any gain is the last coil in the series.

The 3rd coil slows it down, the 2nd coil nothing changes !

One thing i've noticed, the coil that performs well with the flyback diode (the 4th coil in the series) is about two-thirds of the thickness of the other three, would that make a difference ?

What could be the problem here ?

Im at lunch

So I assume you have all the diodes facing the same direction in the circle of coils. If some are wrong, the drive current will go through the diode and not the coil, thus increasing current and slowdown.

If all are correct, then I suspect some diodes are shorted.

I am doing this as we speak, with 3 coils in series and 6A diodes(RS) across each coil and I get the increase with each and no problems.  You can try 1 diode across all coils as a whole, but the diode may have to endure higher voltages.

I suspect you have a few bad diodes or the current/voltage limits of the diodes are being breached.

I can show a simple vid later after work.

I have no doubts that this works 100%   ;]

Mags

Hmmm   

doesnt make sense..   I would very quickely try the other coils with the diode in the opposite direction, just for sake.  Do you have a pic of the setup so I can see when I get home later?   

If you dont get it , ill be back on this evening to see  what is up.
Are the coils that different from each other?

Something is strange..  My coils are beefy.  .44ohm 18awg 2.5in dia 1.75 thick  around 2mh  in series. I do the diodes when in parallel also..   

This is a way to get more to the rotor by using less than normal. We could capture the excess, but instead we are applying the extra into getting more motive force to the rotor that the gen will work with.  Its all about doing all we can to increase efficiency.  ;]




Mags

Three of the coils are 800 turns bifilar, trigger wire is finer than power wire, can't remember gauge maybe trigger around 0.25mm and power 0.35.

The fourth is about 600 turns of same.

My generator cores arrived today and they are the wrong bloody size, my fault, i was going on bolt diameter not shield diameter ...

Here is pic :

http://qvision.pwp.blueyonder.co.uk/setup.JPG

I have a callout so won't be back until 10pm.

Hmm  maybe if trying the diodes in the other direction doesnt work, then maybe there is something up with bifi that this doesnt work...  dunno   I cant fathom why.

Mags

Well   something simple to try.  Undo the bifi and run the coils separate strands in parallel.  If you only have 3, try those in unbifi mode, in series. 

Your power in will go up and rpm. But it gives you something to try with what you got.

Maybe you get over 2000 rpm. maybe more than .97w but then add the diodes. Just to see what you get as compared to before.  Be careful at any high rpms. At least goggles.   ;]

Was thinking, the possibility that the diodes kept the coil running too long, due to the config of the coil in this case.
If you have a scope and possible way to shorten the pulse to the bifi, its very posssible that you could use a lot less power and the freewheel has time to stop before the next mag comes and is not forcing the rotor in reverse, as would be if the coil is still spinning.

Sounds nuts, but if you have those diodes in the right way, it is very possible that the flywheel is going still, when the next mag is approaching and cause slow down. And higher current, cuz the thing is still spinnin, new current can flow more and easy from the source.  There is no git up lil doggie on the coil.  No impedance to kick into a gradual current increase.  Its already goin.  Gotta scope this if you can. 

This is all theory, but they are things, clues that you gave me, that I question, and these are possible answers.  ;]

Just to get some experience with it.   ;]

Im going to try some things while I have my coils up.

But try the parallel on the bifi's  If you have a way to lower the on time, this will reduce input. Then diodes. I suggest 600v n up and probably a couple amps to try.

Mags