wye to delta on the go m-controller ?

hello guys
sorry if my questions are stupid .
i wonder if today there are m-controllers available that can change the motor configuration from delta to wye and vice versa on the go , in order to get max efficiency for torque and rpm .
what i mean is the motor has 6 wires , and controller can reconfigure the winding on the go for wye or delta .
is such a thing available ?
do you also think this can improve efficiency ?
thanks

It is simple to do with a triple pole, double throw switch or relay. Use a normal controller for the 3 top ends of the windings, and tie the three bottom ends to the switch or relay. The switch wiring is very easy to figure out. It works.

Why? It's affect is more like having a common speed switch with slow set to 58% and fast to 100%, than having a change in gearing. There's no change in torque, so it won't help you accelerate harder, or pull a heavier load, or climb a hill better. In fact, it would just add extra wire to the motor windings outside of the housing with no productive benefit, ie increase resistance. Recumpence uses delta/WYE switching to good benefit on an excessive power bike to make it more tame to ride when in low, but common speed switching which changes the result of throttle voltage range has the same benefit.

Don't think of delta/WYE switching as an electronic transmission, because it's nothing of the sort.

John in CR said:

Don't think of delta/WYE switching as an electronic transmission, because it's nothing of the sort.

Click to expand...

Sorry, but it is a speed/torque change exactly like a gear change. The energy in/out stays the same with the Kv and Kt moving oppositely. When the top speed is reduced 58%, the torque at 0 RPM is increased 58%. Exactly the same as the difference in a hard wired delta compared to a hard wired wye motor.

You get the exact same kind of transmission-equivalent gear change of 2:1 (50%) if you double wind the motor, and use switches to switch between two parallel winding strands or one serial wind. That requires three DPDT switch per wind, or one 6PDT switch. Both the delta/wye and the parallel/serial switches can be used in the same motor to have four unique "gear ratios" which is four Kv/Kt values for the motor, but that is a huge range of Kv for a bike motor, even when going from steep climbing to very fast down-hill street racing!

so no solid state electronics that can do that inside the controller , without relays and such ?

Alan, what you're saying is only true if you're driving the motor with an inadequate current source. Most ebikes are torque-limited by motor heating - there is usually enough current available from the battery/controller to exceed what the winding can take. In this scenario, changing from delta to wye won't give you any more torque capability to climb a steep hill. It's the same argument as "slow" or "fast" motor windings with equal copper fill - they all have the exact same power/torque/efficiency potential when driven by an appropriate voltage/current source.

Punx0r said:

...they all have the exact same power/torque/efficiency potential when driven by an appropriate voltage/current source.

Click to expand...

Most people have a power source with a fixed maximum voltage and a more-or-less limited current capacity at (near) that max voltage. Changing Kv will allow one to get the most out of that power source over a different range of speeds, with corresponding change in torques. Exactly as a mechanical transmission allows one to optimize the torque and HP (speed time torque) limits of any given ICE. That is, for a fixed voltage and current, the delta/wye switch will result in a different terminal speed and a different acceleration to that speed due to a changed torque. Exactly as a mechanical two speed transmission with a square root of 3 to one difference between the gears on the output shaft of the motor would change the final velocity and acceleration to that speed. Keep the power supply (battery and fully capable motor controller) exactly the same and the results are virtually indistinguishable from a mechanical transmission.

alan said:

Punx0r said:

...they all have the exact same power/torque/efficiency potential when driven by an appropriate voltage/current source.

Click to expand...

Most people have a power source with a fixed maximum voltage and a more-or-less limited current capacity at (near) that max voltage. Changing Kv will allow one to get the most out of that power source over a different range of speeds, with corresponding change in torques. Exactly as a mechanical transmission allows one to optimize the torque and HP (speed time torque) limits of any given ICE. That is, for a fixed voltage and current, the delta/wye switch will result in a different terminal speed and a different acceleration to that speed due to a changed torque. Exactly as a mechanical two speed transmission with a square root of 3 to one difference between the gears on the output shaft of the motor would change the final velocity and acceleration to that speed. Keep the power supply (battery and fully capable motor controller) exactly the same and the results are virtually indistinguishable from a mechanical transmission.

Click to expand...

