Author Topic: Delta, star, Jerry riged? (Read 18480 times)

Jerry · « **on:** February 11, 2006, 01:50:09 AM »

As promised I'm prersenting my thoughts on the above. I tried to find a computer program for drawing schimatics.

A freind gave me Electronics work bench but the computer and I are not freinds so I quite trying to use that program. It would be a few years before I figured it out so I've hand drawn some schimatics and then took pictures of those with my digital camera.

Sorry guys thats the best I can do for now. Please bere with me?

My insperation for posting the story was a result of my diary on my 8" dual rotor.

The debate was reborn and I thought I should spend some time explaining my thoughts.

The first thing I did this morning was to confirm my suspisions. The test subject was a GM car alternator refited with NEO magnets (Hornet). Some time ago I had seperated all 3 phases in this alt. There are 6 wires comming out of the alternator.

I chuked it up in the lathe and did a no load AC voltage test of each phase. Phase 1,2 and 3 are completely seperated from each other and have no common conection.

At 500 rpm each phase was producing 14.59 volts ac. I tested 2 phases wired in sires

This is 2/3 of star. Now a voltage reading of 25.11 v is showing on the meter.

This is going to be a long diary therefor I'll brake it up into several postings.

Befor I brake for the next post here are a few pix of the alt and the meter readings.

The next post will have more discusion and the schimatic photos.

JK TAS Jerry

Jerry · « **Reply #1 on:** February 10, 2006, 07:07:10 PM »

As seen in the last photo the star reading is 25.11 volts. Now normaly you would think that two 14.59 volt coils wired in sires would produce 29.18 volts. They would if we were talking singel phase but these coils are not totaly in phase. As fack each coil/phase has a 120 degree relationship with its neighboring coil/phase. Neither one is at peak voltage when its neighbor is.

Heres a couple pix to help visulize this.

If the coils were in phase we would see 29.18 volts.

Here is a normal star conection.

The circle here is around one fullwave bridge rectifier. This is why 2 bridges are used. Reality is 1.5 bridges for normal star or delta.

Now on to the next post. I hope the brakes don't confuse things.

JK TAS Jerry

Jerry · « **Reply #2 on:** February 10, 2006, 07:50:23 PM »

Although the are 3 phase it is imposable for the voltage of all 3 phases to be at there peak potential at the same time.

Dan B has stated he like the fact that all coils have current flowing through at all times. This is true.

An example of this is phase 1 is at zero potential at some point in its output do to rotor rotation. At this same point in time phase 2 and 3 are at 120 degrees either aproaching there peak output or leaving there peak output. This means voltage has to pass through phase 1 and at that instant phase 1 is just a coil in the way for current to pass on its way to the doides its connected to.

As a mater of fact the only way any of the ac voltage can reach a fullwave bridge is

to pass its current through a coil that is partialy an inductor only, or has current that is slightly in opsistion.

Puting littel inductors in sires with each phase is not a benifit. For those how understand phaseing try this wire your speaker out of phase in your audio system to see what happens to the acoutic energy.

If we could stop the alt intantly a measure what was happening at that instant we could acctuly see a DC voltage. At one end of a coil a positive voltage would be present and at the end of that coil there would be the negative. This alternates with the alternating magnetic flux but lets do that freez frame thing and check the relationship between phases at that nano second.

We would see that while one phase is producing it peak volktage the phase next to it would be produceing less. They can't all be on top of the hill at the same time. They are a manditory 120 degrees apart.

Its like wireing two 12 volt batteries in sires. Every time you tried to connect the 2 batteries one would instantly drop to 10 volts. You could never get the to be 12 volts at the same time. So you get 22 volts not 24 volts.

My proposal is to make each phase have 2 times the voltage as one phase. This solves the lost voltage isue. Then provide each phase with its own fullwave bridge rectifier. This keeps the voltage high (no phase to phase lose) The perelelling of the lower amperage higher voltage will provide the same amperage as the lower voltage higher amperage phases wire in sires.

A single phase example. A power transnformer will help to understand this aspect.

Lets say I have 2 transformers. They both have 120 primaries.

One has 2 sendaries at 10 volts 5 amps. If I wire these secondaries in perelell I get 10 volts 10 amps.

The other transformer has 2 secondaries at 5 volts 10 amps. If I wire these secondairies in sires I get 10 volts at 10 amps. This example is equal to star with one big exception I can never get 5 volts at the same time from the 3 phases or coil groups.

By wireing a seperate bridge to each phase a phase will never have to send its power through another phase that is not in total agreement.

Time for a couple more pix. Proposed coil change ans stock coil preformance.

Time to brake.

