Books
 |
Metal Detectors
 |
| Garrett GTI 2500 |
Prospecting Tools
 |
Outdoor Gear
 |
Radios
 |
GPS
 |
Power Inverters
 |
Fun Stuff
 |
Clocks
 |
It may seem a bit odd that I chose to write about this subject but I have spent a large portion of the last 20 years designing methods for producing and testing brushless dc motors. I have designed automated stator winding equipment both of the needle variety and the "feed-winding" variety made popular by Electrowind, Inc. during the last decade specifically to wind brushless DC motors.
A friend of mine Peter Thomas, builds custom stator-coating equipment from his company Creative Method and he and I have discussed many times over the years what makes a "producible stator" and we felt that some discussion might shed some light on the subject. We chose now to do so because of many factors but one of them is the emerging emphasis on brushless motors for robotics applications. Pete has also designed and built several automated stator winders and other automated motor production equipment. So hang in there and follow the discussion to see what our view is.
Most motor designers (magnetics people) always seem to want to design the most efficient motor ever done. Unfortunately, many of these designs can be nearly impossible to produce economically. This has led to all kinds of schemes including cutting the stacks in half and then inserting coils just to weld the stack back together again to make it back into a round stack. This is not particularly good for the magnetics plus it introduces another painful production operation. Many people have chosen to hand-wind the stack which ultimately has to be farmed out to a low-wage country such as China. This leads to all kinds of headaches including delays, loss of control, and acceptance of varying levels of quality. Hand-insertion of coils requires that the coils be much larger than the stack and this naturally increases the resistance of the winding which is not usually desired. Plus all that extra wire needs to be dealt with requiring coil forming or lacing. I have heard of rejection percentages of 50% or more when stators are hand-wound. This is easy to understand if you have ever tried to wind a complicated stack with many turns...I get lost all the time doing this.
This brings us to the subject of designing for automatic production. Take a look at the stack below. The slots are fairly narrow and there are 6 teeth-slot combinations to wind.
Depending on the type of winding you are doing
getting the wire through this slot can be easy(automated) or difficult(manually). The key is "slot-fill". Most motor designers want to maximize the slot-fill of their designs.
This causes a range of production problems. I was once asked by a motor designer if his stack could be wound automatically. I took one look at it and basically told
him no. The problem, I could not see light through the windings. This was undoubtedly a very efficient design but very difficult to produce. The wire had to be "pounded"
into the slot to get it all in there. It was costing them over $200 per stack to wind for a roughly 8" O.D. stack. This stack went into a machine
that was approximately 4´ x 6´ x 8´ in size so I asked if he couldn't design the stack a little larger, say a ½" or so. The
response was absolutely not! In mine and Pete's opinions, designs should take into consideration the production requirements anytime more than a couple of
pieces will be made.
The key in stator winding is slot-fill. Automated winding can be performed when the slot-fill is up to about 90%. After this the wire begins to "bump" into the coils as the wire is inserted (feed winding) or the needle will scrape the wire (usually with needle winding this will happen long before a 90% fill is accomplished). When preparing the stack for automated winding we have found that powder coating produces the most reliable insulation coating. You can use paper(nomex) or plastic-molded inserts but usually these present their own problems including having the wire hangup while winding.
Automated feed winding is one of the best ways of winding a brushless dc motor stack when narrow slot-widths are desired. Feed winding also produces very small end-turn sizes which contribute to a lower over-all motor resistance. Another advantage is the ability to control the cross-over lead length when winding more than one set of coils per stack. This alone reduces lead connect time in production and lowers the overall cost significantly for this operation. When properly wound on automated equipment no or virtually no forming or lacing is required for the finished stator.
More to follow but if you want to contact me about this article, please use the Contact menu found on the Train Logo above or use this link to Creative Method. Thanks for visiting.
