Timing advance. Sensor-less vs sensored

[Electric Eel]

I just did a bit of a forum search on this subject, but came up empty handed.
Searching around for information on timing advance, I found this article:

http://electronics.stackexchange.com/questions/53637/why-have-non-zero-timing-on-a-bldc

This got me wondering. I have read here, that using sensorless rotor position reading has advantages because it then gets commutated ' when it wants to be.'
Looking at the picture from the link, I was trying to imagine if the back emf signal from the non energised coil would get the same degree of magnetic field distortion as around the energised coils... But I can see why running sensorless using back emf may be better than sensored at higher rpms.

Does electronically advancing the timing as rpms increase if you are using position sensors might achieve the same thing ?

 

[kenkad]

Electric Eel,
Your comment about the 'non-energized coil' is misleading. When you have block, trap, or 120 degree sine wave modulation, then you do have only two coiis ever energized. But, when you have 180 degree modulation sine wave drive, there is only only one point when only two coils are energized, the waveform zero crossing points, where the other two coil drives are balanced (or supposed to be). I cited a EE Mag design article earlier that discussed this.

FOC is timing/phase advance as a function of RPM and has been stated as such many times in threads. In addition, you need to deal with waveform distortion, a coil/driver electronics inductance issue, an influence that varies as a function of RPM also. I cited a Korean Academic paper on the forum regarding that (accounting for waveform distortion). I consider everything in the chain (Halls, Bemf, drive waveform) has some type of delay/distortion associated with it as a function of RPM.

PunxOr also started a thread today titled 'Controller Operation: Trapezoidal, Sinusoidal, FOC'. The paper he cites is OK, but, makes the whole process sound very complicated and it not. All designers, both motor and controller, make choices they justify to themselves. This only really works very well if the designer was responsible for both the motor and the controller and thus blended their individual characteristics. Of course, I want to state these are my opinions.
kenkad

 

[Electric Eel]

Kenkad,
My question relates to 6step (or some modification of) commutation methods.
Hmm, I just found out about FOC yesterday...

What I was really wanting to know was in sensorless operation, does the initial timing advance get automagically optimised (albeit with a degree of approximation ) by surfing the wave of field distortion. 8)
Like you are saying, the rotor position sensing is getting distorted, but in this case it works in your favour perhaps?
Could you post a link to the Korean paper discussing this?

The Eel.

 

[kenkad]

Electric Eel,
I would suggest more reading. All BLDC waveform drives are 6 step, period. 120 degree modulation means that only two phases are ever driven simultaneously. 180 degree modulation means that all three phases are driven all the time (except as I said earlier, at the zero crossing points).

In sensorless operation (which I would never do under any circumstances), it is easy to check the positioning (rotor/stator) when the motor is stalled (0 RPM). The common comment is that at higher RPMs, the absolute knowledge about positioning is less critical. In other words, any stuttering is overcome by the rotating mass so it is not 'felt'. Read Lebowski's threads and you will see how he approaches sensorless operation.

Please understand that although I have commented on your thread, my comments and opinions are my own. Others on this forum may take different positions. That is OK because it says that you can choose your approach and live with the consequences. Every day you buy a product, you are making such a decision. That is why there are many kinds of differing BLDC motor and controller products.