JK TAS Jerry

Jerry · « **Reply #3 on:** February 10, 2006, 08:10:04 PM »

Just read the above post. Sorry about all the miss spelling and left out letters. I think its a speacial CODE. I've seen others hear on the board use it but I think you regulars can desifer it.

Hiker asked if it was still 3 phase with a fullwave bridge per coil. Actully a my fellings and thoughts about this is not nessesaraly a bridge per coil. A bridge per phase is fine. I did it because I had enough small bridges and as Dan B I think pointed out there are slight and small diferances between coils.

If 1 coil is making 10 volts and 1 is making 9.7 volts and 1 is making 10.2 volts why make the suffer like this. The high volt coils are held back and draged down by the low volt coils. Let the all participate freely without hastel.

With this home brew stuff it would be rear to see all coils at the presice same voltage.

However I don't think the diferance is enough to warrent a manditory bridge for each coil. I just did it cause I could.

A couple more pix.

The last pix is the diode bridge per phase.

Now time for more discusion.

OK guys pick me apart.

JK TAS Jerry

sahlein · « **Reply #4 on:** February 10, 2006, 09:48:44 PM »

In my dim mind it seems that what you have done is to make three single-phase circuits in the space formerly occupied by a three phase machine.

The only "Phasing" remaining is just from the fact that the magnets and coils are spaced such that the outputs of each "leg" of your machine are at varying values relative to each other at any given instant.

Joe S.

hiker · « **Reply #5 on:** February 10, 2006, 10:53:20 PM »

nice draw up their JERRY.....

so you can use one rect for each phase and get the same results--as using a rect. for each coil..? seems like you would have a slight voltage gain by using one rect. for each phase..sence there is some loss using a rect..so the more rects you use = more loss in voltage--and i guess you could add a few more windings to couter act this..

damm guess its time to scroung up some wire and give it a try.............

scottsAI · « **Reply #6 on:** February 11, 2006, 01:35:42 AM »

Hello Jerry,

Some more great testing.

I thought you were going to show the bridge per phase produced more power than star or delta.

But you changed the subject before anything was demonstrated??

What were the Battery currents of the three methods? The battery is what it's all about?

How can we decide what is better unless we know what was?

"to pass its current through a coil that is partialy an inductor only, or has current that is slightly in opsistion."

The generator is not an inductor. The impedance of the generator is very low, it does not behave like an inductor. A generator has some small inductance like any wire will have, major effect is not an inductor it's a voltage/current source.

"My proposal is to make each phase have 2 times the voltage as one phase. This solves the lost voltage isue. Then provide each phase with its own fullwave bridge rectifier. This keeps the voltage high (no phase to phase lose) The perelelling of the lower amperage higher voltage will provide the same amperage as the lower voltage higher amperage phases wire in sires."

Boosting the voltage can get you a low cut in speed. When the wind is in your power range, the generators efficiency is going to suffer. Lets assume the blade is large, can produce the power listed.

We have a 12v system, generators impedance is 1 ohm to keep the math simple. The batteries impedance is 0.0001 ohm, ignore it more than two order mag smaller.

Cut in 6mph we have 13 volts generated.

The current into the battery is 1 amp, or 12watts, power as heat in gen is 1 watt.

mph = 26v, (26 - 12)1 = 14a, 168wa into battery, 14x14 = 196w as heat in gen.
mph = 52v, (52 - 12)1 = 40a, 488w into bat, 40x40 = 1600w left as heat in gen.

Good idea to make a chart with wind vs calculated power into battery. Results are quite surprising.

The danger of boosting the voltage too high is generator burn out. The higher speed winds helps cool it and maybe the blade may be stalled by the excessive power the generator is trying to produce, which will make the machine noisy.

The use of the star to delta switch is trying to get the best of both worlds. Star or the higher voltage for low cut in speed, delta for the lower voltage with higher current at the higher speeds, switching configuration has it own problems.

I would love to see the three configurations tested?? (needs to be same coil)

Have fun,

Scott.

Flux · « **Reply #7 on:** February 11, 2006, 03:01:38 AM »

Jerry

I like your drawings, modified Dancad. I may try it, much quicker than drawing things by computer.

It is inevitable that the 3 voltages of a 3 phase system add vectorially.There is no way round it.Rectifying the things separately doesn't get round this. There is no way to have all coils contributing current at the same time. Any 3 phase rectifier will have the coil with highest voltage conducting at any instant and neglecting overlap( doesn't happen here), the current is always carried by a pair of diodes, and with a full bridge, one diode swops to the next every 60 electrical degrees. The same conduction path will exist with your system but the diode conduction sequence will be more complex.

If you use 6 phase rather than 3, you can achieve the full twice phase voltage and the cut in speed is reduced. The ripple is also reduced but the operating efficiency is not, in fact it it is slightly worse. The winding has to supply a higher peak current for a shorter time to maintain the same average current.

