how to find out the correct FWEAK without knowing the parameters of the ACIM motor?

[Romale]

actually, the essence of the question is in the title of the topic. I took an industrial acim motor and changed the winding to a much smaller number of turns with an increase in the phase cross section for high currents.
But I absolutely can't figure out how to calculate the FWEAK parameter correctly, previously the motor was designed for 380 volts and 50 hertz (4 pairs of poles) but now I want at least 400 hertz from it with a constant battery voltage from 100 to 135 volts first use / 320 - 404 volts second use.
how, knowing only these two components, to find out the FWEAK parameter? and you need to first set up the boost by accelerating into the hill, and then fweak, or vice versa, first prescribe it, and then go to the hill setting up the boost?

thanks in advance for the competent explanations of this situation.

[Romale]

it is also not clear whether the fweak should be greater than the fweakstrt (is this the fweak start?)

and the last. fconst, what is it and how to understand and configure it correctly?

[catphish]

In therory, fweak is the frequency at which the motor is rated to run at your full battery voltage.

Here are some tings to consider:
1) Rated parameters of motors are more like suggestions than exact values, so don't worry about getting it exact.
2) That a lower value of fweak will give you more current and more torque though the motor.
3) Multiply AC voltage by sqrt(2) to get DC battery voltage.

So for example, if your motor is rated for 380 volts (AC) and 50Hz, then it's rated for 50Hz at 537V DC. If your battery voltage is 120V then that's 22% of the rated voltage, so fweak will be 22% of 50Hz, =11Hz.

I'd say 100V DC is much too low to run a 380V induction motor. I run a 230V motor at 120V for testing, and it only gets up to its 50Hz rated speed.

As for boost, there's really no way to calculate or guess this. Just start it at zero (where the motor likely won't move at all) and increase it until you get about the same torque from a standstill as you do at higher speed.

Edit: I completely missed the part of your post where you said you were rewinding the motor! In that case you're basically going to have to guess and adjust. Ideally you should measure the current in the motor and adjust fweak so that the current matches the rated current. And then adjust boost so that you get that rated current from a standstill. This is how I tuned my motor, by graphics AC current and adjusting fweak as needed to match the rated current. You can probably start by calculating the original fweak, and then scaling it by your change in motor windings. Ie half as many windings = double the fweak (and half the boost).

[Romale]

Edit: I completely missed the part of your post where you said you were rewinding the motor! In that case you're basically going to have to guess and adjust. Ideally you should measure the current in the motor and adjust fweak so that the current matches the rated current. And then adjust boost so that you get that rated current from a standstill. This is how I tuned my motor, by graphics AC current and adjusting fweak as needed to match the rated current. You can probably start by calculating the original fweak, and then scaling it by your change in motor windings. Ie half as many windings = double the fweak.

Thanks, I'll try it that way.
it's just that when I tried to start the motor from 70 volts for the test, it didn't start even with a boost of 3000+ until I lowered the fweak to 50.
Probably it is really now to find out only by experience.
but how do I find out which motor current is now rated?
and at what frequency to watch the measurements?
after all, it was previously made for 50 hertz, but now I intend to use an average of about 200 hertz and a maximum of 400 - 500 hertz.

[Pete9008]

If you want to maintain similar performance you need to keep the same flux in the motor. I seem to remember flux increases with N squared (where N is the number of turns) so if you have reduced the number of turns by a factor of 3 the current in the windings needs to increase by a factor of 9 (worth checking the N2 bit, it could just be N but seem to remember it being N2).

Edit - and I think the back emf goes up by N so for the above example the motor would generate a third of the back emf or need to spin 3 times faster to make the same power.

Edit2 - there is something not right there, the power doesn't work out right. Will leave it here to promote discussion though!

Edit3 - think flux must vary with N not N2, power seems to work out ok then. Can't find the reference equation to definitively confirm each way. If it is N then the current goes up by N, the voltage goes down by N and the power stays the same (for the same rotation speed).