Nope. It's totally different. With a mechanical transmission you use a lower gear to climb hills or otherwise pull a heavier load. If you do what you're talking about to climb a hill, then you're going to burn the motor up when switched to lower speed. What you're not taking into account is that changing Kv means changing resistance in exactly the proportion that the motor will make the same heat for the same torque. If you send the same current through a lower speed winding or phase wire termination, then yes it will make more torque, but also more heat. You can easily accomplish that same result with a Cycle Analyst or more modern controllers without the added expense or additional failure points.

The true result is simply a fast or slow setting, and that's what a throttle is for. If you need more fixed limits, then simply use a 3 speed switch. The newer controllers I use now have 3 switch on the fly current limits along with the common 3 speed switching, and they also have 2 switch on the fly throttle response settings. These features give me all the variability I need, though I wouldn't say any are like a mechanical transmission, which is simply unnecessary with a properly geared electric motor of appropriate size.

I wish delta/WYE or series/parallel switching had the benefits you speak of, but they simply don't. Otherwise every manufacturer would put it on their vehicles.

John in CR said:

alan said:

Punx0r said:

I wish delta/WYE or series/parallel switching had the benefits you speak of, but they simply don't. Otherwise every manufacturer would put it on their vehicles.

Click to expand...

Click to expand...

Click to expand...

I believe the only reason you don't see winding switches is because you need a pretty hefty switch or relay for gains that are easier to mostly achieve with only a small cost in efficiency by means of a controller, as you say. However, I do believe such switching would make excellent use in very large motors, such as are used in locomotives where startup torque needed is radically different from cruise conditions and the optimization of efficiencies might be worth while. Note that I believe the mechanical switching will be more efficient with less motor heat per final output power generated than possible with a speed controller, but that does not in any way eliminate the need for the controller. Since the differences will be small, the switch only adds complexity to the system with little gain.

However, threads discussing this in Endless-sphere (search delta-wye switch) are very clear that the switch will provide the user with two distinctly different motor characteristics. The wye provides max acceleration at low speeds, while the delta provides much higher top speeds with more docile acceleration. You just won't see such differences with controller changes that limit max current, or basic throttle control.

I can't speak for locomotive drives, but large industrial AC induction motors frequently used wye-delta switching for start-up and it's purely to lessen the peak current draw from the supply. It's not used on smaller motors because the available current supply is sufficient to start in delta. So yes, wye does produce greater torque, but only where you are supply-limited.

Punx0r said:

I can't speak for locomotive drives, but large industrial AC induction motors frequently used wye-delta switching for start-up and it's purely to lessen the peak current draw from the supply. It's not used on smaller motors because the available current supply is sufficient to start in delta. So yes, wye does produce greater torque, but only where you are supply-limited.

Click to expand...

The difference between delta and wye wiring is exactly the difference between the Kv and Kt of the motors differing by the square root of three. You can get the same difference with a different wind count and appropriately adjusted wire gauge. The peak efficiency of a motor wound either way is going to be when the motor is loaded down to near one half the Kv no-load speed, assuming the electrical supply and controller are appropriate.

For whatever reason one is trying to match a motor Kv to some application, the delta-wye or parallel-serial type of switching allows for multiple Kv choices from the same physical motor for the cost of the switch itself (cost, weight, reliability, size, etc.). Any voltage or current power supply limitations or motor controller current limitations can only be helped by such a switch. Exactly the same benefits possible with a mechanical transmission, except the mechanical transmission offers far more gear ratio options and far more cost and less reliability than do the switches. The speed/torque switching is the same.

It's really a bit misleading, especially to people who don't know this stuff, to claim that changing between Delta Wye has "exactly the same benefits possible with a mechanical transmission". This is absolutely not the case. A variable mechanical transmission allows the torque required from the motor to be lowered. Delta/Wye switching does not.

alan said:

The peak efficiency of a motor wound either way is going to be when the motor is loaded down to near one half the Kv no-load speed, assuming the electrical supply and controller are appropriate.

Click to expand...

No, that's wrong. Peak efficiency is achieved when the motor is loaded to the point where copper losses are at parity with parasitic losses. If a motor was loaded to half the no load speed, the efficiency would be close to 50% - same for all PMS motors....

IIRC peak power is at 50% of no-load speed? So perhaps a simple confusion.

Miles said:

A variable mechanical transmission allows the torque required from the motor to be lowered. Delta/Wye switching does not.

