ELECTRIC TRANSMISSION DELTA /WYE and SERIE/PAPALLEL stator

[Arlo1]

We need to see results already! Doc you should be just about ready to test yours? I am afraid luke might be right so if there is gains is maybe very small almost to the point you cant feel it. Some one actualy ride this already. If I wasnt preping for a car show I would be trying this my self.

 

[liveforphysics]

btw, i'm correcting all my previously understanding of Phase current and current multiplication. LFP and Justin are now convincing me to correctly better play with the dark side of electrons and copper!

Doc

This is outstanding news Doc! You're going to have the insight now to make better choices and matching up battery with motor with controller than ever before now, and the results are going to kick ass harder than ever before!

 

[Doctorbass]

My point on that is that if i would only go with a 5303 and try offroad with alot of hard accelerations and stop ( worst conditions), then the controller would overheat and probably blow.

By using the motor in 5306 mode the controller will less overheat and i will get the same torque or a bit more i think.

5306 will be most for the offroad ( it's a DH ebike after all).. and the 5303 will offer me a bit more speed... 5303 100V 24" wheel with 3.0" tire

The tire and rim are the exact same as the stealth bomber ( i want to try that setup style for fun)...

I remember that the controller used with my 5302 at 172A burst become VERY HOT!! and that the same conditions with my 5305 , make the controller still cold or a bit warm... A HUGE DIFFERENCE!!

Doc

 

[Doctorbass]

btw, i'm correcting all my previously understanding of Phase current and current multiplication. LFP and Justin are now convincing me to correctly better play with the dark side of electrons and copper!

Doc

This is outstanding news Doc! You're going to have the insight now to make better choices and matching up battery with motor with controller than ever before now, and the results are going to kick ass harder than ever before!

Hey Luke, What i can say and share with the E-S now about motor winding choice and controller limit is this:

Setup no 1:

Speed record Killabicycle project ( summer 2009)
5302 on 20" wheel
18 fets controller 4110 that burst at 172A fluke shunt calibrated with kelvin 4 wire method
100A battery current and 160A phase current
Block time : 1
100V battery
Acceleration: 0-70 kmh in 5 sec

-Result after 4 or 6 accelerations to 70 kmh for test: The controller get very HOT !!!! but not blow




Setup no 2:

Mongoose with new 18 fets controller ( summer 2010)
5305 on 20" wheel
18 fets controller 4110 that burst at 221A fluke shunt calibrated with kelvin 4 wire method
100A battery current and 250A phase current ( 2.5 ratio)
Block time : 1
100V battery
Acceleration 0-60 kmh in 5 sec

-Result after 4 or 6 accelerations to 70 kmh : The controller still cold or warm.. no real feel of any danger or heat on the case





my conclusion: in very close conditions, the 5302 offer similar performances as the 5305 but the 5302 make the controller become very HOT !!! and the 5305 dont.

My understanding is that higer turn winding motor are better but if you want to have both torque and speed and not overheat something.. we must use dual winding setup like relay switched stator


Doc

 

[liveforphysics]

I remember that the controller used with my 5302 at 172A burst become VERY HOT!! and that the same conditions with my 5305 , make the controller still cold or a bit warm...
Doc

I bet! That controller running the 5302 was pumping out 2.5times the current the 5305 controller would be doing when starting from a stop if they are both drawing in 172amps at the battery! Easily could have been in the >600phase-amp range starting from a stop.

Like I said, the series-parallel switching plan you've got has the ability to take away half or more than half the heat load on the motor for the first 0-20mph of acceleration or so, that could be pretty handy. But then the question becomes, is it easier to mechanically switch motor windings with relays inside the motor, or just make a controller FET stage capable of more currents (like maybe using some TO-264 package FETs on a custom heatsink).

 

[andreym]

I remember that the controller used with my 5302 at 172A burst become VERY HOT!! and that the same conditions with my 5305 , make the controller still cold or a bit warm...
Doc

I bet! That controller running the 5302 was pumping out 2.5times the current the 5305 controller would be doing when starting from a stop if they are both drawing in 172amps at the battery! Easily could have been in the >600phase-amp range starting from a stop.