[Romale]

If it is N then the current goes up by N, the voltage goes down by N and the power stays the same (for the same rotation speed).

this is true for motors with magnets on the rotor. but there is some misunderstanding with the acim motor. I have reduced the number of turns by 13.5 times!!! and I actually want to use the same voltage as before. only the ampere can be much more, since the phase cross-section has also increased by 13.5 times, besides, the frequency of use will be higher than the calculated by the manufacturer.

[Pete9008]

this is true for motors with magnets on the rotor. but there is some misunderstanding with the acim motor. I have reduced the number of turns by 13.5 times!!! and I actually want to use the same voltage as before. only the ampere can be much more, since the phase cross-section has also increased by 13.5 times, besides, the frequency of use will be higher than the calculated by the manufacturer.

I think it's the same for ACIM as the induced rotor current is driven by the flux. If you keep the same flux I think you get the same induced current and so the same rotor field. Although if you start running at a different slip angle/higher speed that all changes.

My best guess at the moment is that voltage will drop by 13.5 and your current will need to increase by the same amount. But I could well be wrong! Hoping an induction motor expert will turn up as I'd like to know the answer too now!

[johu]

It's inductance that has the square L ~ N² (https://en.wikipedia.org/wiki/Inductanc ... a_solenoid)

Half the turns, quarter the inductance, and, I'd think quarter the rated voltage and 4 times the current at the same frequency

[Romale]

It's inductance that has the square L ~ N² (https://en.wikipedia.org/wiki/Inductanc ... a_solenoid)

Half the turns, quarter the inductance, and, I'd think quarter the rated voltage and 4 times the current at the same frequency

unfortunately, it didn't help the stupid me to get a simple way to calculate the fweak .
I will connect a 220 volt battery soon. and just experimentally, I will first put a minimum boost and a large fweak, then I will start reducing it until the motor starts starting from pressing the pedal (I noticed at 70 volts that the motor does not start at fweak values above 65), and then I will start adding boost.

[Pete9008]

It's inductance that has the square L ~ N² (https://en.wikipedia.org/wiki/Inductanc ... a_solenoid)

Half the turns, quarter the inductance, and, I'd think quarter the rated voltage and 4 times the current at the same frequency

The equations for both motors and inductors turn out to be very similar. You get a factor of N on the way in (amps to flux) and then another factor of N on the way out (rate of change of field to back emf) so the total ratio is N squared. For motors this means that if you halve N then the current increases by a factor of 2 and the voltage drops by a factor of 2. It's the total ratio (V/A) that changes by a factor of 4.

Fweak is an odd one because it depends on multiple factors. An induction motor can spin at almost any speed you want (well up to the point where the rotor explodes anyway), with no currents in the stator there is no current induced in the rotor so no back emf. As you increase load the current in the stator has to increase to provide the power. The current in the rotor then increases and so the back emf increases which limits the speed it can run at (with that load and at that voltage). I believe fweak is the speed the motor can run with the rated load at rated voltage. If you reduce the load then you can run it at higher speeds, if you increase the load you are limited to lower frequencies.

So if N is reduced by a factor of 13.5 and if you keep the load the same then the current will go up by 13.5, the voltage will come down by 13.5 and fweak will be 13.5 x 50 = 675Hz. You can then chose to change the load to change the fweak. If you double the load I think fweak will need to halve (338Hz) and current will have to double too. If you halve the load fweak can double but current will halve.

Edit - corrected equation: 13.5 x 50 = 675Hz not 13.5 x 60 = 675Hz!

[Romale]

hello everyone. today I successfully burned the controller
the last time the motor was running at 70 volts already in the car, but now, when I assembled the cooling system and connected 240 volts, it refused to work. or he always just squeaked softly when pressing the pedal and did not spin, or immediately went into overdrive over current. I tried boost from 0 to 5000, I tried fweak from 0 to 700... nothing helped to make the motor turn. then I set the settings from the time when the motor was running from 70 volts, but I lowered the fweakstrt to 0, it was worth turning on the ignition and pressing start, as I immediately got an exploded controller

[catphish]

I lowered the fweakstrt to 0

a lower value of fweak will give you more current

fweak and fweakstrt control the speed at which 100% voltage is applied. Set one of them to zero and you will get 100% voltage at zero speed. That's certainly a quick way to destroy your controller, sorry

It's not clear to me why it didn't work with the increased voltage. In theory by simply decreasing boost to about 1/4 and increasing fweak by 4x it should work exactly the same. Perhaps the inverter was already damaged. I've certainly seen a Tesla inverter exhibit exactly the same symptoms, worked at low voltage then seemed to just die when 360V was used. Replacement inverter worked perfectly.