If you look at a force commutated rectifier you may be able to improve things, but otherwise no.

Flux

oztules · « **Reply #8 on:** February 11, 2006, 03:11:35 AM »

sorry jerry,

but before you head north for the patent office, connect the three coils in "The last pix is the diode bridge per phase." to delta. the voltages shown will remain identical........jerryrig equals= delta rig (without bothering to connect up the coils).... ie join the bottom of coil 1 to the bottom of coil 2. connect the top of coil 2 to the top of coil 3. join the bottom of coil3 to the top of coil 1. what do you have ....delta (coil 2 is inverted....ask ed (windstuffnow for the explanation).

All the voltages remain the same as in the "jerry rig" configuration.. if you hook them up in star, the ac voltage will become 36.98v. ring a bell 14.59v delta and 25.11v star....... now you simply have 29.18v delta ( or jerryrig) and 36.98 star.

in delta the minor changes in voltage dont push pull either, as it is the rectified voltage that is joined together, the diodes prevent looking back through the coils.

In both jerryrig and delta (effectively the same thing in this application really) the dc will pulse out of time with each other. ie for a 2 pole machine, there will three +dc pulses per cycle appearing at the output. they don't add together because they arrive at the output at different times and are not in series. Averaged over time you have three times the output of a single coil.

so all the advantages and disadvantages attributable to the star/ delta argument, are identical for the star/jerryrig arrangement.....hope this makes some kinda sense.....oztules

hvirtane · « **Reply #9 on:** February 11, 2006, 03:24:37 AM »

I think that the whole issue is,

if it is better to connect coils

of different phases before or

after the rectifiers.

If you'll connect them after

the rectifiers

(the way of Jerry and Hugh),

the currents are

not fighting against

each other at any time, as

it is the case with the star

connection or even with

a delta connection.

So I'm thinking that

connecting them after

the rectifiers is better

as regards the maximum power.

The other thing is that

in practice you often need

high voltages at low RPM,

not the best power in Watts.

So in practice star is the choice

in many cases.

(Even if it is giving less

power at high speeds.)

The other problem is that

you would need more rectifiers

and put them on the top

of the tower, if you'll

do it the way of Jerry and Hugh.

- Hannu

Gary D · « **Reply #10 on:** February 11, 2006, 08:01:16 AM »

In Jerry's other test (wind truck), he got 1425 watts at 35 volts.... He only had 3 rectifiers rated for 6 amps each per phase, so they would have fried (or been at max. output) at 630 watts. There HAS to be some adding of amps from the other coils since this was close to double their ratings??? This was why I thought going to 10 coils such as hughs 5 phase machine would give a bit more load sharing (coils closer to peak,, not 120 degrees out). My inept way of looking at things. Hope not to offend either viewpoint, just think Jerry is on to something! Wonder if increasing the lathe test rpm a bit to get beyond the amp rating of one rectifier might clear things up? If a 6 amp rectifier should blow at x speed single phase, exceed that speed by a bit. Hook 3- 6 amp rectifires( one per phase) and see what the output is, smoke or more than the 6 amps therorectically possible.... I tend to think increased amps.... Simple quick belt change might clear things up perhaps? Gary D.

DanB · « **Reply #11 on:** February 11, 2006, 09:02:10 AM »

"If you'll connect them after

the rectifiers

(the way of Jerry and Hugh), ...."

Actually Hugh is using a conventional 'star' connection, but two coils are running in parallel, so there are 10 leads coming out - but the output of any two leads is not through one coil, but through two coils which are out of phase through a start connection, just like any multiphase machine wired in Star.

johnlm · « **Reply #12 on:** February 11, 2006, 09:03:42 AM »

I tend to agree with Oztules. Effectively Jerry ended up with a three phase machine essentially connected in a Delta type configuration, except he added the phases (in Delta case the lines) after the rectifiers instead of before the rectifiers. Because he used a 12 magnet/9 coil arrangement he indeed has a polyphase (3 ph) machine so in fact at any given point in time there were coils that were not at peak voltage while others were. He increased the number of turns on each coil (as also stated in his discussion above) to offset the lower voltage one gets out of Delta. I think (correct me if Im wrong) Jerry is assuming that all the coils are at peak voltage all the time in his configuration which they cannot be because there are points in time that there is no magnet directly over a given coil. The ones that are not at peak voltage are not contributing full current if any to the summation node after the rectifiers. So the bottom line comes down to what I think Hannu stated and that is: Is it more effecient to add the currents together before the rectifiers or after. In the standard Delta case you have fewer diodes active but running at a higher current. In the Jerry-rig case you have more diodes active at any point in time but running at a lower current so it is probably about a wash on power lost in the diodes.