Click to expand...

Nicely put. That seems to be the essence of it.

alan said:

However, threads discussing this in Endless-sphere (search delta-wye switch) are very clear that the switch will provide the user with two distinctly different motor characteristics. The wye provides max acceleration at low speeds, while the delta provides much higher top speeds with more docile acceleration. You just won't see such differences with controller changes that limit max current, or basic throttle control.

Click to expand...

People talk about all the wonderful benefits while in the planning stage, and then once they pull it off it's a big letdown. That's because in the real world the controllers behave differently pumping into the different load. Your view also doesn't take into account the steeper torque curve of the slower winding/termination, which in your perfect scenario results in torque of the faster setting quickly surpassing your so-called "high torque" mode. With acceleration times in minutes like with locomotives, it makes sense that they'd take advantage from the small difference of delta/WYE switching.

I have several motors with mechanical switching of how they are terminated, and here in the real world with the controllers I've used the result is only barely noticeable greater torque on launch, resulting in nothing more than a fast setting and slow setting. I never bother with slow, because without a lowering of the controller's current settings the motor is at far greater risk of over-heating while in low. A number of people have done delta/WYE switching on ES, and as far as I know only one person is pleased with the results, and that's Recumpence. He's uses it in exactly the opposite manner you mention, probably because the RC controllers he uses deliver current differently. He uses the slow setting for tame easy riding (not greater torque on launches).

If it was any good, everyone would do it. The steel and magnets determine the torque potential of our motors, and how the copper is wound and terminated determines the combination of voltage and current to reach that potential. Changing windings or termination has no meaningful benefit. In the OP's case, he's talking about switching done at the controller, so it's actually a detriment due to the added wiring and unnecessary resistance.

Miles said:

alan said:

The peak efficiency of a motor wound either way is going to be when the motor is loaded down to near one half the Kv no-load speed, assuming the electrical supply and controller are appropriate.

Click to expand...

No, that's wrong. Peak efficiency is achieved when the motor is loaded to the point where copper losses are at parity with parasitic losses. If a motor was loaded to half the no load speed, the efficiency would be close to 50% - same for all PMS motors....

Click to expand...

OK. I (mostly) give! I do not agree about peak efficiency, as getting most motors over 90% out power/input power is not going to happen near min or max speed, and the system can most easily be designed, as for as I have seen, to get these high efficiencies around half peak speeds. I acknowledge there are tons of factors that determine peak speeds of a motor.

I agree that the torque required from a motor does change with a mechanical transmission, but not from a delta/wye switch. However, the current needed to achieve a given torque does. This does affect efficiency, and most of this reasoning will break down near current limits of the motor or controller or battery, of course, where a mechanical tranny can alleviate some of these issues.

My experience was a single prototype motor suitable for a small RC plane that I built with the delta/wye and parallel/serial switches. I was surprised how well it worked, but did not have all the test equipment I needed to accurately measure efficiencies (no calibrated load). I used a lab power supply and a much larger controller than needed (20A for a motor that could barely handle 10). I was surprised that you can even mix Kv by switching one or two, rather than all three winds with the parallel/serial switches and get the mixed Kv working and not destroying the controller, but I did not stress test that!

Hi Alan,

The loaded speed (relative to N.L. speed) that peak efficiency occurs at, varies with the efficiency itself. A 90% peak efficiency motor will achieve that when loaded to approximately 90% of N. L. speed. As PunxOr said, if a motor is genuinely loaded to half its no load speed then it's outputting its maximum possible power and the efficiency will be close to 50%. Obviously, the voltage sag of the batteries needs to be taken into account, too.

If you read Chapter One of this book, things might become clearer.
https://books.google.co.uk/books?id=jjuTYtKokc8C&printsec=frontcover&dq=fundamentals+austin+hughes&hl=en&sa=X&ved=0ahUKEwjlyNaWhZ_LAhVHthoKHaAuDtIQ6AEIJjAA#v=onepage&q=fundamentals%20austin%20hughes&f=false

Regarding heat generated per unit of torque, changing from delta to wye makes absolutely no difference. The change to Kt is exactly balanced by the change in phase to phase resistance.

