The decogging tutorial basically grew after after these ( http://www.fieldlines.com/story/2006/6/23/94016/5550  and http://www.fieldlines.com/story/2006/6/23/185558/084) discussions about 2 years ago where I tried to express my thoughts with only limited succes. I had a lot of trouble expressing myself correctly in a way that people understood so I decided to make the effort once and for all, write it down properly in a small document and be done with it. So, if you say you don't understand what I mean, I'm at my wits end because I don't think I could explain any better than it was written in there :)

But I'll give it a go...

Firstly, I'm not sure I know what you mean by 'mounting flush' so will ignore that part of your question.

As far as the imaginary magnet goes... Let's do a little thought experiment.

Zubbly's (simple) rule states that, on a row of magnets, the first magnet should sit under one stator tooth and the last magnet sit again under a stator tooth, but the next tooth.

Follow me ?

Now, let's imagine a rotor with just one pole. On that pole there are two magnets (on one skewed row). Magnet A sits exactly under one tooth (in the image below, imagine it sits under tooth 3), magnet B sits exactly under the neighbour tooth 2.

Do you see how this would cog terribly ?

So, apparently, the simple rule goes wrong, especially in cases where there are only very few magnets per row (2 or 3). The error becomes progressively smaller with more magnets per row and is hardly an issue in practice for those cases. However, if you use the 'simple' rule (simple not being a pejorative description, btw), you will go wrong in the case of only 2 or 3 magnets per row and it will cog.

In our previous example for it not to cog, one magnet would have to sit exactly under a stator tooth, while the next magnet should be exactly under the slot, i.e. half way -between- two stator teeth:

If you understand this part, you'll see how adding the extra 3rd imaginary magnet and siting that (like the real first magnet) under a stator tooth, will serve as an aid for laying out the rotor and determining the correct skew angle.

Let me know if this was clear or not.

Regards,