Rotating Magnetic Field's and Inductors.

[TinselKoala]

Here's what the induced voltage looks like when a rotor magnet passes a coil.

As the pole of the magnet approaches the coil, it induces a voltage whose polarity depends on the polarity of the magnet facing the coil. As the magnet gets closer and closer the flux changes faster and faster so the amplitude of the voltage increases. As the magnet passes "TDC" or closest approach, the polarity flips (because now the flux is decreasing instead of increasing) and the amplitude starts high (fast change in flux) and decreases as the magnet moves further away. 
If the rotor magnet is flipped so that the other pole is facing the coil, the induced voltage pattern is flipped: first positive, then negative.

[MileHigh]

Webby:

TK posted a "down-up" double pulse.  So let's say that was for a north-out rotor magnet.  The zero-cross of the double-pulse is the exact TDC of the rotor magnet.   So when the south-out rotor magnet passes the sensor coil you would get an "up-down" pulse.  That gives you all the information you need to know where the rotor angle is after you correlate it with the EMF induced in the drive coil.

The interesting thing here is that the coil + rotor is slaving to an external signal generator pulse train.  That's in contrast to what we normally see were the rotor magnets themselves or optical markings on the rotor generate the timing.  So knowing the precise sync mechanism would be interesting because it's a thing that you don't see every day.

The real way to do it would be to use your DSO and then roll up your shirtsleeves, get some graph paper and a pencil, and construct a real timing diagram, or do the computer equivalent with some kind of graphical program and load in DSO captures, etc.

MileHigh

[MileHigh]

There is no reason you can't put the actual torque waveform on your timing diagram.  Once you have the basic timing understood then you can do anything you want.  For example, you can hold the static rotor at a certain angle, energize the coil, feel the torque with your hand, and then put that rough sample datum point on the timing diagram.  Then advance the rotor by 10 degrees and repeat the process and record another rough datum point and so on.  You could easily make a basic torque waveform on your timing diagram using just your wits and your hands as "torque sensors."  That would give you a very good idea of when the rotor was undergoing a gentle acceleration.  Apply yourself and keep on analyzing the setup thinking about how to tackle various problems and you should also be able to make a very decent estimate for the deceleration of the rotor.

There is almost no limit to this stuff.  Once you understand the basic timing, just the EMF waveform in the drive coil is telling you which (presumably) virtual pole is approaching the drive coil.  As a bonus, now you know exactly what angle the rotor is at when you look at the EMF waveform.  You also know exactly what the current waveform looks like and the timing for the energizing of the drive coil.  So you don't even have to measure the torque like I state above.  You have all the information on the timing diagram to deduce what the torque is just like that.  Therefore you can also do a decent job of deducing the acceleration and deceleration of the rotor.  Then, if you are thorough, you make sure that your deduced torque is the same as your measured torque as stated in the above paragraph.  If they are in accord great, if not, then you still have to do more thinking and investigating.

Note how this can be done without printing out a fine-pitched black and white strip of encoder bars to glue to the rotor and then using an optical encoder and presumably some turn-key software to spit out the acceleration and deceleration for you.  It's arguably more fun to "build it yourself."

After six years on the forums, I don't expect to see this level of measurement and analysis for the operation of a pulse motor.  Even the "pros" never dare enter this territory.  And I am just making it up as I go along.

[tinman]

I was just about to address this issue, so I'm glad you brought it up.

Consider a rotor with magnets, turning inside a ring of cored stator coils. First, just spin the rotor but don't provide any power to the coils. Consider what happens to the inductances of the coils, and the induced voltages in the coils. The inductance goes down as a magnet approaches the coil and goes back up as the magnet recedes from the coil. This effect does not depend on polarity of the magnet. (Faraday's Law in action again.) This effect _does_ depend on the polarity of the magnet facing the coil. That is, whether the voltage goes from + to -, or from - to + as the magnet passes, depends on which pole of the magnet is facing the coil as it moves past.

So you have a very complicated situation. In some configurations the induced voltage from the magnet's motion will aid the voltage applied to pulse the coil, and in other configurations it will oppose it. So depending on which pole of the magnet is facing the coil, and what polarity of voltage is applied to the coil to pulse it, you can have enhanced "pulling in" or reduced "pulling in" in an attraction-type pulse motor, and enhanced or reduced "pushing out" in a repulsion-type PM. If you have alternating rotor magnet polarities it gets really _really_ complicated. Then there is the effect of changing inductance/permeability from the approaching and receding magnets. These effects can aid or reduce the effects due to induced voltage!

The most interesting type of pulse motor that I know about is the Steorn Orbo core-effect motor. As you know, a carefully wound toroidal coil on a high-permeability core will not have much leakage of magnetic field when it is energized. So you wouldn't think it would be useful for attraction or repulsion type pulse motors, and it's not. BUT.... the magnetic field caused by energizing the coil does change the core's permeability... and this affects how strongly a magnet is attracted to the _core_. It isn't attracted to the magnetic field from energizing the core, but to the core material itself. So the Steorn Orbo core-effect motor works by having the rotor magnets attracted to the toroidal cores while the power is _off_ during the approach, and just at dead-center the power is turned _on_ and this _reduces_ the permeability of the core material, making it less attractive to the rotor magnet. So the magnet is pulled in more strongly as it approaches, than when it has passed and is receding. So the rotor speeds up. This core effect does not depend on polarity of either the voltage applied to the coil, or the polarity of the rotor magnet passing it! So you can have alternating rotor magnet faces, or same faces, to the toroidal coils and it will work the same.

But also, as the magnet approaches the coil it induces a voltage in the coil, that increases up to the nearest approach, then flips sign and decreases as the magnet goes away again.

That is incorrect.

[TinselKoala]

That is incorrect.

Argue with an oscilloscope, why don't you.

Explain this trace, which I'm sure you have seen many times in your own work.