Miles said:

If you read Chapter One of this book, things might become clearer.
https://books.google.co.uk/books?id=jjuTYtKokc8C&printsec=frontcover&dq=fundamentals+austin+hughes&hl=en&sa=X&ved=0ahUKEwjlyNaWhZ_LAhVHthoKHaAuDtIQ6AEIJjAA#v=onepage&q=fundamentals%20austin%20hughes&f=false

Click to expand...

Looks like a good book. Chapter one was complete there but most of the rest was hidden. Might have to buy the darn thing. I do find it interesting the author's use of "motional EMF". That might have helped explain things to Mr. Bowser a few weeks ago.... Nah, probably not.

Re: Wye/delta. To me, just a method of voltage control. It's useful for machines run on the mains for starting, but what's the sense for electronically controlled motors where you have control of the voltage?

I think, Chapter One is the best introduction to motor theory that I've come across. I'll add the link to "Resources".

Miles said:

The loaded speed (relative to N.L. speed) that peak efficiency occurs at, varies with the efficiency itself. A 90% peak efficiency motor will achieve that when loaded to approximately 90% of N. L. speed. As PunxOr said, if a motor is genuinely loaded to half its no load speed then it's outputting its maximum possible power and the efficiency will be close to 50%.

Click to expand...

Hi Miles,

I'm just a little vague on why efficiency is related to load at a percentage of N.L. speed. Seems like a strange metric. I am used to seeing efficiency maps overlaid on the speed-torque grid like this:



So from this I see, at say 60Nm, 90+% eff from 400 to 5000 RPM and 95+% eff from 800 to 2600 RPM. There is a pretty large area of high efficiency performance. I wonder why some hunt for "peak" or "sweet spot" and want to add things like shifting transmissions to accomplish a gain of a few percent in the motor when their method brings in additional loss, mass and cost.

Regards,

major

Hi major,

Torque is pretty much proportional to phase current. So, let's a assume constant input voltage. Because of the phase to phase resistance, the bEMF and shaft speed fall until enough current can flow to balance the load. All the losses are balanced by the drop in shaft speed.

That's all I was saying and it was in response to:

alan said:

The peak efficiency of a motor wound either way is going to be when the motor is loaded down to near one half the Kv no-load speed, assuming the electrical supply and controller are appropriate.

Click to expand...

Change the throttle or the input voltage and you change the speed and the torque for peak eta.

major said:

There is a pretty large area of high efficiency performance. I wonder why some hunt for "peak" or "sweet spot" and want to add things like shifting transmissions to accomplish a gain of a few percent in the motor when their method brings in additional loss, mass and cost.

Click to expand...

Me too. Usually, the only reason that justifies having variable gearing is if the motor is power limited and incapable of providing the torque required.

That Brammo graph looks like it's heavily influenced by the controller as opposed to what the motor would naturally do. Is that because it's an AC motor with inverter vs the BLDC motors we more commonly use and talk about on the forum?

OR can we just look and see where that peak 96% efficiency falls, ie in the region of the lower half of the torque range, to conclude that it's a throttle position result. Best efficiency falling in the lower half of both torque and rpm, just feels weird to me.

John in CR said:

That Brammo graph looks like it's heavily influenced by the controller as opposed to what the motor would naturally do. Is that because it's an AC motor with inverter vs the PMDC motors we more commonly use and talk about on the forum?

Click to expand...

Hi John,

By more common PMDC motors do you mean commutator type without controller? Speed torque would just be straight line. Otherwise aren't all electronically commutated motors heavily influenced by the controller? Anyway, I just used that particular graph because I had it handy and I thought it was "typical". I think I lifted it from elmoto (.net) and it was from something said about the Brammo race bike. I don't know anything more about the motor, although it is very nice. I especially like how the peak efficiency regions fall in zones I think would represent highway cruising.

John in CR said:

OR can we just look and see where that peak 96% efficiency falls, ie in the region of the lower half of the torque range, to conclude that it's a throttle position result. Best efficiency falling in the lower half of both torque and rpm, just feels weird to me.

Click to expand...

I agree it is odd looking in that regard. But actually nice in my opinion. It may be that it is an IPM rotor??? Don't know and probably never well. I don't understand your reference to throttle position. Isn't the entire map except the perimeter dependent on throttle position?

major

Sorry, I meant to type BLDC. edited

Comparing these two, drop the 94-96 contours from the top map and cut off the bottom map at 5000RPM, and they don't look so different. I think the Leaf motor is PMAC.