Field Weakening- can we measure effects with no-load info?

[John in CR]

I kinda understand what field weakening does to get more rpm, but I'm interested in predicting the results. Will no-load current and rpm info give us what we need, so we can plug the info into Miles' motor spreadsheet and get the torque, rpm, and heat losses?

 

[John in CR]

Since there's no responses, I take it that it was a stupid question and no-load rpm and current don't change with field weakening forced by the controller. If so, does all it do is force current at higher rpms that the motors BEMF wouldn't allow it to normally draw, and the result is simply getting the motor under load to achieve closer to no-load rpm at pack voltage?>

 

[Punx0r]

I'd normally stay out and let those more knowledgeable and experienced speak, but since you short on any kind of response I'll chip in: To my understanding, field weakening energises the coils out of time in order to counteract the magnetic field from the magnets. Lower field = high Kv & lower Kt which is a handy compromise for applications where you're limited by available motor speed, but not torque. The field weakening current is entirely wasted and contributes to motor heating. Justin used it to simulate load to heat up motors in his wind tunnel testing.

So I guess it depends what you're asking. If your motors natural no-load speed is 1000rpm and you use 50A of field weakening to raise this to 1500rpm, then power at both the old and new no-load speed will be much greater.

I stand to be corrected on any of the above

 

[John in CR]

That was kinda along the lines I was thinking when I made the first post, ie can we use the spreadsheet to predict results of field weakening based on a new higher no-load rpm. I don't get how sending out of time pulses could work though, so now I'm thinking it just forces extra current through at the normal timing and no-load rpm doesn't change. That way there's more torque at the high rpms so we get closer to no-load rpm while pushing a load. That also seems like it coincides with the "about 20%" potential increase in top speed using it.

If the latter is correct, then the spreadsheet already gives us the predictive results, since we can enter any torque number and rpm we want, which leads to a part we're missing in the spreadsheet. That is the normal tapering of torque at rpms above peak power. Unlike the young guys who want pictures and graphs, I prefer numbers.

 

[Lebowski]

Field weakening gives a 'virtual' increase of the supply by the amount

V_extra[V] = 4*3.14 * f[Hz] * L[H] * I_fieldweak[A]

With
f[Hz] = erpm/60
If I is in rms then V will be rms.
L[H] the inductance for a single winding (phase) of a WYE wound motor.

In my controller IC fieldweakening is applied automatically upto a max I_fieldweak (user settable)
I recoomend I_fieldweak to be 70% of max motorcurrent.

The extra supply results in extra speed based on the normal Kv

 

[Lebowski]

https://endless-sphere.com/forums/viewtopic.php?f=30&t=72692&p=1099890&hilit=fieldweakening#p1098494

 

[Punx0r]

John, there's a reasonable explanation on this page (I just went searching for the vector diagrams I've seen before): http://scolton.blogspot.co.uk/2009/11/everything-you-ever-wanted-to-know.html

It's a bit complicated and really helps if you're familiar with the basics of AC electrical theory i.e. how voltage and current can be out of phase due to an inductor (or capacitor). The system appears to work by altering electrical timing (equivalent to altering the physical position of the hall sensors) to shift the relative phase of the voltage and current to increase BEMF and so motor speed.

The line near the bottom about the amount of speed increase you can get being dependent on motor inductance will probably ring a bell and ties the theory together: Inductance causes phase shift in AC voltage/current, the shift can be worked to our advantage to increase motor speed.

 

[John in CR]

Ok, I get probably only enough of it to get me in trouble. I do understand some about phase shift of inductors from work on speaker crossovers in my previous hobby of speaker building. This is all related to FOC controllers, which I'm still to cheap to own or to short on knowledge to build, but what do more common trapezoid controllers due to force the overspeed? The factory I'm dealing with now calls what they do in excess of 100% speed setting "field weakening", so is it the same thing but just estimated like they do for phase current?

 

[Punx0r]

Yep, that's about where I am: knowing just enough to know it works, but not really how...

Good question on the trap' controllers. One of Jeremy Harris' old posts suggest the 120% speed setting on the "Infineon" controllers did nothing clever, just increased the restricted PWM duty cycle upto a maximum of around ~95% duty. It sounds like the controllers are unstable at or near 100% duty so they're deliberately limited under standard settings.

https://endless-sphere.com/forums/viewtopic.php?f=2&t=12249&p=190254&hilit=110%25+speed+waveform#p190223

Whether the controllers you're looking at do similar, or in fact alter the timing to gain speed I would love to know. I wouldn't be surprised if advancing the timing gave a useful speed increase - it does on brushed motors and IIRC there's a false-positive hall combination on BLDC motors that gives very high speed (and current draw) thanks to grossly advanced timing.