Also to clarify a point that Gary made, regarding the wind truck test with 1350 W (or whatever it was) and 35 amps overloading 3 bridge rectifiers. I think Jerry has 9 bridge rectifiers on his unit he tested on the truck, with the currents not equally divided up between them at any one point in time, but likely any given rectifier pack was running under 6 A.

John

Jerry · « **Reply #13 on:** February 11, 2006, 09:18:12 AM »

Yes the circuts i've show above are for the normal 3 phase conections not my diode bridge per coil.

But even with a doide bridge per coil its still a 3 phase alt. There are still 9 coils and 12 magnets so the electrical and physical arangment is still there.

Many newbys have asked questions like...do I conect the start to the end or start to start or end to end or lay my coils this way or that way. All of that has no importance with a bridge per coil it just doesn't matter. Each coil and bridge is its own littel ac to dc power supply with all its dc out put contributing to the alts total output.

The sum of all there power is collect to 1 point. The amount of copper is the same, the amount of magnet is the same. The diferance is how the power is collected from a group of 9 coils and 24 magnets (12 per rotor) Rectify each phase seperatly and the phase diferances are done away with.

Hough has discused this with delta. I think he refures to it as hystrisus isuse or some thing like that. If there are phase isues with delta then there are phase isues with star.

JK TAS Jerry

Jerry · « **Reply #14 on:** February 11, 2006, 09:23:48 AM »

Hi hiker.

There would be more diode loss if all the rectifiers were wired in sires then the losses would add up but they are wired in perelell. So the total rectifier loss for all 9 is equal to one rectifier loss only.

JK TAS Jerry

DanB · « **Reply #15 on:** February 11, 2006, 09:41:10 AM »

"Hough has discused this with delta. I think he refures to it as hystrisus isuse or some thing like that. If there are phase isues with delta then there are phase isues with star."

Not really. Delta is basicly a sort of 'parallel' connection, so if we have coils with slight variations in them (or coils that are not placed perfectly) then currents can flow between the coils and create inefficiency (two coils basicly in phase, but one is producing slightly more than the other, so the lower output coil is a small load on the one with slightly higher output). If the alternator was perfectly built - Delta would be fine, but perfection is about impossible so we always avoid parallel, or Delta connections. Star is a series connection and this is not an issue. It has nothing to do with hystoresis.

Jerry · « **Reply #16 on:** February 11, 2006, 09:58:52 AM »

Hi Scott.

I didn't change the subject. The subject was explore the diferances between delta, star and jerryriged and my thoughts on the idea.

As I see it there is a point when any given coil or phase is not producing power. It not at full positive swing all the time Its not at full negitive swing all the time.

In its transition between negative and positive its in a momantary limbo or non productive state and at that point it is just a coil of copper wire. However at that point it is still wired in sires with another coil that is producing power and the power producing coil must send its power through this none produceing coil.

This happens in the 3 phase sequence. This is the 3 phase relationship.

The reason for booting the voltage was not to make the voltage higher then the star voltage.

I tried to show this comparison with the transformer idea. Its also nown that the delta arangement produces half the voltage as star but twice the current as star and this is with the same coils.

You could make a delta wire alt produce the same voltage and the same current by using smaller wire with twice as many turns as star the end result will be the same.

Thats all I've done here. What I've done here is the equivelent to delta without the hystreesess problems (phase canceling problems) that Hugh refures to on ocasions.

Delta is shuned because of this problem. My aproach simply alows you to do the delta scheam without the phase conflict isues.

The actul out is the same as star but without the same phase isues.

You may have missed the transformer idea above but the 2 transformers have diferant secondaries but one is configured sires one is configured perelell the end result is the same output both in amps and volts.

Thats all I've done here. I have not made a higher voltage alt. The 2 are identical in terms of amps and volts more with one acception. One has done away with phase isues. Or the lack of compatability between coils wired in sires that are not in phase with each other.

In terms of temperature. This littel alt was stone cold after returning from a 1225 watt test run. Diodes cold and coils cold.

As far as needs to be the same coil, no thats not true in this case. Please refure back to thew tranformer discusion.

If you use the exact same coil for delta and star the end result will not be the axact same voltage and amperage. My bridge per coil is equal to a delta configuration there for it to have the same output as star it must have twice the turns count. This will mean half the wire size.

JK TAS Jerry

Jerry · « **Reply #17 on:** February 11, 2006, 10:20:34 AM »

Hi Flux.

I realize all coils are not contributing power at the same time. Thats my point. If you have a coil that is contributing power wired in sires with a coil that is not contributing power or not contributing power at its full potetial then the power coming from these 2 coils is not the full potential that is in the coils but not being harvested.

