Max safe Flux Weakening

hias9

1 kW
Joined
Jul 11, 2018
Messages
422
How many amps of flux/field weakening would you consider safe without risking to destabilize the magnets?
Motor is a MXUS 3k v3 3T
 
hias9 said:
How many amps of flux/field weakening would you consider safe without risking to destabilize the magnets?
Motor is a MXUS 3k v3 3T

As far as I remember, in a perfect scenario you can increase rpm with flux weakening until the point the motor's speed would be capped by available power.
So in reality the required torque for the set rpm(thus road speed) should be achievable with the amps you can supply. If you don't have enough, your motor can seize.
 
I think the maximum value I can set in the Sabvoton software would be 150A.
For the Lipos this would not be a problem.

Isn't there a risk to destabilize the motor's magnets with this amount of flux weakening?
 
hias9 said:
I think the maximum value I can set in the Sabvoton software would be 150A.
For the Lipos this would not be a problem.

Isn't there a risk to destabilize the motor's magnets with this amount of flux weakening?

Flux weakening uses a lot of amps and that cooks your motor magnets faster. So it is amps>heat>magnets. I don't think there is a amps>magnets link where the magnets die directly. So as long as you aren't overheating, it shouldn't matter.

On concrete max values for the mxus, I would ask for someone with one, or in the mxus thread if there is one.
 
There is data in the MXUS 3K / 45xx threads that the saturation current is around 250A, though I don't recall which wind that's for. Saturation is basicaly the point above which further current doesn't make more magnetic field, but instead starts making more heat.


I'm testing the SFOC5 with phase currents of 250A (150A presently due to a broken axle on one side) in an MXUS 4503 v2; it heats up pretty quick at the 250A level with just a few 0-20MPH accelerations down my street. Not using any field weakening, as I don't need speed, just torque. ;)
 
If max. phase amps would be 350A and max field weakening would be 50A, would the max amps going through the phase wires be 400A or still 350A?

If flux weakening if activated, is it also active and drawing power at lower speeds?
 
Torque current and Field Weakening current are 90 degrees apart, so the sum would be the vector sum rather than the linear sum, which is more than either but less than their arithmetic sum. (Square root of the sum of the squares in this case).

The algorithm in the controller will decide when the Field Weakening current is applied as well as how much, and so depends on firmware, but there is no reason to apply it when the torque current is high (and speeds are well below the motor's normal full speed). It is when the torque current drops off (back EMF is high and the motor is nearing full speed) that there is advantage in adding FW current. Depending on how they did it, the total may well never exceed the 350A value, it is just distributed between the two vectors. But it is hard to know for certain.
 
Related Flux Weakening question...

If I have a 50A battery limit due to my BMS, can I go faster by using a portion of the 50A for field weakening or would I go faster using the entire 50A without any field weakening?

My particular application is a 6T MAC running on 52v.

Thanks...I realize this is an old thread just hoping somebody is still monitoring it :D .
 
Although you've probably found out already, I'm just going to put this for posterity:

BLDC / FOC controllers with field weakening only start to apply field weakening once the back-EMF from the motor reaches the battery voltage. This only occurs when you are going fast enough to generate enough BEMF that the BEMF plus the voltage drop due to phase resistance equals battery voltage.

So if you're using no field weakening and not pulling 50 amps at your top speed, field weakening will make you go a little bit faster.
If you're already pulling 50 amps without field weakening, the controller will never have a chance to apply field weakening even if you enable it.

Edit: Correction on phase resistance (in italics)
 
thorlancaster328 said:
Although you've probably found out already, I'm just going to put this for posterity:

BLDC / FOC controllers with field weakening only start to apply field weakening once the back-EMF from the motor reaches the battery voltage. This only occurs when you are going fast enough to generate enough BEMF to equal battery voltage.

So if you're using no field weakening and not pulling 50 amps at your top speed, field weakening will make you go a little bit faster.
If you're already pulling 50 amps without field weakening, the controller will never have a chance to apply field weakening even if you enable it.

Thanks...and yes, I am still around. A little older but still playing with electricity and trying not to get shocked :lol:.
 
thorlancaster328 said:
BLDC / FOC controllers with field weakening only start to apply field weakening once the back-EMF from the motor reaches the battery voltage. This only occurs when you are going fast enough to generate enough BEMF to equal battery voltage.

They actually apply field weakening much ealier (but not before 50% of no-load speed) so there is more torque and power available above 50% of no-load speed.
 
Based on my understanding of FOC (I have built a few controllers), Field Weakening (AKA injecting D-axis current to make the effective magnet strength less) is pointless unless you are trying to push more current into the windings than the phase resistance and back-EMF will allow you to. Before the output stage begins to saturate (reach near 100% PWM), it is always more efficient to increase the Q-axis current to produce more torque.

