3 Phase Motor Redesign Winding

Well I just tossed some basic math at it and it suggests that you are trying to get about 20 Hp continuous and 50 peak out of a 10 HP rated motor. Unfortunately I suspect that you will run into some core saturation and winding over temp issues long before you get to the power levels you are hoping for even with a liquid cooled outer core.

You can cheat your HP some by running your motor over its stock RPM but there usually comes a point of diminishing returns for the effort as you get further over your motors stock frequency. Maybe 50% more HP at 2x the stock frequency dropping to maybe 60 - 70% more at 3x the base frequency provided the rotor does not fly apart first or you hit your cores peak realistic working frequency.

However this is just an educated guess at this point though.

Yes i understand about over saturation, here is the numbers, but will not be running

these kind of numbers very long 4 to 5 seconds

200 amp draw 200Hz

tooth density is 193387, max is 120000
Back iron density is 163228, max is 120000
Air gap density is 92285, max is 55000.

Do you have any relationship between winding turns and voltage and IN HAND turns.

Thanks Ivan

Ivanbennett.com

Here is more info on my motor. My goal for this motor is 80 volts rms.
The winding on my web site will not work, 15 gauge wire 15 turns all in hand,
at 60 volts the motor is pulling 50 amps, go to 90 volts amps are over 100 amps,
witch tells me we need more turns smaller wire. Only got 1 ohm resistance on coils,
delta connect. Tryed a Y connect still only 1.5 ohm. This motor needs to be delta
for torque.
My calculations for volts per turn shows 15 gauge 15 turns should of been ok, but going with full in hand turns seems to change the voltage..

Any Ideas

Ivan

http://ivanbennett.com Ivan@ivan5.net

This still sounds like that particular motors iron core is going into magnetic saturation which tells me regardless of how you change the winding numbers and sizes you are still going to be stuck with a simple issue of the Amp Turns magnetic field strength limitation due to core saturation.

Amp turns is that 1 amp going through 100 turns of wire generates the same magnetic field strength as 10 amps going through 10 turns or 100 amps going through one turn.

Basically you need to use a physically bigger motor if you want to get much more than 10 or so continuous HP.

I'm glad you are sticking with this pretty ambitious project.

at 60 volts the motor is pulling 50 amps, go to 90 volts amps are over 100 amps,

Are these figures with no load?? (If so, they are incredibly high.) If under load, what are you using for load?

15 gauge 15 turns should of been ok, but going with full in hand turns seems to change the voltage..

15 turns in hand (by which I assume you mean in parallel -- on your site you seem to show all wires connected to one lug) would be effectively 1 turn of very thick wire. I'd expect back EMF voltage to be very low, amperage very high. 1 ohm seems like high resistance for such stout windings -- but maybe you are actually measuring impedance with the rotor turning? (Are you measuring this with an ohmmeter of deducing it from amperage?) What is the length of each 15 gauge piece of wire?

K Fry

15 turns in hand (by which I assume you mean in parallel -- on your site you seem to show all wires connected to one lug) would be effectively 1 turn of very thick wire. I'd expect back EMF voltage to be very low, amperage very high. 1 ohm seems like high resistance for such stout windings -- but maybe you are actually measuring impedance with the rotor turning? (Are you measuring this with an ohmmeter of deducing it from amperage?) What is the length of each 15 gauge piece of wire?

I am measuring this both ways with a meter and a meger.

The lenght of each phase all four poles is 21.5 feet. 15 15gauge.

Had to lay the coils in one continuous lay, because the pole connections

would of been to large.

Ivan

My books say that 1000 feet of 15 ga copper is 3.25 ohms so given that 15 strands of 21.5 feet in parallel should work out to ((3.25/1000) x 21.5 = .0699)/15) = ~.0047 Ohms per phase. I think your ohm meter does not read low enough.

At 80 volts that should give you a theoretical peak DC amperage of around ~17,000 amps per phase. Now granted that at 20 amps on 15 ga copper is pushing its limits so even then that should give you around 300 amps per phase with proper cooling provided that you are using high temp enamelled copper wire that can handle the heat.

Still all that amp capacity wont break a stator core magnetic flux saturation limit.