One worker is sweeping the floor while the other is just draging his feet and has not picked the broom up yet. This just happens to be ocureing in a group of 3 missfits.

As far as the 6 phase thing and trying to double voltage. Thats not what I'm after here. As far as complexity. In my 9 bridge system its actuly very simple as per electrical flow is concerned. Its like geting 9 littel 5 amp 14 volt power supplies and wireing them in perelell. The amperage and voltage output ofcorse depends on wind speed. They have no conflicts with each other. Same can be done with a single bridge per phase rather than per coil. The bigest conflict is between phases. A very small conflick between coils within a given phase.

Jk TAS Jerry

Gary D · « **Reply #18 on:** February 11, 2006, 10:36:43 AM »

John, you are correct if you are conceeding that more than one phase is contributing to the overall amps produced. That's what Jerry has been trying to say (I think).

There is a reason for my madness... If this is correct, going to a 16 magnet per rotor duel rotor assy. could "standardize" things a bit. With 12 coils (4 per phase), it would give anyone the option at any time to change output voltages with a set of jumper wires, no need to recast stators. You could go from 12 volts to either 24 or 48 that way (if my thinking isn't muddled). The resistance issue with line loss tower to batteries not calculated in of course. That is beyond my current knowledge. Then again, what isn't?... :-) Gary D.

Jerry · « **Reply #19 on:** February 11, 2006, 10:49:44 AM »

Hi oztules.

Yes my configureation is equal to delta. But without some delta isuses that Hugh has talked about.

My quest is a phase configuration without conflicts that are present between 3 coil groups wired is star that have phase diferances.

I realize the delta or jerryrig dc out puts will be out of time with each other.

There again thats what I'm talking about in star the coils from the 3 phases are out of time with each other. They are not in sequence or phase or on the same page or in agreement or producing the maximum power at the same time. That can't this is 3 phase were talking about here.

So one coil must send its power through the other coils that are not peak output.

You said "they don't add together" Tell me then how my 8" dual rotor alt with 9 littel coils of 22 gage wire with a 6 amp diode each coil was able to as a group produce 35 amps at 35 volt accross 2 Optima batteries wired at 24 volt sires.

If I do the math that means that each coil was producing 3.8888 amps each. You said "I would have 3 times the output of a single coil averaged over time". Well in this case the output for a single coil was 3.88888 amps. So lets see 3.8888 amps per coil X 3 = 11.6664 amps. Could you tell me then were the other 23.333336 amps came from ?

JK TAS Jerry

Jerry · « **Reply #20 on:** February 11, 2006, 11:02:42 AM »

Thank Hannu.

I don't now if Hugh will like to be thought of as in the same school of thought as I. Right now I'm the outlaw or black sheep when it comes to this contraversy.

However I'll keep plunking away at it till I finaly see the light or maybe the light will come on for the rest of the world?

As far as diodes on the tower. You don't have to have 9 bridges to deal with these phases diferances. Only 3 bridgres one per phase. That does meen 6 wires down the tower if you want bridges on the ground. I'm flying a genny with 4 wires down and thats not to bad a couple more is doable I think and the 6 can be a smaller gage then 3 wires down would requier.

I just did 9 bridges cause I had them and thouhgt 9 littel free bridges was cheaper then 3 I had to pay for. It also releaves the slight diferances between coils.

JK TAS Jerry

Flux · « **Reply #21 on:** February 11, 2006, 11:14:32 AM »

Jerry

I have some sympathy for you reasoning about the issue with star, but it is the most efficient. One coil is not really fighting the other, there is a reduction in voltage but it does not produce a loss. Any winding that is distributed has this phenomena. The coils in one of your phases also suffer from this as they slide in and out of the magnetic field. We would get more volts if you could wind them all at one point like a single turn, but having a distribution with a reduction in voltage doesn't cause losses.

There is a much better case for considering your version against delta. There are 2 basic problems with delta, firstly there is an issue of circulating currents due to slight differences in voltage due to errors of spacing etc. This is a practical issue and is only very significant if the construction is bad. It mainly causes drag below cut in and may hinder start up. Your connection will prevent that.

The more serious issue is that with other than a star winding the rectifier waveform causes harmonic currents to circulate within the winding and these also cause loss.

There is no loss from this cause below cut in but it comes into play as soon as the rectifiers conduct.

I tried your method several years ago to see if it would block this problem and it didn't. At the time the thing was so similar to delta that I decided not to pursue it as star doesn't have this problem. My results do show your method slightly better but at that time I didn't have means of keeping battery volts constant and I can't measure the small changes in torque to the accuracy required to be certain. I will have another look, but the effect is so small that I don't think it is a major issue.

