How does it work

[MetalMan]

Some brushless can change frequency through programming, like the Schulze U-Force. Generally, 8khz is good for 2-pole motors (Feigao, Lehner) and 16khz is for 4-pole motors (Plettenburg). If you are *stupid* then that's all you really need to know. (In fact, if you judge stupidity in that way, then I am stupid :) LOL). I think that the frequency has to do with how quickly the controller checks on the rotor's position.

[Serum]

The frequency changes the amount of pulses that are put in one motion. It are the better quality magnets that can benefit this higher frequency. my guess it has got something to do with eddy currents, because of the reaction speed of the magnet.

[GriffinRU]

Originally posted by coolhandcountry
I got a question. I like to know how an esc works. Does it regulate the amount of volts or the amount of amps?

It regulates output voltage by PWM (pulse width modulation) and this automatically regulates current as well.
There are modes when ESC can regulate current, to maintain constant RPM.

Brushed (Digital) or Brushless ESC works similar (brushed has one phase, brushless 3 phases).

Switching frequency selection based on motor inductance. Higher turn motor has higher inductance. Higher frequency allows getting more steps in regulating output voltage and makes motor to run smoother. But running at higher frequency taxes ESC efficiency, more times switching occur more power loss on FET. Details can be provided later.
To run ESC in the most efficient way you should select highest frequency for low turn motor and the lowest one for high turn.
For a example 7XL – 9 kHz and 8S – 19 kHz (or even 38 kHz) Schulze example.
Testing will tell you which one is the best but do not set 38 kHz for 7XL. When changing frequency only ESC temperature should be affected because frequency over 1-2kHz doesn't effect motor to much if at all. You can run low turn motor on low frequency but then it should be less smooth.
If you do not know what motor you have and what settings to use, start from low frequency and step up until your ESC start radiating significant amount of heat and then step back.

BEC is a simple voltage regulator. Most ESC produces 5-6V and can deliver 1-5A based on particular model. Key factor here is power dissipation. If it is not build based on DC-DC switcher then most TO-220, D2-PAK rated to 1-3W (without heatsink or extra cooling) and if BEC generated from 7.2V then you can easily draw 1A without overtaxing BEC regulator (7.2-5*1=2.2W). But if you start from 14.4V and 1A you do the math…
Most Voltage regulators have built-in over-temp protection which reduces output voltage to limit output power. Extra cooling helps but not much.
Ultimate BEC (with DC-DC converter) is the only one solution.

That is why if you increase BEC voltage output it is better from both sides, better for us (more power to servos) and better for regulating part less heat dissipation. But it is hard to maintain 6V (for BEC) if you run from 7.2V, not enough voltage for reliable regulation, not mentioning voltage drop upon battery use.

Artur

P.S.
Some links from my BLDC web collection

http://www.torcman.de/peterslrk/SPEE...l#Anker1591256

http://www.servomag.com/flash/2-pole...ldc-motor.html

[Serum]

And for the startup the esc works on a lower frequency. (sensorless)

At a certain point, the possition of the magnet can be determined by the speedo and at this point the controller steps up in frequency.

I think that is the reason that the elder type of controllers are not as smooth as the new ones. i think (I'm not sure about this, but it sounds very logical to me) that the elder controller just worked on a fixed switching frequency.

[GriffinRU]

Originally posted by Serum
And for the startup the esc works on a lower frequency. (sensorless)

At a certain point, the possition of the magnet can be determined by the speedo and at this point the controller steps up in frequency.

I think that is the reason that the elder type of controllers are not as smooth as the new ones. i think (I'm not sure about this, but it sounds very logical to me) that the elder controller just worked on a fixed switching frequency.

I am confused with your comment here...
What rotor position have to do with frequency? At start up to find out position ESC firing each phase in sequence (with reduced power by PWM), and acquiring EMF feedback to match later.

The goal here is to make as less switching as possible, but to run motor smooth you need to control voltage in very fine increments (higher freq., more switching)
And depends on your overall usage (most time 50% or 25% or 100% load) you can choose better frequency for particular motor and your needs.

Old ESC's had fixed low frequency due to the fact of available microchips and flexibility in programming(PIC's). Now you can get better (New gen PIC's and new FET's with better Rise and Fall time) components which can operate at rates capable to control motors in fine steps (512-2048) up to +150000 RPM with 2 poles. And yes you can adjust switching frequency on-fly but I haven't seen one yet with such capabilities.

Artur

[Serum]

from as far i know, they downgrade the sample frequency (lets say from 38 to 12 khz) with the startup of the rotor. once the rotor spins, the esc is capable of determining the position of the rotor.

(i've got this PCB layout here that is developed like this) it uses the ST7mc1 (not that that has got anything to do with this)

this controller uses a standard startup procedure for making the motor start. it start on a certain frequency, with a fixed voltage and current, and once the communication signal is capable of determining the possition of the rotor, the esc steps up to his normal 'sample' frequency.

it does reading/communication from the motor with a 1mhz signal to determine the possition of the rotor. (on the unpowered poles)

[GriffinRU]

Originally posted by Serum
from as far i know, they downgrade the sample frequency (lets say from 38 to 12 khz) with the startup of the rotor. once the rotor spins, the esc is capable of determining the position of the rotor.

(i've got this PCB layout here that is developed like this) it uses the ST7mc1 (not that that has got anything to do with this)

this controller uses a standard startup procedure for making the motor start. it start on a certain frequency, with a fixed voltage and current, and once the communication signal is capable of determining the possition of the rotor, the esc steps up to his normal 'sample' frequency.

it does reading/communication from the motor with a 1mhz signal to determine the possition of the rotor. (on the unpowered poles)

Let's start from scratch, tell me how motor and ESC works. It would be very clear where we are and what is what. It would be great if you cover relation between PWM and switching frequency as well.

Artur

[Serum]

the startup of the motor, like i said. When the esc doesn't know the position of the rotor.

[GriffinRU]

Originally posted by Serum
the startup of the motor, like i said. When the esc doesn't know the position of the rotor.

Well, I will make it simple brushed motor. What now?

Artur

P.S. Motors with low inductance, almost instantaneously response to PWM from ESC. The low inductance may cause excessive current ripple and cause lots of trouble (motor heating, ESC over voltage). This effect increases at lower PWM frequencies where the current may swing from pos to neg during each PWM transition. That is way switching freq. and motor inductance match important.
In real world to gain performance extra inductors are added in between motors and drivers. Not practical for RC.