Go Down

### Topic: [answered] 3-phase permanent magnet alternators (Read 1 time)previous topic - next topic

#### JimboZA

##### Oct 19, 2014, 05:51 amLast Edit: Oct 19, 2014, 05:09 pm by JimboZA Reason: 1
Just started fiddling with these, prompted by the fact that my son's motorbike has one for charging. I had just kind of assumed all alternators were like car ones (field coils), whereas making a PM version (perm magnets on rotor, windings on stator) for experimenting is within even my powers.

(Real reason: I'm still looking for stuff to stick my new 'scope probes into....)

Anyhoo, to my question, after much Googling and YouTubing. 3 phases often means 9 coils (3x3 in series, spread out round the clock) and it also seems common to go for 12 magnets. I'm wondering if the 12 magnets for 9 coils is deliberate / significant?

Look at the attached pic (and I apologise for quality: it's a screen shot of a paused YouTube vid). It seems (if you look at 10, 2 and 6 o'clock more-or-less) that 12 magnets for 9 coils gives you (they way they're spread out) a magnet between 2 coils and a magnet each entering and leaving those 2 coils.

So is that by design? Is that why 12 magnets is common with 9 coils? I suppose this arrangement gives each phase a fair shot, and also smooths out the AC into a nice sine-wave? Another video kind of implied that.

Johannesburg hams call me: ZS6JMB on Highveld rep 145.7875 (-600 & 88.5 tone)

#### nilton61

#1
##### Oct 19, 2014, 07:47 am
A electrical rotational machine (motor or generator) is characterized by its number of poles. That is magnetical poles in rotor for ac machines or magnetical poles in the stator for dc machines (A BLDC is actually a synchronous machine with the ac generated by the driver). Since maxwells laws state that there are no magnetical monopoles this is always a even number. The number of poles decide the ratio between number of electrical periods and angular advance. A  2 pole machine rotates a full revolution for each electrical period. A 4 pole machine rotates half a revolution for each period and so on.  This is independent of the number of phases used to drive the coils. So far for the simple theory.
In practice the layout of the coil-slots affects torque ripple, harmonics and zero sequence currents.
http://jimhendershot.com/Jim_Hendershot/Articles_files/brushlessdcmotorphasepoleslotconfigurations.pdf

#### MarkT

#2
##### Oct 19, 2014, 02:40 pm
The number of poles is even because they alternate N-S-N-S, nothing to do with monopoles!

Talk of pole-pairs if you want, then one electrical "revolution" per pole-pair.

There's a multiple of 3 coils in a 3-phase machine, and there is a phase coil (usually)
1 or 2 times per pole-pair.  There are many ways to wind a motor though, all that matters
is that the phases are 120 electrical degrees apart in the finished stator:
http://www.bavaria-direct.co.za/info/
[ I DO NOT respond to personal messages, I WILL delete them unread, use the forum please ]

#### JimboZA

#3
##### Oct 19, 2014, 03:37 pmLast Edit: Oct 19, 2014, 05:08 pm by JimboZA Reason: 1
What's "teeth" in that table, coils? Looks like it.

Also it's not clear there if blue's "good combination" is better or worse than white's "it works"  , but what's clear there is that for 9 coils (it that's what "teeth" are), the all 9 are wound the same direction with 12 magnets.I was going to do that in the absence of any other info.

Nice blog there, def worth a bookmark.

edit: I emailed the blogger to ask those 2 questions.)

more edit: And got answer. Yep teeth are coils and blue is better than white. So in fact for 9 coils, 10 magnets is better than 12 (reason unstated). BUT.... with 10 magnets the wiring is a bit more complex since the middle coil of each triplet is opposite hand; with 12, they're all the same.

That really answers my question so thanks to MarkT for the link to my compatriot's site.
Johannesburg hams call me: ZS6JMB on Highveld rep 145.7875 (-600 & 88.5 tone)