The test was done on an air gap alternator and it is conceivably possible that an iron cored alternator with slots would be worse affected by the harmonic currents.

Any chance that you could complete the storey you started with. It would be good to have an exact comparison between delta and Jerry rig on that converted car alternator that you show in the picture. Do you have the means to measure torque, just comparing output without being able to measure power in would not help.

Flux

Jerry · « **Reply #22 on:** February 11, 2006, 11:16:45 AM »

Hi Gar D.

Acctuly there are 9 coils of 250 turns each of 22 ga. wire. There are allso 9 each 6 amp diode bridges. At 35 amps this means 3.8888 amps per bridge. The bridges a rated for 6 amps each continuas and a much higher surge each. 6 amps X 9 X 35 volts would have been 1890 watts. So at 1225 watts I was quite under the amp rateings of these littel bridges.

The 22 gage seemed to handle the 3.8888 amps per coil just fine also. I had a bridge per coil not per phase. The wind test went way behond the laithe test in terms of power.

JK TAS Jerry

Jerry · « **Reply #23 on:** February 11, 2006, 11:36:49 AM »

Hi John.

Yes I will correct you.

The point is missed again. I realize none of the coils in a 3 phase system are at peak power at the same time. Not even in my 9 coil 9 bridge arangement.

That is not my point and yet it is my point. My point has to do with the star arangment of coils being wired in sireies that are not in phase with each other.

I know even with my arangement take any 3 coils of one phase at a given time they may be at there peak voltage while the other 2 phase will not be. And that doesn't mater if the coils are wire delta, star, sires perelell , seperate or what have you.

The coils of one phase are at the same peak at the same time. And that doesn'nt mater if its 1 coil per phase, 3 coils per phase, 6 coils per phase. It just makes no diferance haw many coils are in that same or one phase they are all at the same peak at the same time.

The diferance is when we compaire the output of any phase to the output of any other phase at a given time.

Thats the isue I'm talking about. Wireing a coil or coil group of one phase in sires with a coil or coil group of the other phase its not in phase with.

JK TAS Jerry

Jerry · « **Reply #24 on:** February 11, 2006, 11:50:52 AM »

Thanks Flux.

I'm releaved that I'm not crazy after all. I don't have a tourque measureing ability but I can do the delta VS jerryriged compairson. In terms of keeping battery voltage constant.

I have in some of my tests placed a load accross a battery to keep its voltage constant. Resistive loads are a problem for wind alt testing as this prevides loading befor the 12 volt cut in. The problem with battery testing is voltage kliming to an uncumfertable level with long term testing. A varyable load on the battery to keep battery voltage constant will work but it also needs to adjust as voltage condition changes with charging time.

I do have a zero to 500 amp load tester however these things a think are designed for very short test times.

I'm still not done doing battel with the star thing.

Thanks again Flux.

JK TAS Jerry

windstuffnow · « **Reply #25 on:** February 11, 2006, 01:56:17 PM »

An interesting topic, definately generated a lot of thought. I've done a bunch of similar testing myself and in the course of learning have created some nice alternators as well as a sizeable pile of junk ( which gets recycled into other projects I might add ). I would tend to agree with flux and the others on the star connection but I believe the rectified phases will out perform the delta connection in efficiency ( and create less heat loss ). I have a tendency to run the AC as far as I can before converting it to DC so the 3 phase star wired seems to be the easiest ( saves a little on large wire runs ). I'm not fond of running DC for the whole run... it would be fun to have a wire box at the bottom of the tower with connections for each coil for testing, changing and comparing. I've never rectified all the coils, sounds like a fun project, it might be a good comparison on my 500 watt kit to run them side by side in both set ups. You should be ashamed of yourself trying to get me to add another project to my "to do" list... I'm on my third roll of toilet paper as it is...

As far as finding the input on your lathe set up, your half way there... you already have the alternator chucked in the lathe and you have to hold the alternator from spinning... simply put an arm on the alternator attatched to a spring scale. If you know the rpm its running as well as the length of the arm and the scale reading you have all the info you need to determine input, output you already have. If the arm is 1 ft long and your scale reads in pounds you have ft/lbs. torque x rpm / 5252 = hp x 746 = watts. You'll have a comparison from watts going in to watts comming out ( pun intended ).

I like posts like this, gets our creative thoughts flowing... Keep up the great work Jerry!!!

.

DanB · « **Reply #26 on:** February 11, 2006, 02:15:16 PM »

I'm pretty sure Star will always come up a bit more efficient. I expect a 'perfect' alternator wired in Delta would also be slightly more efficient.. but perfection is next to impossible in these machines. I've been screwing around this morning trying to make some drawings to show this - but it's been frustrating without graph paper - maybe Ill get around to it. But Im quite sure current is flowing more of the time through any given coil with the star connection that it would be if you were rectifying each one seperately. The moment a single phase coil drops below cutin voltage it's wasted space. In a 3 phase machine, even if the coil is below that it can still be contributing voltage, and it's still conducting current. Hard for me to get across w/o drawings - maybe Ill draw something.