[Romale]

what I'm sure of is the inverter.
I tested it on a gen 2 leaf motor from a voltage of 360 volts with a firmware of 5.24 foc.

the reason is clearly somewhere else. maybe I did not correctly switch the windings to the middle point?
I will make a new controller and check the ends and beginnings of the phase pins. I think I'll try to run it in a single-phase network from an outlet through a powerful transformer lowering to 12 volts. if everything is correct, then in any three connections the motor will have to rotate

[Romale]

Half the turns, quarter the inductance, and, I'd think quarter the rated voltage and 4 times the current at the same frequency

Greetings. since then it's not just a discussion of the fweak, but also of the motor itself with the controller, maybe it's worth moving all the messages from here to my main topic ( viewtopic.php?t=2710&start=25 ) , and closing this one?
it will be difficult to sort out what to write here and what to write there.

[Pete9008]

Sorry to hear that Would it be worth measuring the resistance and inductance of all three windings just to check that they are all in the right ballpark?

[Romale]

Sorry to hear that Would it be worth measuring the resistance and inductance of all three windings just to check that they are all in the right ballpark?

I personally performed the winding myself. for all three phases, I created identical bundles of wire from parallels 10 meters long each. each phase is one hundred percent the same with the neighboring one in terms of resistance and inductance. The only thing that confuses me is the possibility of not correctly understanding what is considered the beginning of the winding and what is the end, and I could mistakenly connect incorrectly to the middle point.

[Pete9008]

Sorry but I'm not questioning your ability here just suggesting a way of checking to see if the numbers make sense without having to pull it all apart. It would also be interesting to know what kind of inductance you have ended up with (too small and the inverter may struggle to drive it).

[Romale]

this is the scheme I used.
but what if the middle point had to be entered - blue phase located between 2 and 3 slots, and not + between 8 and 9?

[Romale]

Sorry but I'm not questioning your ability here just suggesting a way of checking to see if the numbers make sense without having to pull it all apart. It would also be interesting to know what kind of inductance you have ended up with (too small and the inverter may struggle to drive it).

unfortunately, I have nothing with which to measure the inductance, but it is not very different from a leaf gen 2 or Tesla ldu motor. probably even closer to the prius motor

[Romale]

in any case, I have already disassembled the controller, removed the power module and will make a replacement in the evening (it's good that I had a spare one )

[Pete9008]

Afraid winding motors is a complete mystery to me. Did find this while googling, no idea if it is of any use but it does seem to show the start point arrangement? https://www.jmag-international.com/e ... ary/013/

Good luck with mk2!

[Romale]

heck! judging by this picture and many others like it, the third phase is really moving in a different direction!

[Romale]

I replaced the power part of the controller, but it didn't help. the processor stopped seeing the position of throt 1, something else burned down along the way. I did not look for the reason and replaced the control board with a new one. I checked at home on a small factory acim motor, and noticed this situation: the udcnom value is 0 defoult, with this value the motor is shaking and slowly rotating as if without a/b sensors, I set the value to 70 (like a used battery) and got the same thing. I remembered that the last time it worked well at a value of 50, prescribed this value and the motor started working as it should.
What kind of parameter is this, UDCNOM?
I thought it was the nominal value of the battery voltage.

[Romale]

now I have to go to the car, deploy the ends of phase C and test everything anew.

[mjc506]

udcnom is the voltage for which the other parameters are set - some parameters are automatically adjusted based on battery voltage.

This perhaps suggests you need to set udcnom to your actual pack voltage and re-tune the other parameters?