You are correct that this happens below no-load speed, the resistance of the phases makes this happen. But for a low-turn-count motor and/or at lower phase currents, the voltage drop across the phase resistance is minimal and these speeds are pretty close.
 
Given all options as being available....

Would it make more sense to simply rewind the motor with a higher kV?

Flux Weakening is a kind of "hack" where you throw enormous waste at increasing motor rpm when a higher kV would get that extra rpm at full efficiency.

I cannot think of a reason to do Flux Weakening if altering the kV is an option.

And heck... if you do a rewind with added copper fill you could change the kV and even improve efficiency at the same time.

Win, win.
 
SafeDiscDancing said:
Would it make more sense to simply rewind the motor with a higher kV?
There is an "ideal" wind where the speed you are most frequently at is close to the base speed of the motor. That way your controller is running with high PWM ratios (which makes it more efficient) and you don't need to use phase advance/flux weakening. In that case, using field weakening to get occasionally higher speeds makes sense.

For a concrete example, let's say you have a commuter ebike where you spend most of your time doing 20-25mph. But you have to get on a main road for a mile, and for that mile you really want to hit 35mph so you can (almost) keep up with traffic. In that case, choosing 25mph as 90% of base speed would probably make sense. That maximizes your efficiency at the most common speed. You lose efficiency on that one stretch of road but you make up for it with the other 10 miles.

If you instead chose 35mph as your 90% point you would often be running your controller at less than 50% duty cycle - and you would start to lose efficiency. You might even end up with lower overall efficiency for your route with the higher base speed.[/quote]
 
JackFlorey said:
You might even end up with lower overall efficiency for your route with the higher base speed.

If you use the Grin motor simulator you don't see huge differences though.

A 10% increase in kV will translate to maybe a half a percent efficiency reduction at the base speed.

Flux Weakening is massively inefficient because you are skewing the phase timing.

My "guess" is in most cases the simple 10% increase in kV is better.

What are the losses for Flux Weakening?

As far as I have heard it's like a massive 40% waste of power being turned into heat.
 
SafeDiscDancing said:
If you use the Grin motor simulator you don't see huge differences though.

A 10% increase in kV will translate to maybe a half a percent efficiency reduction at the base speed.
Agreed. They are not huge. They may be significant for longer commutes.
 
JackFlorey said:
Agreed. They are not huge. They may be significant for longer commutes.

A half a percent?

Compared to a 40% loss of efficiency using Flux Weakening?

Like I said on my first response I think the Flux Weakening was always a "hack" for people with a situation where they didn't want to rewind their motor but get a little more top speed anyway

Unless I have the basic idea wrong ALL the Flux Weakening creates nothing but wasted heat and the further above your default kV the worse it gets.

One thing we seem to see repeated often is ideas that are "trendy" but are actually nonsense.

If someone can explain how Flux Weakening would ever be a good idea if you were starting from scratch I'd love to hear the argument.

Is the heat loss as bad as 40%? Or is that number wrong?

To really justify Flux Weakening you need to compare doing that with a properly configured kV motor working as a normal motor should work.
 
Flux weakening is good on IPM motors where a considerable portion of the power is made using reluctance torque anyway.
https://www.automate.org/industry-insights/interior-permanent-magnet-motors-power-traction-motor-applications

So, SOME motors paired with SOME controllers (middrives, basically) are indeed capable of quite effective flux weakening.
On a typical PM BLDC motor, flux weakening is more of a gimmick and is a huge battery drain.
 
SafeDiscDancing said:
A half a percent?

Compared to a 40% loss of efficiency using Flux Weakening?
Like I said, you have to do the math and look at the total overall efficiency for the trip. If you lose 40% (which you don't) for 1% of the trip, but pick up .5% for the rest of the trip, you still come out ahead.

Like I said on my first response I think the Flux Weakening was always a "hack" for people with a situation where they didn't want to rewind their motor but get a little more top speed anyway
Flux weakening is just an extension of centuries of actual motor drive experience. In the days of separately excited rotors, you just reduce the excitation current and get a higher base speed. It actually saves energy at the expense of torque. When considering induction motor control you explicitly control the two magnetic vectors separately, so the relative strengths of the fields are whatever you make them.

The one advantage that BLDC motors have is that the field is already present via the permanent magnets, so you don't have to "waste" magnetic flux (created via amp-turns) to create the basic field. But as with induction motors, you can still control the field to get the performance you want. You are just enhancing or retarding the existing field. And even if you (slightly) oppose the field via field weakening, you are still getting the benefit of not having to create the field to begin with. And since losses go by I2R, opposing the field by 10% incurs a very small penalty due to the low delta currents required.