If it was me I would have started out with a 20 - 30 HP OFC type motor that was rated for VFD service duty and went with a simple external fan cooling system opposed to the liquid cooling system.

This may be worth your time to read, http://www.itrc.org/reports/motor/motor.pdf ,being you are attempting to run a 50 -60 Hz designed motor stator core at up to 200 Hz.

tcmtech

My books say that 1000 feet of 15 ga copper is 3.25 ohms so given that 15 strands of 21.5 feet in parallel should work out to ((3.25/1000) x 21.5 = .0699)/15) = ~.0047 Ohms per phase. I think your ohm meter does not read low enough.

Ya but (dont ya love ya buts) that is all 3 phases 1000 feet, each phase

is only 333 feet / 15 wires 21.5 feet long, 1 ohm. These are kind of just round numbers.

I realy should have higher ohms per phase, resistance.

Ivan

tcmtech

The original question was, how does one figure the voltage for a motor
winding with 2-3-4 ect multiable inhand turns, if you did not no then
why didn't you just say so. Instead you had to chime in on saturation.

Ivan

The voltage to turns ratios stay about the same until you reach the energy level where magnetic saturation in the iron core starts to take over. After that it does not matter what you do to your winding ratios or how much more power you dump into it.

My reference books say a 10 Hp 3 Ph motor will draw around ~ 27 amps per phase at 230 VAC 50 -60 hz. Reduced down to an 80 volt input that would work out to around a 75 amps per phase at a 10 HP output.

So here is how this relates to turns ratios. Once you go past that designed spec HP and or frequency numbers you are reaching into the area of limited extra HP due to the iron core reaching its designed magnetic saturation point which is for the most part is a physical dead end limit to how much HP the motor can produce mechanically regardless of how much more electrical energy is forced into it.

You can dump loads of extra electrical power into it in the form of crazy high amps but still the magnetic forces acting on the rotor to make it turn will not follow the added amps input or power input once the saturation point it reached. That is once you reach magnetic saturation of your core the mechanical output stays about the same despite huge increases in electrical energy input. Past that point the differences between input power and mechanical power just makes waste heat and nothing else usable.

That's why I keep going on about core saturation and that's why I suspect your input amps will keep going up and changing the turns ratios will have little affect on it above the core saturation point. The point is changing turns ratios will not cure a core saturation or HP efficiency limit.

From what your website says I think you have your ratios about right for a 80 volt system but still your motor likely just has a natural limit on how much mechanical power its going to produce and in this case the engineers that designed its magnetic properties probably did not give it much extra headroom before magnetic saturation occurs.

I would be happy to hear any other theories from any of our other electric motor experts here to see if they have a different theory than mine. I could be way off the mark too!

I'm not sure what you mean by "in hand turns". As I understand it, one has wires in hand, not turns. But the relationship I think you are asking about is very simple. Say your motor is initially rated at 400 volts 20 amps and is wound with 4 wires in hand (i.e. in parallel, treated as if they were one larger wire) and say those 4 wires do 20 turns per pole. Then if you wanted it to become a 100 volt 80 amp motor (a "rewind factor" of 4), you would need to have 16 (of the same) wires in hand and only do 5 turns with them all. Everything just scales either up or down by the rewind factor.

And yes, you can get 4 times the instantaneous power out of the same motor, by doing such a rewind then running it at 4 times its nominal speed and its original nominal voltage (which is 4 times its new nominal voltage) provided the bearings can take it and the rotor doesn't break up with the centrifugal force. So you'd want a 4-pole motor for this, or more than 4 poles. A 2-pole would never survive.

The continuous power would not go up anywhere near in proportion (as another poster guessed) because the iron losses (read "heat") increase with frequency. But for electric vehicle use, this doesn't matter so much. Full power is only ever needed in short bursts.

There are several people who know about this stuff on http://forums.aeva.asn.au

As others have said, basically,"You cannot make a silk purse from a sow's ear."

There are physical limits on HP depending on design of the motor itself.

A motor may be slightly over-engineered for reliability, but not to the extent you are wanting.

It is not economical to manufacture a standard motor with that amount of overhead.

You can re invent the wheel, if you just like to tinker, but in my opinion, it is better to start with current technology and try to improve efficiency from there.

There are many links to EV projects on line, and sources for specialty motors,controllers, couplers,etc.

