Delta Connection improves performance

[SparWeb]

I had plenty of time to make many changes an improvements to my genny during the holidays.  Results are slowly coming in, because it's hardly been windy at all in the past 2 weeks...

Apart from putting up some special christmas lights:

http://www.fieldlines.com/story/2007/12/26/14242/395

:^)

Most importantly, I changed the motor conversion connections from Star to Delta - big and immediate improvement!

Connecting the genny in Star was a deliberate "limiting" of output because at the time I only had a pair of Trojan 105's to charge.  I regularly came home to find them bubbling away past 16V.

Now I have a big set of AGM's to charge, so the sky's the limit when it comes to charging these!  Changing to Delta has matched up the cut-in speed of the blades with the generator's open-circuit voltage much better.  There's almost a constant trickle of 2-4 Amps, as long as there's enough breeze to feel on your face at ground level.

I also piled a pound of weight onto the tail.  This should put an end to the pitiful furling in 15 mph wind!  Having changed from Star to Delta connection will have an indirect effect of prop thrust, which will also affect furling speed.  All my reading of tests Flux made on this topic and I still don't have a "feel" for which way this effect will turn.

I put in a C40 charge controller, with 400 watts of dump load (four 2-ohm resistors in parallel).

Nothing stands in my way, now, when I get a 24V inverter and re-connect the batteries accordingly.

I'm considering a Magnum 4024 Inverter.  Any experience to share with these (relatively) new products?  It seems to beat the Outback's in cost-per-watt, and when used as a charger, can deliver twice the DC.

Title shortened on request. TW

[SparWeb]

Wind suddenly came up an hour ago, and 12 Amps were flowing in a modest wind.  It got the C40 blinking 3 flashes at a time - not bad!  The book says 0.50 Volts below the Bulk voltage setting.  Accounting for temperature (-2C) the measured voltage is bang-on, so all seems well...

...except that the wind just died again.

Now the ammeter is twitching over 3-4 Amps.  Never seen it do that before.

Inverter is turned off so it's not that.  Going to pull the front panel of the control box and see if there's a weak phase or blown rectifier...

Stay tuned folks.  Don't touch that dial!

[SparWeb]

Rats!  That is what it was.  The moment I opened the control box, my nose told me all I needed to know.  The heat sink was actually warm (and it's below freezing tonight).  First time for that, too.  Tested each phase, and one didn't match by a long stretch.  Pulled the wire off the offending side of the bridge and presto - back up to 10 Amps.  The rectifier from ~ to - is shorted.

Well, you can't run safely with one phase on only half the bridge that's already burned out, so everything's all shut down, now.

After all that work to raise the output, and I neglected the rectifiers that put it all together!

[sorry about the long subject title, seems to cause a software error.  Can it be shortened?]

[ghurd]

Regular Delta or Jerry Rigged?

G-

[SparWeb]

The rectifiers are Jerry rigged, I think.  Always been a bit confused as to what "Jerry-rigged" means.  Three wires coming down the pole.  The generator is connected in Delta.  Three AC wires split across three bridge rectifier blocks for DC output.

With three standard block bridge rectifiers, how else do you connect them?  My diagram shows for example, phase A being split to one side of the lower bridge, and the other side of the upper bridge.  Since then, I put both sides of Phase A on its own bridge; same with B and C on their own bridges.  Don't think this mattered in the failure mode.

It's more likely that my "test" circuit, run through the transformer, is what doomed me.  It adds an inductive load across two phases.  It, too, didn't matter back when I couldn't collect more than 10 Amps!

[TomW]

Steven;

"Jerry Rigged" is simply a way of saying "individually rectified phases". At least thats my understanding of it.

Cheers.

TomW

[ghurd]

You have done some interesting things with the bridges, but it ain't Jerry Rigged.

That's regular Delta.

Jerry Rig needs all 6 wires.

Each phase goes to its own bridge.  Then each bridge goes to the battery, one at a time.

Just like a single phase PMA, but 3 of them in the same holder.

If that isn't vodka clear, I'll do one of my world famous "MS Paint" diagrams.

I probably should have done that a long time ago.

Delta has parasitic currents flowing.  That is bad.

Jerry Rigged does not.

I do not know how it would relate to a reasonable dual rotor because everyone argued to the point of confusion.

I never made a real nice conversion, but all of mine worked better in Jerry Rig than Delta.  A lot better.

I don't see why the 1 ohm resistors are in there?  

Conversion shouldn't need it.  8' dual rotor that needed it in Star shouldn't need it in Delta or Jerry Rigged.

A bit of looking and now I don't know if this is a Baldor conversion or an 8' dual rotor.

BTW, 4W for the Christmas lights was watts well spent!

G-

[ghurd]

Dog Sitting...

This is Jerry Rigged...

[wooferhound]

Here is my version

I think this is also Jerry Rigged, but in series for more voltage

[Flux]

I seriously doubt that you would need any form of series resistor with a motor conversion that was even vaguely matched.

I don't think your transformer can be blamed for rectifier failure as it is before the rectifier. In delta the frequency will be higher and it would load the winding less than in star. Look at heatsinks or more likely poor rectifiers from doubtful sources.

Flux

[SparWeb]

Some would call is luck, others "providence".  Last night winds over 90 km/hr (45 knots) blew through overnight.  If my windmill had not been shut down, due to this earlier failure, I don't know if it would have survived!  Rectifier bridges failing one after the other could have let to a total run-away.

Even with the kill switch shorting it out, there was enough torque for the rotor to spin maybe 200-300 RPM and fully furl during the strongest of the wind.

Guy wires humming like guitar strings...

...

Six wires down the pole are required for Jerry-rig.  Somewhere in the back of my head I'd learned that, then forgotten.  I can only fit three down my tower (long sad story).

Are "circulating" currents of concern in a motor conversion?  The windings are evenly spaced, down to the thousandth of an inch, and the magnets were placed on the rotor with CNC machined holes, so there's little variability, there, either.

Parallel-Star is a reasonable option - there are plenty of wires coming out of this motor housing!

I have picked up new rectifier blocks, NTE brand.  Fingers crossed.  Maybe it's time for one of those bigger 3-phase blocks.

Transformer will soon be eliminated, anyway.  While picking up the bridges, I bought the pieces required for CompDoc's tachometer circuit.

[ghurd]

Stray thoughts.

Fully furled while shorted?  That means get rid of those resistors!

"Circulating" currents.  Someone far smarter than me says we are "~making power with something intended to use power, and making power is a side effect~".

I had better (bench) results Jerry Rigged than Delta with a GE ECM.  Factory made, same idea.  And Jerry did too.  Depends on the voltages.

So yes, I think it is an issue.

Bridges can be at the top of the tower.

Parallel-Star.  Looks to me like it splits the difference between Delta, Jerry and Star.

I never had the option.

But I still stand on Jerry's side of the fence, unless the Star output meets the voltage:RPM needed.

G-

[SparWeb]

Stray responses:

"Fully furled while shorted?  That means get rid of those resistors!"

Actually, that diagram is from October - those resistors are long gone.  Sorry about that.  Phase resistance is 1.2 Ohms now in Delta, and yes it was turning in a >90kph wind.  I wouldn't expect otherwise with an 8-ft prop.

Thanks Ghurd.

(What's a GEECM?)

[TomW]

General Electric Electronically Controlled Motor?

"GE ECM"