Unless I have the basic idea wrong ALL the Flux Weakening creates nothing but wasted heat and the further above your default kV the worse it gets.
That's true. But it's just as true to say that all induction motors create nothing but wasted heat with the induced field in the rotor. In a very narrow sense that's true - but that's also why they work, and the power they create is pretty useful.

Or consider it this way. Every switch-based controller out there incurs losses due to the finite on-resistance of the switches and AC losses that happen due to that switching. You could eliminate them with a commutator. Still, using a controller gives you some advantages - wouldn't you agree?
 
BalorNG said:
Flux weakening is good on IPM motors where a considerable portion of the power is made using reluctance torque anyway.
https://www.automate.org/industry-insights/interior-permanent-magnet-motors-power-traction-motor-applications

So, SOME motors paired with SOME controllers (middrives, basically) are indeed capable of quite effective flux weakening.
On a typical PM BLDC motor, flux weakening is more of a gimmick and is a huge battery drain.

I think this is the definitive answer.

Flux weakening was from the induction motor background.

Essentially the entire "slip" mentality of the induction motor was built around shifting the relative advance of the phase timing.

As you move further away from the induction motor through the "in between" designs where the magnets are embedded deeply inside the rotor you still can get away with it, but it's losing it's effectiveness.

By the time you reach the fully "permanent magnet surface mount" PM BLDC you have nearly completely lost any argument for doing it.

Try to imagine what Flux Weakening is ACTUALLY doing...

In order to advance the kV above it's default rotation speed you are literally trading heat for speed.

That's nuts.

That's just throwing battery power away for nothing.

----------------------------

And finally... if you then try to argue:

"Oh, but I'll be good. I'll only use Flux Weakening for 1%."

...well you know that's bullshit.

Human psychoplogy means everyone pushes their bike at full power pretty much every opportunity they get it.

So in the "real world" if you add Flux Weakening the losses will be huge because the temptation to use it will be too great.

Better to just get the right kV in there from the start and abandon this "super trendy nonsense".

It's like Regen.

The topic has been analyzed forever and in reality an ebike due to it's physics (high wind resistance, low weight) means you cannot recapture more that a very low percentage to the battery.

One thing we see across all areas of humanity now is this obsession with things that "sound pretty" and so people conform to the idea because the official narrative calls it "pretty or a fact" when in reality it's nonsense.

Objectively we can measure what really works and what doesn't.

Flux Weakening and Regen are both "feel good, happy talk" but not really the real world.

The real world is you design your motor with the right kV and make it as efficient as possible as a baseline. That's it.
 
It’s actually a lot more difficult than ”all motors are the same”

As balorNG already said: IPM motors can gain efficiency and torque in every speed range by field weakening. That’s why MTPA control algorithm or lookup tables of d-axis&q-axis current vs rpm is needed for optimal drive of these motors. They overturn the whole concept of ”what’s the kV, multiply it by volt for the max rpm”
 
SafeDiscDancing said:

Try to imagine what Flux Weakening is ACTUALLY doing...

In order to advance the kV above it's default rotation speed you are literally trading heat for speed.

That's nuts.

That's just throwing battery power away for nothing.

----------------------------

And finally... if you then try to argue:

"Oh, but I'll be good. I'll only use Flux Weakening for 1%."

...well you know that's bullshit.

Human psychoplogy means everyone pushes their bike at full power pretty much every opportunity they get it.

So in the "real world" if you add Flux Weakening the losses will be huge because the temptation to use it will be too great.

Better to just get the right kV in there from the start and abandon this "super trendy nonsense".

It's like Regen.

The topic has been analyzed forever and in reality an ebike due to it's physics (high wind resistance, low weight) means you cannot recapture more that a very low percentage to the battery.

Although I can understand your reasoning, I do not agree. It is too narrow of a view. To me, both regen and flux weakening are just extra design options that you can use when building your ideal ebike. To start with regen: regen is actually very useful and effective when you use it for what it actually is: an electronic brake. It saves your brake pads (very useful for heavy, fast ebikes), provides a third brake option for extra braking power and safety in addition to your front and rear brake discs, and when properly tuned it can act as an anti-stall brake system in slippery conditions like a snowy road. And additionally you get the free benefit of some extra mileage, 2-10% in my applications. So that’s a win-win-win in my view; I use it on all my bikes!