You may find that someone else has already been in your position, and the direction they took.

I am not saying everything has been invented or developed yet,the field is very young,and many advances will be made, but it pays to sometimes stand on the shoulders of others that have come before.From that vantage point you can see much further.At least look around from that vantage point.

If you see something from there that is a eureka moment, that will advance the art or science,even if it means discarding the status quo, you will have accomplished much.

Good luck.

If I read Mr. TCMTECH, and Mr. Weber's posts correctly, I believe the key area that need for you to examine and study is how the core was originally designed. The core's BH curve characteristics, the # of poles and size of core openings. As claimed, there is a need to increase the copper wire gauge that will enable higher amperage capacity thus will result to a higher HP rating. The # of turns/pole has to be decrease simply due to the limited available core open space!

I believe that you've addressed most of the factors needed in your project, except altering the physical field core design of the motor.

I also believe that you are well aware that there is only a maximum # of magnetic flux that can be produced for every turn of wire, by using a particular type of core material? This flux density characteristics of the core material maybe the one limiting factor for your project?

Beautiful workmanship on this project.

It is the AC reactance and not the DC resistance of the windings that is of concern.

I would run a simple test to measure the core saturation. With the rotor in place, use a variac to drive a single winding. Measure and plot the applied voltage -vs- current.

A better test would be to drive the motor no load with your inverter and vary the applied voltage. Measure and plot the applied volts/Hz -vs- current.

For a factory production line, we used to get 20 hp+ out of a 5 hp motor frame during acceleration. The duty cycle was low, about one 15 second acceleration every 15 minutes. The core of the motors could handle a boosted V/Hz and high acceleration current without going into saturation.

BTW - I don't understand the need for the transmission. Electric motors have a very high low end torque and the torque speed characteristics can be changed by varying the V/Hz.

Just want to thank everyone for there input.

I have found out that the in hand turns only count as one turn. So with that
in mind this is what i have calculated for my new wind.

1-I have 107184 cirular mils of space in each slot.
2-3.333 volts per turn. Calculated from original winding.
3-80 volts rms is the new wind. Curtis controller.
4-80 volts divide by 3.333 = 24 turns, we need to double the turns (slot fill)
because half of those turns (wires) are the in hand turns.
5-We will have 48 wires total slot fill, this a dual layer winding so each phase
will be 12 inhand 2 turns for 24 wires.
6-107184 divide by 48 = 2233 (is the circular mils of my new wire)
7-number 17 gauge magwire is 2052 circular mils. I will be using 17 gauge.
8-48 times 2052 = 98496 circular mils total slot fill.

This will leave me with enough room for a coil separator stick..

Ripped out that new winding today. Oh well, Practice makes perfect.

Ivansgarage
Ivanbennett.com

Contact:

David Whitelaw

Bestway Electric Motor Service

728 S. Campbell Ave

Tucson, AZ

This gentleman can answer your questions and if not he will point you in the right direction as he is very experienced, well educated, and extremely knowledgeable.

CR4 ADMIN - email address and phone number removed

From the CR4 Rules: Do not post phone numbers or email addresses. The CR4 Admin will delete all phone numbers and email addresses posted in threads or comments. You can share this information via the CR4 internal messaging system.

Bestway Was one of the first places I stoped by about 3 month ago

dont no who i talked to, but all they could tell me is they would not warranty

it (da) and they did not think I new what the hell i was talking about "100

150 contnuous amps 10 hp" no fricken way. I walked out.

I got my supplys and my first dip and bake from Jim at Arizona Electric Motor,

really a nice guy..

Any way after this next wind i will test it before i dip and bake..

I layed this whole thing by hand so I would not have any pole connection.

I use to wind motors some 30 years ago, and trying to find the relationship

between in hand turn and turns has been a real chalenge, I think I got it know.

K Fry said,

15 turns in hand (by which I assume you mean in parallel -- on your site you seem to show all wires connected to one lug) would be effectively 1 turn of very thick wire.

I have to say thanks K Fry, That one little sentence really rang a bell "would be effectively 1 turn" Got me thinking 1 turn, 1 turn, .. 1 wire or ten wires..

Ivan

You're welcome. Happy to help. I wish you great luck with your project.