There's the old - but important argument I think, that normally alternators are wired in Star and I believe there's a reason for that. If they could make car alternators smaller, and cheaper by rectifying each phase seperately they'd be doing that. Generally a 3 phase machine is lighter weight and smaller compared to its heavier single phase equal - but if you rectify each phase seperately you lose that advantage. The only advantages you have over a single phase by rectifying each coil seperately are smoother operation, and better use of the line - but there is no efficiency advantage over single phase in the alternator itself I don't think.

There are a couple of clear advantages to doing things the way you do:

if you have 4 coils per phase (or two) you can hook in series or parallel to match your alternator to different battery voltages.
you can wind with finer wire.

I think the drawbacks are lower efficiency - and rectifiers at the tower top.

I expect efficiency is similar either way - and both ways have their own set of pros and cons.

It would be fairly easy for you to continue your tests there Jerry - but the load for Star needs to have higher resistance to be equal because the voltage of the alternator will be higher. But I expect if you measure power in that machine you tested there - with loads that are properly matched to both conditions (that might be a bit tricky) you would see that you get more power per given rpm with Star. With Delta you can test with the same load - that would also be interesting. The tests you've allready done only show voltage so the key point is missing there. Great discussion - both of your last diaries were excellent Jerry - thankyou for putting fourth the time!

oztules · « **Reply #27 on:** February 11, 2006, 02:52:47 PM »

Jerry we must keep our frame of reference cohesive. three coils add up thier pulses averaged over time for the example you set in your pics.

You have now moved on to a nine coil arrangement. Lets look at this for a moment. There are two ways of looking at your 9 coil twelve magnet arrangement.

Your coils are 40 degrees apart, your mags are 30 degrees apart . This means coils 1,4 and 7 are in phase ie north pole will be directly above each of these three coils at the same time. at 0 degrees, 120 degrees, 240 degrees notice 40 and 30 divide into these degree marks, so that is where they line up or are in phase. the same angular argument exists for the other mag coil arrangements as they are symmetrical. so for delta in this arrangement, coils 1,4,7 are wired in parralell, coils 2,5,8 are parallelled coils 3,6,9 are parallelled. they are then wired into delta, and presto, the voltage and current will conform to your output figures.

ie treat each set of "in phase coils" as a single coil as per your gm alternator. We can do this as the power in each of the three inphase coils is identical at each point in time within the set. Their currents are cumulative as they are in phase, their voltage is not as they are in parrallel. When these three sets are presented at the output, the three sets of current pulses will arrive at the output at different times and average out over time as your 35 amps.

Case 2 if you have a different assymetrical arrangement of mags and coils, such that no coils produce in phase with any other coils, you will have produced 9 phase power. here it can be wired into 9 phase delta i guess, and each delta phase will be rectified seperatly, just as it is usually done. Now 9 pulses will arrive at differing times at the output. The currents will add, the voltage will not, and we are back to where we started..........why wouldnt we use this assymetrical arrangement? probably because the arrangement would not allow us to uniformly get a north pole and a south pole over the opposite legs of each coil uniformly, and so we would sacrifice some of our generating capacity. I think flux also alluded to the fact the polyphaseing over three was less efficient and thats without asymetrical magnet/coil placement. If we design symetrical 9 phase we could then parallel those in phase coils as per your design (ie 18 or 27 coils = 2 or 3 parallel sets times 9)

so yes i still think that the jerry rig is equivelent to delta. However it may offer some advantage in multicoil per phase applications ( say your nine coil arrangement) if the coils are a bit wayward (turns or placement). Here it will then be possible that an advantage would be gained by independently rectifying the outputs and parallelling their outputs. In your GM exanple, with three coils only, delta push pull would not be an issue, in multi-parallel, poor coil specs or mag placement may cause push pull between coils in phase. .............running out of finger prints if i dont stop typing soon..... oztules

Jerry · « **Reply #28 on:** February 11, 2006, 02:59:55 PM »

Well, Jerry left to work. Since I work out of my office at home most of the time, I decided to get on Jerry's computer and see what has been taking so much of his time. It's all "jibberish" to me! However, I have to tell you that Jerry is a man that gets an idea--and is not going to let go of it until he has satisfied his own curiosity. His mind is always working, and he is very dedicated to trying things that he thinks might work. He is not afraid of "trial and error". If we didn't have such a busy life with our retail store and such, I'm sure he'd be well on his way to something new and exciting in the "windpower" field. He really seems to be facinated with it all--and enjoying his experimentations.

