How to install and tie together magnet wire coils

[aidrenegade]

I suggest using the rectifier referred to above at a single coil output first (obtain at least 10A 40V Schottky diodes, from computer power supplies etc) and load the DC output with power resistors (but 12V car headlight bulbs may be used, remebering they are nonlinear), you may wish to use a DC voltmeter in parallel with the resistor or the lamp load and a separate DC ammeter in series with the resistor or lamp load. For the two electrolytic capacitors shown in the link, I suggest at least 10000uF or higher, with 25V or rather 35V DC ratings. 

I've got to order these bits online as the local Maplins shop doesn't carry them. Guya, can you confirm I've got suitable components before ordering them via these links:

Schottky Diode 45V, 16A (not sure about the 'to 220v' bit): http://www.conrad-uk.com/ce/en/product/163689/Vishay-MBR1645-SCHOTTKY-DIODE-TO-220AC-Schottky-Diode-45V-16A-TO-220-AC.

Elecrolitic capacitors: http://www.conrad-uk.com/ce/en/product/446191/15000F-Electrolytic-Capacitor-10mm-Vishay-2222-058-57153

Thanks.

[gyulasun]

Hi aidrenegade,

I would suggest using several smaller capacitors in parallel, this has two advantages: cheaper overall price and lower ESR as a result (ESR is equivalent series resistance of any capacitor, the lower this value the less loss is created inside a capacitor).

So at  your local Maplin you may get 10 (or more) quantity of a 4700uF 35V (105°C) for  Ł1.15 each, (quantity dependent), order code N97KF

So you may wish to parallel 5 such 4700uF and call it as C1 group and paralell the other 5 and call it as C2 group.  One such group would work as one of the caps in the rectifier circuit I referred to earlier and the other such group would work as the other cap there.

Regarding the diodes, it is a good choice you show (TO 200AC refers to the case style) but you may find similar at Maplin, see MBR3045PT  (45V 30A)  price  Ł0.99 if you buy 20 quantities it is Ł19.80 order code N18CC   (you need 20 because of the 10 coils)   Notice that these are dual diodes, actually two diodes are manufactured into a single case with common cathode and you can connect the two anodes together to get a single diode from the two, this decreases the overall diode loss some percent (7-8%) too.  (Data sheet for the diode: http://www.onsemi.com/pub_link/Collateral/MBR3045PT-D.PDF )

I think if you can stay with the 10 coil - 12 magnet pairs it would be better than using equal number of coil-magnet pairs. Further if you could keep the safety tapes wrapped around each magnet, your setup would be much safer at higher rpms.

Gyula

[aidrenegade]

Dam it! Build quality let me down! At about 700 rpm one of the magnets came loose, took out 2 coils and loosened another 2 magnets. And on the day my diodes & capacitors turned up in the post too. Wish I had maybe £500 or £600 to get an engineering firm to build a decent, ballanced engineered part of my build to take out the weak areas! So, 2 new coils to wind and 2 magnets to re-fix into position (guess I need more supa glue....)

[aidrenegade]

OK, I finally built a new assembly to hold the magnets (see pic below). I went for 10 pairs of magnets as I have 10 coils but I'm still producing ac voltage, 6v at 532rpm. The whole thing does now seem more balanced in operation and a bit less noisy. Time to install the diodes / capacitors and get the voltage doubled and converted to dc.

(QUOTE) "Notice that these are dual diodes, actually two diodes are manufactured into a single case with common cathode and you can connect the two anodes together to get a single diode from the two, this decreases the overall diode loss some percent (7-8%) too."

Does this simply mean solder the input side of the magnet wire to both outer legs as per the wiring diagram?

[aidrenegade]

So I've fitted 5   2 diode full wave rectifiers / voltage doublers to 5 of my coils. Needed to order a few more capacitors to finish the job but started to experiment with these 5 outputs connected in parallel. I'm finding it really difficult to find the 'sweet spot' with the 400w rated dimmer switch. Momentum of my motor / magnet arrangement combined with the slower discharge of the capacitors is making using the dimmer switch to 'fine' control the motor feed voltage to control the dc output difficult.

Any ideas for a voltage control system to replace the dimmer switch are welcome! As I type, I've been trying to find the perfect setting of the dimmer switch, it seems to be settling down but it's already taken 1 1/2 hours! Using a dimmer switch may infact be a built in fault if it can't output a finer level of voltage. Add to this the fact that the motor is controlled from a UK 240v ac mains (at present) when my inverter will have a 230v ac output I see a lot more swearing and knashing of teeth ahead!

Currently output voltage is varying between 12.25v and 12.68v (dc) but it may still be slowing down over all. At least I know the fan motor will keep going and no where near it's 240v max imput voltage. Not going to waste time trying to actually measure this voltage till the rest of the full wave rectifiers are fitted. In regard to your typical 12vdc / 230vac inverter, the book doesn't state a +/- tolerance for the 12v dc input, does anyone have any ideas on this tolerance?

As always, in advance, a big thankyou for your input! Don't be shy with your comments, one may point my mind in another direction of thought that you were not thinking of!