Simplified star - delta switch

[wphfl]

A few months back, Windstuff Ed ran an experiment where he rectified the voltage from the star point of a two-phase generator.  Zubbly identified this configuration as a "Scott T" connection and Ed called it "Pseudo-delta".  Unfortunately the thread drifted off in to the relative merits of two-phase generation, and the significance of the experiment was lost.  After analyzing the current flows, I realized Ed had shown us a simple and elegant way to provide a star - delta switch from a single set of stator windings.

In a star configuration, current flows through all three-phases when a voltage is induced in any one phase - full current in the induced phase and half current in each of the other two adjacent phases.  This is why the resistance in a star wired stator is 2x the resistance of any single phase.

In a delta configuration, the three phases are in parallel so the current will flow through the two end points of the phase and the phase resistance is equal to the resistance of any single phase.  Since the phase resistance is half that of star, the current flow is theoretically twice that of star.  (Not entirely true in practice, I am ignoring the effects of parasitic current flow in the adjacent phases for the sake of simplicity.)  Delta provides more current, and the induced voltage is virtually identical in each.

Now let's wire our stator in star, but bring a forth wire down the tower from the star point to our rectifiers.  In this line, insert either a mechanical or electrical switch between the star point and the rectifier.  When this switch is open, the generator performs as a star wired stator.  However when the switch is closed, the three phases are now paralleled and the generator performs as a delta wired stator.  This may be a simple way to extend the range of the system in varying wind conditions using a single set of windings.  There are of course implications for stator losses, heating, and prop performance, which I shall leave to others to explain.

Regards,

Kevin

[dinges]

Been there, done that.

It's a good idea. You've got 6 wires coming out of your stator, and with a manual switch (or relay) you can switch from star to delta (delta is not really parallel, I describe it as 'a bit in parallel', by a factor 1.73 (sqrt(3))

It took me an hour of thinking and making quick sketches, but I've got a schematic somewhere on how to wire your switch. It's in AutoCAD format though. If anyone's interested, I could make a quick sketch of it and scan it in. Though I'm pretty sure I haven't invented the wheel, and others have done this before. The wiring of the switch seems to be the hardest of it all (but not very difficult, just follow the drawing).

You would need a 3-pole switch (3pole, double throw) that's a bit uncommon. If you use a relay, there are plenty of 3-pole relays out there (though they will have to be able to handle the current...)

I'm planning on building an automatic star-delta switchover in the future, but so far I haven't seen any good designs yet (and I don't intend to buy completed switches or project kits. If I can't build it myself, from scratch, I don't bother. Because I have to be able to service it myself too, if necessary. PICs would be ok, as long as the software is freely available.)

Peter,

The Netherlands.

[willib]

Thank you admins for your prompt removal of my previous garbage calculations ..

this one is correct.

 wphfl  your statement "In a delta configuration, the three phases are in parallel so the current will flow through the two end points of the phase and the phase resistance is equal to the resistance of any single phase."

is incorrect because

http://www.otherpower.com/images/scimages/2965/delta.GIF

the delta resistance between any two points is the resistance of one phase in parallel with the other two..

the formula for parallel resistance  is R1*R2/(R+R2)

R1= the resistance of one phase R.

and R2 = (2*R)

so plugging into the formula gives us R*(2*R)/R+(2*R)

this equals 2R^2/3R and simplifys to 2R/3.

So the delta resistance between any two points is two thirds the resistance of one phase..

so sorry ,  for any confusion.

[scottsAI]

Hello wphfl,

Not sure I understand what your neutral is connected to?

Assumption: Star is connected to 3 phase bridge rectifier.

Neutral connected at the rectifier; can't connect to + or -, will cause a short. (tried it)

So, added another set of rectifiers like the ones on the other phases. Result was no current flow, in the neutral wire. Voltage is always at some mid point voltage relative to the other phase wires.

Simulated 3 phase generator with 17v out,  3 ohm output impedance to limit current.

Connected to a 13v battery, with 0.001 ohms.

The star connection simulated properly.

Have fun,

Scott.

[windstuffnow]

  Kevin,

    I built a 3 phase unit that uses the 4th wire as an experiment.   I wasn't sure if it would work the way I thought it would.   It seems to increase the capability of the alternator without creating more heat as the delta circuit.   Basically what I did was to connect a lead to the neutral (star connector) and bring it down the tower with the other 3 and connect it to the spare tab on the rectifier.  I assumed current wouldn't flow through this lead until the voltage of each phase was higher than the battery charging voltage.  