The same goes for flux weakening: it is also far too simple to say that it is a bad design choice to use it. I have both a 3T mxus running at 72v 150 battery amps and 300 phase amps, as well as a 4T mxus running at 72v 120 battery amps and 200 phase amps. Both have equal acceleration (200 Nm; just keeping the front wheel on the ground when leaning very far forward over the handle bar) up to some 40km/h. Then the 3T becomes faster. I mostly use my bikes up to 50km/h in hilly forests but also want them to go at least 80 km/h on flat roads occasionally. The controller, battery, cables and motor of the 4T stay cooler at low speeds (up to 50km/h), but it cannot reach 80km/h. The 3T is faster, but also less efficient at high amps (slightly less copper fill, more losses + voltage drop in the battery, cables and controller). So maybe in this use case the better design choice would be to use the slower 4T motor + flux weakening for high torque and high efficiency at low speeds in the forest, while still being able to run 80 km/h on flat roads where the additional amps that both the battery and controller can provide can be used for a nice speed boost. To me, this is also a win-win and it is not a no-brainer to immediately think the 3t motor is the better design choice, although all parts of the bike are able to provide the required amps.

So in summary, it is not a black-and-white story, both regen and flux weakening can be very valid design choices. With foc controllers, we just have the luxury that we can choose our optimal combination of motor + controller + regen + flux weakening for our specific application!
 
[/quote]
Although I can understand your reasoning, I do not agree. It is too narrow of a view. To me, both regen and flux weakening are just extra design options that you can use when building your ideal ebike. To start with regen: regen is actually very useful and effective when you use it for what it actually is: an electronic brake. It saves your brake pads (very useful for heavy, fast ebikes), provides a third brake option for extra braking power and safety in addition to your front and rear brake discs, and when properly tuned it can act as an anti-stall brake system in slippery conditions like a snowy road. And additionally you get the free benefit of some extra mileage, 2-10% in my applications. So that’s a win-win-win in my view; I use it on all my bikes!

The same goes for flux weakening: it is also far too simple to say that it is a bad design choice to use it. I have both a 3T mxus running at 72v 150 battery amps and 300 phase amps, as well as a 4T mxus running at 72v 120 battery amps and 200 phase amps. Both have equal acceleration (200 Nm; just keeping the front wheel on the ground when leaning very far forward over the handle bar) up to some 40km/h. Then the 3T becomes faster. I mostly use my bikes up to 50km/h in hilly forests but also want them to go at least 80 km/h on flat roads occasionally. The controller, battery, cables and motor of the 4T stay cooler at low speeds (up to 50km/h), but it cannot reach 80km/h. The 3T is faster, but also less efficient at high amps (slightly less copper fill, more losses + voltage drop in the battery, cables and controller). So maybe in this use case the better design choice would be to use the slower 4T motor + flux weakening for high torque and high efficiency at low speeds in the forest, while still being able to run 80 km/h on flat roads where the additional amps that both the battery and controller can provide can be used for a nice speed boost. To me, this is also a win-win and it is not a no-brainer to immediately think the 3t motor is the better design choice, although all parts of the bike are able to provide the required amps.

So in summary, it is not a black-and-white story, both regen and flux weakening can be very valid design choices. With foc controllers, we just have the luxury that we can choose our optimal combination of motor + controller + regen + flux weakening for our specific application!
[/quote]

Both motors will be capable of the exact same torque per battery amp...although the 3T motor will need more phase amps to match the torque output of the 4T. At higher speeds, the 4T motor will require flux weakening to match the 3T speed. For a given motor design, in almost all cases it is better to select the faster wound motor i.e. the 3T in this case.
 
hias9 said:
How many amps of flux/field weakening would you consider safe without risking to destabilize the magnets?
Motor is a MXUS 3k v3 3T
This is a really good (and hard) question. The answer is hard because the magnetic properties are actually not fixed for a magnetic material but are affected by the magnetic circuit which makes it impossible to state a general limit. In my experience demagnetisation rarely happens on neodymium magnet motors and if it happens the motor has been heated too much. This is based purely on my own motor cooking and the user posts here.

In theory it's like this in open circuit (magnet is in free air, not in a motor)
Since demagnetisation is dependent on field strength and temperature it’s required that you use motor temp limiting in your controller at less than 80deg C since above 80 the normal neodymium magnets lose some of their strength anyway. Then one can do estimations from demagnetisation curves for the magnets. The size of the magnet is needed, then it can be solved by using the BH graphs on k&j magnetics site.

My guesstimate is that if you have temp limiting set at 80deg C then you won't need to worry about demagnetisation from the coil fields. Use temp limiting controllers and motor temp sensor :wink:
 
On IPM rotors, field weakening RPM boosting even to over 2x base RPM speed can still be massively efficient.

I keep increasing field weakening until the motor or controller has control faults, then back it down a whisker.
 
Back
Top