Like I said, the series-parallel switching plan you've got has the ability to take away half or more than half the heat load on the motor for the first 0-20mph of acceleration or so, that could be pretty handy. But then the question becomes, is it easier to mechanically switch motor windings with relays inside the motor, or just make a controller FET stage capable of more currents (like maybe using some TO-264 package FETs on a custom heatsink).

In that case i believe the controller should be INSIDE the motor because just imagine the wires thickness to allow say 500-1000amps through phase wires? 1awg? 0awg?
and with controller located inside (at least power fets) you could easily power this with say 20kw at 100v and 200amps using 10-8awg)

 

[Doctorbass]

I remember that the controller used with my 5302 at 172A burst become VERY HOT!! and that the same conditions with my 5305 , make the controller still cold or a bit warm...
Doc

I bet! That controller running the 5302 was pumping out 2.5times the current the 5305 controller would be doing when starting from a stop if they are both drawing in 172amps at the battery! Easily could have been in the >600phase-amp range starting from a stop.

Like I said, the series-parallel switching plan you've got has the ability to take away half or more than half the heat load on the motor for the first 0-20mph of acceleration or so, that could be pretty handy. But then the question becomes, is it easier to mechanically switch motor windings with relays inside the motor, or just make a controller FET stage capable of more currents (like maybe using some TO-264 package FETs on a custom heatsink).

In that case i believe the controller should be INSIDE the motor because just imagine the wires thickness to allow say 500-1000amps through phase wires? 1awg? 0awg?
and with controller located inside (at least power fets) you could easily power this with say 20kw at 100v and 200amps using 10-8awg)

Not exactly .. Like Luke explained before, The phase current is not 100% time on every cycle of the AC .. it's duty cycle..

otherwise with 600A at 80V that would be 48000W and the battery dont do that... continuous battery current is transformed into Phase current that is just pulsated partially oin time on each cycle of the high frequency that the controlelr send into the motor winding...

Doc

 

[Doctorbass]

I remember that the controller used with my 5302 at 172A burst become VERY HOT!! and that the same conditions with my 5305 , make the controller still cold or a bit warm...
Doc

I bet! That controller running the 5302 was pumping out 2.5times the current the 5305 controller would be doing when starting from a stop if they are both drawing in 172amps at the battery! Easily could have been in the >600phase-amp range starting from a stop.

Like I said, the series-parallel switching plan you've got has the ability to take away half or more than half the heat load on the motor for the first 0-20mph of acceleration or so, that could be pretty handy. But then the question becomes, is it easier to mechanically switch motor windings with relays inside the motor, or just make a controller FET stage capable of more currents (like maybe using some TO-264 package FETs on a custom heatsink).

or... maybe just make a better mosfet dissipation... :wink: ON my present 18 fets controller i am using thick copper bar instead of aluminum and also i use kaptin tape instead of that grey rubber stuff as mosfet insulator... I guess it already help ALOT !!

but guess what.. .the doubled phase wire from the motor to the connector ( dual 12 gauge for all phase ) and ( all in 10 AWG for the controlelr side that also become REALLY HOTT!)

That 5302 is also a PURE current monster eater!

Doc

 

[John in CR]

That 5302 is also a PURE current monster eater!
Doc

The 5302 is a much better motor as far as I'm concerned. Look, it out performed the other one despite being handicapped by a lower current limit. Imagine what it would have done with a controller manly enough to feed it the current it wanted. The results wouldn't have been even close. It wanted and could handle double the current, but it got less current. As long as you're not bumping up against saturation limits of the stator it's capable of 0-70 in under 3 seconds with double the current.

John

 

[steveo]

That 5302 is also a PURE current monster eater!
Doc

The 5302 is a much better motor as far as I'm concerned. Look, it out performed the other one despite being handicapped by a lower current limit. Imagine what it would have done with a controller manly enough to feed it the current it wanted. The results wouldn't have been even close. It wanted and could handle double the current, but it got less current. As long as you're not bumping up against saturation limits of the stator it's capable of 0-70 in under 3 seconds with double the current.

John

5303 all the way!!!... Dual winding x5306-x5303 ... hard to beat!