    The testing that I had done with the small 2 phase and 3 phase alternator showed a much higher output using the neutral.   Like you said sort of an automatic star-delta switch but without the switch or delta connection.   The Lenz turbine is using this alternator right now which is a fairly inefficient alternator that was actually designed for a smaller turbine.  I expected 200-300 watts max out of it but I've seen as much as 450 watts from it with no signs of overheating.   So the "Psudo delta" or "Scott T" does actually improve the performance of the unit.   I spent one real windy day where the turbine was really working the alternator hard up and down from the roof checking the temp of the stator.   It never went over 100* F.   I can't say I completely understand what is going on with it wired this way but it seems to work quite well.

.

[DaveW]

Thanks all, this is a idea worth following up on, what with my concerns with stator heating. But I guess this means I have to replace my nifty 3 phase recifier.

[Flux]

Ed

This should work fine for the 2 phase case, but it is in effect changing from single phase to 2 phase in the rectifier.

It shouldn't work for the 3 phase case, it should be as Scotts has pointed out above, with a balanced 3 phase the neutral should sit mid way between the dc rails and no current should flow in the neutral.

Are you using some non standard rectifier arrangement like Jerry ? I am not sure why you are seeing any change with a 6 diode 3 phase bridge.

Flux

[ghurd]

Can someone direct me to some more information on the Scott T connections for a 2-Ph alternator.

Or is it just connecting both phases to their own bridge, and joining one AC connection of one bridge to one AC connection of the other bridge?

G-

[Flux]

I wouldn't get too excited about this, The Scott connection is a way of converting from 2 phase to 3 phase or vice versa.

One winding of the 2 phase is centre tapped and the other phase has one end joined to this centre tap. To work properly the windings have to have a ratio of 1: root3/2

This will give a balanced 3 phase. It was commonly used as a bodge to run a 3 phase motor on 2 phase and in that case none of the ratios work properly and it is not balanced but it worked well enough.

Two phase can be rectified with 2 single phase bridge rectifiers or one end of the 2 phases can be commoned and this and the other 2 ends fed to a 3 phase bridge, it saves 2 diodes and seems to have the same effect. If the commoned point is removed from the rectifier, the thing operates as single phase with the 2 phases 90 deg apart as compared with 3 phase with one lead missing where the single phase vector angle is 120 deg.

If the winding was wound for the true Scott ratios then it should work with the rectifier as true 3 phase ignoring the common ( star) point.

Seems to me that you use 2 phase warts and all or make it true 3 phase.

Flux

[windstuffnow]

  Hi Flux,

    I'm using the standard rectifier blocks but using all 4 AC inputs.  I ran a few experiments with the 2 phase and 3 phase a while back.   I ran the alternators at different speeds using both configurations and when the neutral wire was connected the current output to the batteries increased considerably at higher speeds as opposed to not connecting it.   At higher rpm you can even hear a change in tone of the vibration created from the alternators ( easy to hear with the roof mounted Lenz turbine in high winds).   In my mind, I see it changing when the rpm reaches a point where each phase is producing a voltage higher than the battery voltage and starts passing current through the neutral wire.   Basically turning it into separate single phases.   I really don't know but it seems to work better using the neutral than with out it, it certainly keeps the heat down.

  The alternator using the 4 lead is a low power, low rpm built for a smaller version of the Lenz turbine.   I used #18 wire with 47 turns per coil and the resistance came in at 1.2 ohms.   I figure it should reach its 50% point at around 12 amps.  Beyond that it should start producing lots of heat as the efficiency drops.   It didn't seem to heat up much at all while pushing over 30 amps into the battery.  It is, however, open to the air so I'm sure it was disipating a considerable amount of heat in the air stream but I really expected to see something burnt.  It's gone through several extended high wind situations producing, what I thought to be, more than it could handle.  Interesting to say the least!

.

[Flux]

Thanks Ed.  It certainly doesn't operate that way under normal conditions where the internal drop is reasonable. I am beginning to suspect that if the internal drop is high enough commutation may transfer to the neutral rather than the normal line to line.

I will try to see if I can get this to happen but the only alternator available to test at the moment is very stiff and I shall have trouble loading it down to the point where this might happen.

Interesting stuff as always.   Flux

[hiker]

what about a simple wire diagram...:]

[Flux]

See if this helps.

4 wire needs 2 single phase bridges, the other 2 can use a 3 phase bridge.  Ignore the star point on the Scott  

[ghurd]

Now I am going to have to dig through the stepper motors... again!

I actually did try the 2-Ph 3 wire a while ago on several styles of steppers and the output went down considerably, maybe because of reactance and hi ohm coils.

[Flux]

Ed

 I did try this, couldn't get any neutral current with the alternator on its own but I did get a small current when I added 1 ohm resistors in the lines to drop the efficiency down to about 30%. Still only an amp or so and my resistors were not exactly matched.

You must be pushing that alternator pretty hard to have a significant neutral current.

Flux