14-11 does 1000 watts.

[Jerry]

I did a truck mounted wind test of the 14-11 today. That is a total of 28 of the Lowes ceramic magnets, 44 three amp diodes and 11 coils wound with 400 turns of 24 ga. wire.

I was using my 4ft diameter plastic "Jerry Blades".

First I did the 12 volt battery load test run. Then I did a 1 Ohm resistor load test at 30 mph only.

Here are the #s. Battery load test first.

 10 mph  38.4   watts

 15 mph  76.8   watts

 20 mph  325    watts

 25 mph  378    watts

 30 mph  562.5  watts

 At 30 mph it did 1179.5 watts.

When I got back  from my run ( most of the 20 minutes at 400 to 500 watts ) I check the coils for heat. They were fairly warm but I could hold my fingers on them continualy.

Not a burnt 24 ga. wire 400 turn coil anywhere, not a shorted 3 amp diode anywhere.

Heres a few pictures.

The NEO 14-11 is going to be a beast. It will be the same speaker baskets and steel disc and weigh the same. But the power will be way high compaired to this unit.

                           Jerry

[GeeMac]

Holy Moly! That is ONE mean looking machine. May your wind blow at thirty more often.

[Jerry]

Hi GeeMac.

Its very large and heavy for the power it does. I think it would be concidered to be a 300 to 400 watt machine.

The wind speed was not accurate. I either had the truck speedometer or the hand held anamoeter on the hood. The wind at the genny could have been mudh higher mph.

However I wanted to use the same wind speed measuring set up that I used on the 4ft Hugh Piggot genny in my diary from a few days ago. Just for compairison.

This is a very large and heavy alt for the power it produces.

I built it just to check out the use of cheap ceramics, small ga. high turns wire, many small single diodes with no heatsink, odd coil count, iron in the coil centers.

It definetly does not go by the book. But it works quite well.

I will build the NEO version in the same 2 speaker baskets. It however will need a much larger blade and will look more in scale.

I'm thinkinh 7 to 8ft ?

Its all for fun and learning.

The 1000 wats was do to the resistor loading allowing the voltage to clime very high.

The battery real world load kept the wattage in perspective.

It was intresting that both machines did the same battery load power at 30 mph.

Could be the blade power limet?

                              Jerry

[ghurd]

Looks like something out of a Sci-Fi movie.

Keep an eye on those 3A diodes.

I figure 30MPH is over loading them by about 20%.

That is something to consider in the neo version.  I expect the neo version will pretty easily overload 6A diodes.

G-

[Jerry]

Hi G.

The NEO version will be a 48 volt so the current should be ok. But I might squeez the 6 amp in if I can.

I'm a littel puzzeld? There are 44 diodes. 4 for each of the 11 coils.

So how much current actuly goes through each diode? If 1 coil  produces 2 amps does each diode pass 1/2 amp?

Well the 6amp fullwave bridges on littel yellow deliverd 1600 watts and for several HRs. There is 9 of those. It was at 24 volts.

This is all great fun.

Jerry

[richhagen]

Hi Jerry, Two of your diodes are both carrying the full current out of the coil at any given time since you are basically hooking them up as a bridge rectifier for each coil.  Say your coil has one connection we'll call A and the other we'll call B.  When the magnetic field strength is changing in one direction A will be at a positive voltage relative to B, Only one of the two diodes connected to each coil connection will conduct all of the current.  When the magnetic field strength is changing in the opposite direction B will be at a positive voltage relative to A and the diode connected to each coil leg that was not conduction before will conduct, and the one that was conducting before will not, this keeps the negative output of the bridge rectifier negative, and the positive output positive.  You have two diodes conducting all of the current in series in your circuit whichever leg is positive.  So for your 2 amp example, each diode is passing (pulsing) all of the current from the coil half of the time.  Rich  

[ghurd]

Oh Boy Jerry,

I hope we did not open another can of "whoop azz", about diodes this time!

A 3A diode is rated 3A average current.

AC power has the current flowing back and forth, so half the time the current is going through the "other" diode of the pair on that output wire.

I had it figured as 562.5W at 15V, or 37.5A.

The 11 coils would be making 51W each.

Battery voltage at 14.4V, or 3.55A per coil.

I take the 3.55A output as the 'average'.

I could argue the average is a lot lower (about half), but I don't believe it quite yet because of diode failures.

It would be good to hear other peoples ideas about this.

I know I heard some before.  I can not recall what they were.

G-

[Ungrounded Lightning Rod]

Remember that the waveforms are peaky so the output current switches around from one set of diodes to another.

The average for a diode may be (average output current) / 2 (active half the time) * 2/(AC terminals) (i.e. only "on" about 2/3 of the time for three phase), but the peak current is significantly higher than the average when it's on, especially during the peak of the cycle.

Voltage drop across the diode is logarithmic plus the bandgap (which flattens it further) so average dissipation isn't raised a whole bunch by the peaks (as it would be in a resistor with its square law of power vs. current).

But a major failure mode for semiconductors is electromigration, especially in the metalization of the connections.  Electromigration is where a bunch of charge carriers hit an atom and knock it out of place.  This goes up steeply as current increases (because the atom will tend to fall back in place unless it's hit a bunch.)  So you really want to size your diodes for the peak current, not the average, if you want them to achieve their rated lifetime.

(Note that bridge ASSEMBLIES have their current rating appropriately derated from the ratings of the individual diodes for the intended service.  So a three-phase bridge can operate at its rated current but single-phase bridges combined into three-phase bridges don't appreciably benefit from the implied current-sharing and should only be operated at the rated current (two bridges with one hanging AC input or three in jerry-rig) or twice that rating (three bridges with two AC inputs per phase line and a GOOD heatsink).

[joestue]

Electromigration is not an issue, in fact you are off by about 3-4 orders of magnitude.

Sending 40 amps through a 3mm^2, 3 amp diode diode is 1.3KA/cm^2.

This is peanuts compared to the 10^6A/cm^2 required to even start electromigration.

Interestingly the last chevy alternator I had my hands on had silicon diodes that were about 3mm square, this works out to 1.1KA/cm^2 "average" at full rated load.

Practical single phase diode bridges in every consumer power supply face peak current levels about 10 times higher than average, this is standard practice. Diode derating has always been based on real life heat dissipation and junction temperature as opposed to theorized failure modes.

With 11 phases, the slightest line inductance or resistive coil losses will spread the current across at least 3 phases at any given time, but you need an oscilloscope to really know what is going on.

If you are looking for the reason why a 100 amp 1000 PIV diode bridge can fail on a 48 volt, 50 amp application, look no further than the fact that you have a 200 foot antenna attached to one end of it.