> ... I have been pondering going direct drive with a slow speed homebrew BLDC motor at some point in the future.
I'll chip in as someone who started making homebrew BLDC motors in 2008, now having molds for the PP-epoxy composite body parts, and drill templates, to duplicate them, and can make a kit available...
I am thinking large diameter with at least 8 maybe more poles, phases Y-connected. Radial flux so warping of the rotor plate from heat and torque won't be as likely to scrape the magnets on the coils, and so I won't have to be so concerned with small amounts of axial play.
With axial flux the gap between the magnets and the stator is generally on the order of 1/2" instead of hundredths, which means having them hit each other owing to minor misalignment is unlikely, plus the magnets (in my designs) have polypropylene strapping epoxied onto them affording good surface protection and a thin composite wall between the rotor and stator compartments protects the coils, plus the rotor can be inserted and removed simply by hand pressure (not forcelessly, but at least up to my 12" O.D. size), and also the rotor supermagnets don't gradually weaken from close-up localized effects as (I understand) is the case with radial.
Water cooled so a slow turning fan won't allow overheating. Yes I know this will be an expensive project and I will end up rewinding the stator multiple times but it is mostly about pushing the envelop and seeing if it can be done practically, using an off the shelf controller like the 300a 48v Kelly unit.
Such a BLDC controller should easily do it for either of my 'pancake' BLDC motors, the Electric Caik (~9"x4", 6 coils 4 poles, 3+ KW?) and the Electric Hubcap (~12"x4", 9 coils 6 poles, 4.5+KW?, both about 0-2000 RPM). I estimate these are up around 95% peak efficency (but still have done no definitive tests), and so they generate little heat inside. The magnets themselves act as the cooling fan blades and the air coming out the holes can be felt even at low RPM s. In general for driving a prop, the motor shouldn't 'stall' or be loaded to too low a speed for the air cooling to be effective.
One question is whether I will get power loss from induced current in the cooling lines if I use copper or stainless tubing. I figure on running the coolant in and back out of each individual coil core. Will induced currents simply cancel each other out?
Currents induced by the rotor magnets do cause heat and inefficiency. That's why I switched from steel stator coil backing plates to the non-metallic PP-epoxy composite bodies. The coils are very low loss iron powder toroids. (and the coils are wound around them as a 'donut'.) The main remnant metal losses are now from the coil holding machine screws, which I shrank from 1/4" to #10-24.
Also it seems that with lots of coils I could tailor the impedance a bit by wiring each phase in series, parallel, or series/parallel. Does this sound right? The larger diameter (maybe 22" or so) should give me lots of low speed torque, correct? And more poles reduce cogging? Going way out in left field how about say 30 poles, 90 coils? Could go series, or 15 parallel pairs in series, or 10 series of 3 coils, or 5 of 6, etc.
Torque for the Electric Hubcap is maybe about 1.5 foot-pounds per 10 amps of current from the batteries (@ 36 volts), and maybe 1 f-p per 10 amps (?) for the Caik at 24v. But several characteristics also depend strongly on the flux gap, which can reasonably be adjusted from under 1/2" to near 3/4" or say 11 to 17mm.
I have a design for a 26" rotor, 12-15+KW unit, the Electric Weel (24 coils 16 poles, all around the outside edge), that might (my ideal) have enough torque to move a car, tied straight to the wheels with no gear reduction. Alas although some work has been done it's still an unbuilt concept and I haven't had time to work on it for some time.
My coils are 3.5 turns/volt, but obviously that depends on the core and size. I usually wind them with 21 turns of #11 wire - 6 volts. In essence, 36 volts for the Hubcap is with 3 coils per phase wired "Y" - 6 coils. I've also used 63 turns of #14 with the 3 coils per phase in parallel for 18/36 volts. Obviously if I put the coils in series they'd be 54/108 volts at 1/3 the current.
With heavy gage connecting wire to reduce losses, might this work, so that the same basic motor could be tailored for different operational needs by rewiring? Another reason for lots of poles is so I could use cheap off the shelf magnets and not have to use custom mags. ...
I'm not sure - ceramic (or ??) magnets might work, probably with smaller flux gaps, but I just buy NIB supermagnets, eg, magnet4less.com.
Craig
http://www.TurquoiseEnergy.com/
Victoria BC Canada
__._,_.___
| Reply via web post | • | Reply to sender | • | Reply to group | • | Start a New Topic | • | Messages in this topic (3) |
Just launched ! Link preview on Yahoo Groups
Visit your Group on the web, simply paste the link to the article, photo or video you wish to share in the message you are composing.
.
__,_._,___
No comments:
Post a Comment