Rectifying 3 phase

[RE Pod]

Is there any difference between rectifying the output from a 3 phase pma in either of the following 2 ways?

A --- Connecting the stator in star configuration, then rectifying through 2 bridge rectifiers.

OR

B --- Rectifying each phase through a seperate bridge rect and then connecting the DC outputs in series.

I would assume the only difference would be some extra voltage drop due to the extra bridge diode. Is that correct or have I missed something?

Many thanks.

Regards Terry...

[drdongle]

B will gain a higher net output voltage, yes you will have more rectifier loss about 2.4 volts more then 3 phase loss.

[finnsawyer]

There is, in fact no law that says he needs to use a bridge rectifier under B.  He could use only a single diode for each winding.  This cuts the voltage penalty down to 0.6 volts rather than 2.4.  If maximum voltage out is the aim, rather than maximum current, then this is the way to go.  Remember, though, that under B you also need one electrolytic capacitor for each winding.  Each capacitor can charge up to a voltage which is nearly equal to twice the peak voltage minus the diode voltage.  But this is true only if you don't draw too much current.  There are other design considerations that come into play.  The larger the capacitor value, the larger the current spike through the diode, which conducts current only for a short time.  The diodes have to be sized to handle that.  While the average power dissipated by the diode may be low, the instantaneous heating at the diode junction may be enough to destroy the diode.

[terry5732]

Most heavy diodes don't like fast cycling and heat  up. If you are rectifying multiple cycles with a single bridge you need to consider that the diode duty is cumulative of the cycles. So if say you have 100-200 cycles per phase, with three phases you are 300-600 cycles. They aren't meant to be heaters and don't last long. I've had bridges with very little power through them heat to where you can't touch them without blisters at as little as 1000 cycles.

[drdongle]

 Using a single rectifier while producing less drop will half the total output, pretty much defeating the purpose putting of the windings in series. To maximize output you have to use a bridge( and cap)for each winding. Depending on the voltage output of the windings the rectifier drop may be negligible.

[RE Pod]

OK.  It looks like there's more to this than I first thought.  My 3 phase pma will be charging a battery bank at 12V.  I think I'll connect the coils in star config and then rectify through bridge rectifiers.  I'll probably have to parallel 2 or 3 sets of rectifiers to be able to handle the amps.  Heatsink too.

I'm learning a lot of this as I go. If I knew a few weeks ago, what I know now.  I'd have gone the 24V route, but I've ordered my winding wire now, so too late to change.

Maybe after a while of running the 12V stator, I'll make a 24V stator.  Amps can be quite high at 12V and buying parts to cope can be expensive.

Experience is a good teacher, but can be hard on the wallet.  At least I'm having fun.

Regards Terry....

[Flux]

Rectifying phases and adding them in series may be a useful trick to get something out of a small motor that is too fast in its original form. It is not a practical way to produce serious power.

You should use star connection and a 3 phase bridge( which can be made from single phase bridges or 6 individual diodes).

What alternator are you making, if it is Hugh's 5 phase design you can convert to 24v with the same wire if you connect opposite coils in series and make it 5 phase with 5 leads instead of parallel 5 phase with 10 leads. You just have to be careful how you connect it, I think Hugh did it the 10 lead way so the connections can't be got wrong.

Flux

[tecker]

 Rectifying single phases may be advantageous if the geometry of your stator or rotor is not ringing in and the voltage of a coil combination anded together is parasitic . If the coil and magnet alignment is correct the drive factor of three phase is very effective in driving loads this is why you use that configuration .As far as diodes go the PN junction is not set up to pass power but is a necessary part in dealing with the power storage that is at our disposal (so bigger is better) . If you want a good power transfer with individual rectified phases or coil outputs add a cap with the highest capacitance and in the working voltages your are needing (the higher the rated voltage the less the leakage).Remember that all these components are susceptible to static and if your going to survive a routine lighting discharge you need to limit the components exposed to this out door environment.