 

[Doctorbass]

That 5302 is also a PURE current monster eater!
Doc

The 5302 is a much better motor as far as I'm concerned. Look, it out performed the other one despite being handicapped by a lower current limit. Imagine what it would have done with a controller manly enough to feed it the current it wanted. The results wouldn't have been even close. It wanted and could handle double the current, but it got less current. As long as you're not bumping up against saturation limits of the stator it's capable of 0-70 in under 3 seconds with double the current.

John

5303 all the way!!!... Dual winding x5306-x5303 ... hard to beat!

Did you tried yours??

 

[steveo]

i came close ... however i could not run the motor because i could not get everything to fit in the case... your pcb is what i want !!!!.. doc put me on the list for those pcbs .. i'll be interested in a few if they work out well!!

-steveo

 

[hunter]

doc i'm interested in getting on the list as well. my x5303 is all apart just waiting for your PCB boards.

 

[fechter]

The resistance losses (and therefore heating) are a function of current squared, so a 5302 is going to result in more heating for a given amount of torque, but will have a higher no-load speed.

Torque will be a function of (phase amps) x (number of turns).

Maybe a 5302/5304 dual wind would be good?

Those latching 60A air conditioner relays I found would be nice if they were SPDT. They do exist like that, but are expensive. To use the cheap SPST versions would take 9 of them for a series/parallel dual wind switch.

 

[tostino]

The resistance losses (and therefore heating) are a function of current squared, so a 5302 is going to result in more heating for a given amount of torque, but will have a higher no-load speed.

I don't think that is actually true fechter... A 5304 has a resistance of 203.4 mohm, a 5302 has a resistance of 61 mohm (about 1/4th of the 5304). If you need double the current in a 5302 to get the same torque, the heating will still be the same in both motors for the same torque, but the 5302 will have a much better top speed.

A 5304 at 150 amps will make 4576 watts of heat.
A 5302 at 300 amps will make 5490 watts of heat.

Both motors should produce the same torque. If there is the same amount of copper in both motors, they actually should produce the exact same amount of heat. I don't know if this is just an error in the resistance numbers that I pulled from ebikes.ca, or what. But you get the idea.

 

[liveforphysics]

The resistance losses (and therefore heating) are a function of current squared, so a 5302 is going to result in more heating for a given amount of torque...

No no no!
You shorten the wire by the same proportion that its thickened up. Shorter wire change in resistance of x, times thicker wire change in resistance of x, means x*x, makeing resistance drop at the square of turns, which balances perfectly with heat increasing at the square of current, to leave you with copper fill being the only thing that matters in determining the continous torque rate, and they all get identical heating for that torque value if the copper fill matches, etc etc etc...

 

[John in CR]

Wouldn't the exception to this heating issue during a given takeoff be that the efficiency of the higher turn count motor will increase faster and produce less heat? Also, since our controllers limit the current below the maximum the low turn count motor can take, won't the higher turn count motor have more real world torque as opposed to the theoretical equivalency? I believe the real benefits will show on steep climbs and during lower speed trail riding, where the higher turn count motor will have a distinct advantage by operating in its efficiency band while the low turn count motor is still trying to kill its controller during current limiting.

John

 

[Doctorbass]

Ok no.. again this debate... :wink:

Guys.. once i finished that PCB thing i'll try it and will post results and will see...

everythig i'm doing now for thst DH ebike need custom parts.. = welsing, laser cuting, cnc cuting, etc.. it take so much time to acheive!!! and guess what... it's vacancy period here so all industrial CNC and metal worker are 2 next weeks off!!! :x

I mist use imagination to finish that DH bike project before next monday!!! ( when i'll go to meet BlackArrow at 1000km fron here and by the time, see my familly)... The next days are crucial!!!

tomorrow night i should get the PCB, new torque arm/steel dropout, 12 lipo battery box and C-A... Alot of work to do . but that project will be the best to test the performances of that serie parallel method.. at high power!!

Doc

 

[John in CR]

Ok no.. again this debate... :wink:

Guys.. once i finished that PCB thing i'll try it and will post results and will see...

everythig i'm doing now for thst DH ebike need custom parts.. = welsing, laser cuting, cnc cuting, etc.. it take so much time to acheive!!! and guess what... it's vacancy period here so all industrial CNC and metal worker are 2 next weeks off!!! :x

I mist use imagination to finish that DH bike project before next monday!!! ( when i'll go to meet BlackArrow at 1000km fron here and by the time, see my familly)... The next days are crucial!!!

tomorrow night i should get the PCB, new torque arm/steel dropout, 12 lipo battery box and C-A... Alot of work to do . but that project will be the best to test the performances of that serie parallel method.. at high power!!