I do have to say that my husband's spelling is terrible--but, I like humor--and his spelling has made me laugh many times over the years! (I don't know if he is going to take this as a compliment!) I'm sure he'd love it if I could understand everything that he shares with me in regards to the windpower. My father died a few years ago--but, he used to tell Jerry to just lift his head once in awhile and say, "yes dear"! That way, I wouldn't know if he was listening or not. Well, I do listen to some of what he tells me--but I can only "pretend" to understand a portion of what he's talking about!

Ok--I gave my two cents! Go get um Jerry!

Jerry's wife--

kitno455 · « **Reply #29 on:** February 11, 2006, 03:09:01 PM »

ma'am, there is a book that you need to get for jerry- 'a feersum injun' by ian banks. there is a character in it that talks just like jerry writes. i had a tough time reading this book. jerry will be able to breeze thru it.

jerry is terribly persistent, and at the same time self-effacing, which makes his posts a pleasure to read. someone using home-brew windpower for part of their unrelated livelyhood is an inspiration for those of us just starting out.

this post in particular has got alot of us thinking aloud.

allan

hvirtane · « **Reply #30 on:** February 11, 2006, 05:41:29 PM »

DanB: Actually Hugh is using a conventional

'star' connection, but two coils are running in

parallel, so there are 10 leads coming out

- but the output of any two leads

is not through one coil,

but through two coils,

which are out of phase through

a start connection,

just like any multiphase

machine wired in Star.

OK. I never seen it exactly how it is done.

I've only seen ready made stators by my

Friend in India according to Hugh's plan.

When there are 10 leads coming out

I've been thinking that

the coils are rectified separately.

---

I think that there is no need to argue

about the basic issues that 'delta'

will give more available power in watts

at high RPM than 'star',

but star will rise volts faster,

so 'star' is often the choice

for low RPM machines even if giving

less power than 'delta' when driving fast.

Please see for example the excellent manual

at itdg by Hugh Piggott:

www.itdg.org/docs/energy/pmg_manual.pdf

A good place about circuits:

http://www.allaboutcircuits.com

---

The difference between rectifying all

the phases separately and 'delta', I think

is really small.

I've read from the publication by Alfred Forbes

('The Homebuilt Dynamo Alfred T. Forbes, 1987')

that in practice he got better power by rectifying

all the three phases separately.

Also 'WindstuffEd' reported some time ago

a similar result.

Contrary Mr. Kragten from NL,

a great designer of small wind machines

told me that according to his measurements

a slightly better performance

was achived with 'delta'.

---

But.

On more íssue remains.

Is it a good practice to rectify

all the phases separately and then

connect the phases in series?

(I think that at least some capacitors

are needed for each phases.)

If this will work well, we could

get the advantage of getting the better

high speed power and voltages rising

quickly with low RPM?

Or would we get too much losses trough

the diodes involved?

- Hannu

johnlm · « **Reply #31 on:** February 11, 2006, 07:08:30 PM »

Hannu,

You said:

On more íssue remains.

Is it a good practice to rectify

all the phases separately and then

connect the phases in series?

(I think that at least some capacitors

are needed for each phases.)

Ive done this before for exactly the reason you stated. I had a rewound car alternator that really wanted to use a 5 ft prop but the cutin rpm was almost too high for anything but a very high TSR prop at 5 ft unless I waited for the wind speed to get to 10 or 11 mph. And the real high tsr would have had difficulty overcoming the cogging except at about 10 mph, so I separately rectified each phase and connected them in series. Each output did have to have a cap on it to work properly and get the output voltage one would expect. The output RPM/V went from 39 RPM/V with standard WYE to 24 RPM/V with the series connection but the output impedance went from 2.4 Ohms (Wye at 60 HZ) up to 6.6 Ohms (Series at 60 HZ)so even though I got the cutin down where I needed the output impedance went up by a higher factor than the voltage did so the power out went down more than I had hoped. I may have been able to improve the output impedance had I increased the size (more uF) of the filter caps but I did not have larger ones at a high enough voltage rating to try at the time. I may go back and revisit that with larger caps some time.

Johnlm

electrondady1 · « **Reply #32 on:** February 11, 2006, 07:36:55 PM »

glad you liked the "jerry rigged" designation. i don't have enough electrical knowlege or experience to take part in the debate. but even as a novice builder i can see were runing a jerry rigged stator could make up for inaccurate coils or misaligned ones. the idea of rectifying each phase separately is interesting . i am puting together a 32 pole machine. with 24 coils, the output of any one phase needs to pass through another 8 coils on it's way out .

News:

Author Topic: Delta, star, Jerry riged? (Read 18480 times)