By Timothy Raneyâ€¦Bald Engineer Guy with Glasses
Last week, we discussed recovering from mistakes and assembled the mechanical parts for a â€œForm, Fit and Function Verification Testâ€. Now weâ€™ll continue this marvelous series and see some progress on the coil winder and begin discussing the electromagnetâ€™s electrical aspects!
Initial Coil Winder Design
In the initial design of the coil winder, I had to check the two motor candidates on-hand. One was a vintage shaded pole induction gear motor. At first, the motor would not run at all. However, once disassembled, I cleaned the rotor bearing and oiled the gears. Afterwards, it ran nicely at 2rpm. Too slow for a coil winder if I wanted to complete the electromagnet coils during my lifetime. Now cleaned and labeled as a 2rpm motor, it is resting comfortably in a bin with the other motors. Next I tried a Pittman DC gear motor last tested in 1998. I really should have a better motor collection. Anyway, it failed the â€œsmoke testâ€ initially. With it pulling over five amperes at ~11 volts and smoking like a chimney, I quickly decided it prudent toÂ cut the power to the motor. To make a long story short, I took the motor apart, cleaned the commutator with fine steel wool and reassembled it. Keep your fingers crossed. Yes! Tested again, the motor operated at 19VDC and only drew ~0.2A under a simulated load - holding the shaft tightly with my fingers â€“ very scientific. No, I donâ€™t have a dynamometer. The motor shaft rotated at 80rpm â€“ much better. Yes, it ran very nicely and it gave up smoking too. Interestingly, many old motors still work well â€“ they just need a little maintenance.
Letâ€™s "Improve" an Old Coil Winder
What an idea. Instead of building a new coil winder, I decided to â€œimproveâ€ an old coil winder I built some years ago. In winding the coils, I estimated each coil could hold up to 5133 turns of #24 magnet wire (0.020â€ actual diameter) based on a 1.1875â€ wide x 1.75â€ deep coil slot, i.e., 59 turns fit in a 1.1875â€ wide slot multiplied by 87 turns high. Using a wire table and a 50% de-rating factor for bundled conductors, the maximum current capacity was estimated at 3A. However, the maximum practical current is based on the supply potential and the coil resistance per Ohmâ€™s law. It is also important to remember the 5133 wire turns is an ideal result given the coil slot dimensions. In practice, the magnetâ€™s design is a series of compromises. We want might to maximize the current, but the correspondingly larger diameter conductor will occupy more space. This result reduces the number of wire turns for the given slot dimensions. Moreover, we might then need to add a liquid or forced air cooling system since the higher current increases the I2R loses, where I is in amperes (A) and R is the resistance in ohms (W), with the product (power) in watts (W).
Conversely, we can maximize the number of wire turns by selecting a wire size that can carry enough current to produce the magnetic flux we need â€“ without needing forced air or liquid cooling. We want the maximum number of turns since the subsequent magnetic flux is proportional to the turns multiplied by the current. The product, known as â€œampere-turnsâ€ is the starting point for determining the electromagnetâ€™s flux (in gauss or tesla) later. Therefore, if we can add enough turns, the magnet will produce the desired flux at lower currents.
Preliminary Coil Winder Design
I worked on the preliminary design for the coil winder. This work included disassembling the old coil winder, selecting parts for the new version and beginning the actual fabrication. The preliminary design for the coil winder is shown below. I did not have a gear motor with shaft speed in the 120rpm range. As I mentioned previously, 120rpm worked well on the old lathe for other coil winding projects. So, instead of buying a new gear motor, I decided to add a gear set with a 1:2 ratio to increase the maximum speed from 80 to 160rpm. The gears were brass and stainless steel â€œshop stockâ€ I had on-hand. The gear outside diameters were 2.6â€ and 1.3â€, respectively. Since the coil winder motor can use a variable DC input, the selected gear ratio will yield essentially any speed up to 160rpm. This is a good feature, especially when winding smaller gauge wire â€“ it gives the operator greater control over the process. The result is a lower probability the bright operator will break the wire or cause an unsalvageable, tangled mess of nice magnet wire. The aluminum bearing block and gear motor was salvaged from the old coil winder. Once assembled, the coil winder headstock assembly functions much like its counterpart on a lathe. I will attach an on-axis stud to the electromagnetâ€™s core and hold it in the drill chuck. Since the ~2â€ long core is relatively short, I donâ€™t anticipate needing a tailstock type support. No, I didnâ€™t draw the drill chuck in the diagram â€“ my lack of mechanical drawing skills is one of my failings at this point in my life, along with not doing a very good job at free-handed twist drill sharpening.
Rough Design â€“ Now On Paper (sort of)
Now that I had a rough design on paper, I could begin modifying existing or fabricating new components. First, I had to enlarge the 0.25â€ bore on the brass gear to 0.375 for the drill chuck shaft. With the gear in the lathe chuck, I reduced the total indicator run-out (TIR) to ~0.001â€ with a 0.002â€ thick shim between one chuck jaw and the gear. This TIR is fine for this low-speed project. Modifying the gear was a quick job â€“ drill 23/64â€ at 750rpm, followed with a 0.375â€ reamer at 240rpm. Again, I follow the practice of reaming a work-piece at 1/3rd of the speed used for drilling. Afterwards, I cut the 0.375â€ diameter shaft to its rough length â€“ just over 6â€ long. I used O-1 tool steel since itâ€™s produced with an excellent surface finish â€“ excellent for running within the journal type bearing surface represented by the aluminum bearing block. O-1 steel is tough too, even in its annealed state (as received). It is used for cutting tools and plane blades. Thus, it is a little tedious to thread with a die. However, using an HSS die works fine, especially after removing ~0.005â€ from the outside diameter. This practice still leaves enough material for full height threads. I only had to thread a 3/4â€ length for the 3/8-24 drill chuck anyway. No other work on the shaft was needed at this point since I intended to mount the gears and a cam with set screws. As weâ€™ll discuss later, the cam closes a micro-switch each revolution to actuate the electromechanical counter â€“ now an 18 year old relic from a long ago visit to Radio ShackÂ® in ancient times.
Next week, weâ€™ll continue with building the coil winder and we might go into more depth on calculating the flux we can expect from the electromagnets. Hope to see you then!