[finnsawyer]

It depends on what you mean by output.  This is pretty standard power supply stuff.  If you assume the diode voltage drops are negligible, then the voltage from one winding across its capacitor will rise to the peak to peak voltage of that winding whether there is one diode or two.  If you subtract the diode voltages then the capacitor voltage will be greater with the single diode.  This may become significant if you're trying to get a trickle charge to a battery at the lowest possible rpm.  Is 1.8 volts significant?  While the full wave rectifier will give two current pulses per winding per cycle, it is not clear that this is necessarily an advantage as the single diode set up only needs to double the charge per pulse to give nearly the same result.  The diode voltage will be only slightly larger per pulse and the capacitor drop between pulses will be only slightly greater.  Presumably the voltage drop in the winding will not be significant at low currents in either case.  My emphasis was on one getting the largest voltage out not the greatest power.

[deloiter]

I hope I don't reveal my ignorance with this question, but here goes. How can the voltages be additive when they are 120 degrees out of phase?

[Ungrounded Lightning Rod]

I hope I don't reveal my ignorance with this question, but here goes. How can the voltages be additive when they are 120 degrees out of phase?

With a bridge rectifier they're only 60 degrees out of phase.  Add a capacitor and the phase doesn't matter a lot and you just add the voltage - minus the capacitor droop since they're not in phase.  When a coil is in the part of the phase where the voltage output is under about 1.2v the diodes bypass the current from the others around it.

But at the frequencies and currents we're running you need ENORMOUS capacitors - or the droop is so large they might as well not be there.

Series means you have 12, rather than 2, diodes in series dropping about an extra 2.4 volts.  And the diodes are conducting all the time, resulting in about three times the heating on the diode and requiring bigger diodes and heatsinks.  (The two effects are related.)

The proposal to half-wave rectify the system doesn't work well at all.  If you don't put in bypass diodes or capacitors you get nothing, period.  If you do put them in, you're still getting less than half the power from the coils compared to full-wave even in the best of cases - and still have extra diode drop.

You REALLY don't want to put separately-recitfied phases in series unless there's no other way to get power from them.

[deloiter]

Thanks for the explanation!

[RE Pod]

Thanks for all the explanations.  Informative as always.

I bought a proper 3 phase rectifier (160 Amp).  So hopefully that's that sorted.

Regards Terry...

[Ungrounded Lightning Rod]

Series means you have 12, rather than 2, diodes in series dropping about an extra 2.4 volts.

Oops:  "6, rather than 2"

The 2.4 volts of extra drop is right, though.

[drdongle]

I agree that it depends on application, but generally with a wind turbine your trying to squeeze out the maximum net power (watts). Losing half the AC cycle is a big loss in my book, total power would be half of that of using both halves of the cycle( not factoring in rectifiers). In some applications such as trickle charging this may be exceptable, not when it comes to brute force power where you want everything you can get.

[finnsawyer]

But if you can get P watts out at a windspeed of X, and I can get P/2 at a wind speed of X - Y due to the higher voltage might that not be an improvement in some cases.  With wind you're really limited by the power in the wind.  Once you reach a usable voltage the velocity cubed effect tends to make most other considerations moot.  Nor is it clear that my scheme would necessarily get half the power of yours.  If the only difference is 12 diodes in your case versus 3 in mine, we can say a few things about the waveforms.  To get the same current (or charge per second) out, the voltage drop between current pulses in my case is twice yours, but  hopefully still much less less than 1.8 volts.  Now to get this result it is necessary to ram twice the charge through my diodes than yours.   Since we have a greater voltage drop across the capacitors, the diodes will start conducting a little earlier in the cycle.  So we get a broader and higher amplitude current pulse than in your case.  Now the winding resistance comes into play.  A voltage loss due to the current pulse will occur, not quite twice as much in my case as yours.  If this drop plus the capacitor drop plus the slight increase in diode voltage due to the greater current pulse is less than 1.8 volts my scheme actually will provide more power than yours.  Such is the advantage of larger voltage.  As output current increases, the voltage drops occurring in the windings will eventually give your scheme the advantage.

[(unknown)]

it may be of interest for all of you that there are diodes that may be better than the common silicone rectifier diodes. These are Shottky-diodes, which are available as double diode in one 3 pin package for 2 x 60 Ampere. Reverse Voltage is up to about 60 Volt. And the big difference is voltage drop: 0.1 to 0.3 Volt depending on current. These diodes come in a flat plastic housing and look like a transistor of the bigger kind, with the pins on one side, surface mountable to a cooling plate, non-isolated.

One diode costs around 3 Euro (around 4 USD) at Conrad Electronic in Germany.

Happy soldering!

WiLi

Germany