Doc

I hope you're riding the bike the 1kkm to meet up with BA. That would be a great shakedown cruise. 8)

John

 

[Doctorbass]

Ok no.. again this debate... :wink:

Guys.. once i finished that PCB thing i'll try it and will post results and will see...

everythig i'm doing now for thst DH ebike need custom parts.. = welsing, laser cuting, cnc cuting, etc.. it take so much time to acheive!!! and guess what... it's vacancy period here so all industrial CNC and metal worker are 2 next weeks off!!! :x

I mist use imagination to finish that DH bike project before next monday!!! ( when i'll go to meet BlackArrow at 1000km fron here and by the time, see my familly)... The next days are crucial!!!

tomorrow night i should get the PCB, new torque arm/steel dropout, 12 lipo battery box and C-A... Alot of work to do . but that project will be the best to test the performances of that serie parallel method.. at high power!!

Doc

I hope you're riding the bike the 1kkm to meet up with BA. That would be a great shakedown cruise. 8)

John

Hey.. nice idea.. but with 1.3kWh of lipo i wold need to charge every 50km.. that's 20 stop... 20 stop x 45 minutes charge is 15 hours of charge plus the 20hour of continous run...... 35hour of ebike..

I should add some confortable spring on that DH seat and bring a sleeping bag with me..!

i'll pass this time.. Justin already prooven that long 7000km ride can be acheived.. i'll leave him the credit for incredible long ride!

Doc

 

[Miles]

Ok no.. again this debate... :wink:

It's not really a debate - more like countering the propagation of a common misunderstanding.

 

[Ariane]

I hope you're riding the bike the 1kkm to meet up with BA. That would be a great shakedown cruise. 8)

John

Hi John,

Doc take his car to going here, but we have allot of quad trails here for his new DH ebike and he can test this series/parallel set up in this hostile environment :wink: . We can try to make some movie during those rides, and I'm really impatient to see and to ride with Doc, it will be fun no doubt about it

Good day!
Black Arrow

 

[fechter]

Ok no.. again this debate... :wink:

Sorry

Yes, for the motor I see your point about fill, but losses in the phase wires and conduction losses in the FETs will still increase.

I'm reminded of the Ecycle 'Solid Slot' configuration where it's like just a big bar of copper in each slot. If properly driven, seemed to remain pretty efficient.

The key to that system was using a dedicated inductor for the PWM instead of using the motor windings. With very low turns, the inductance of the motor gets pretty low, which you could counteract by using a higher PWM frequency. Motor windings tend to be pretty lossy at higher frequencies (mostly core losses), so they added a high quality inductor/freewheel diode ahead of the motor. Supposedly the lossed avoided in the motor windings more than made up for the losses in the separate inductor.

Long story short, with 2 turns, you might see some improvement if you added some kind of inductance in series with each phase wire.

 

[John in CR]

Ok no.. again this debate... :wink:

Sorry

Yes, for the motor I see your point about fill, but losses in the phase wires and conduction losses in the FETs will still increase.

I'm reminded of the Ecycle 'Solid Slot' configuration where it's like just a big bar of copper in each slot. If properly driven, seemed to remain pretty efficient.

The key to that system was using a dedicated inductor for the PWM instead of using the motor windings. With very low turns, the inductance of the motor gets pretty low, which you could counteract by using a higher PWM frequency. Motor windings tend to be pretty lossy at higher frequencies (mostly core losses), so they added a high quality inductor/freewheel diode ahead of the motor. Supposedly the lossed avoided in the motor windings more than made up for the losses in the separate inductor.

Long story short, with 2 turns, you might see some improvement if you added some kind of inductance in series with each phase wire.

Like what?....Do some smallish loops of our phase wires? If so, how many might do the trick?

 

[Doctorbass]

The motor serie parallel mod is finished!!!

Now i'm working on the controller and battery.

i'll post results soon!!

DOc