|
Sections: Introduction | Power Electronics | Driver | Control | Interface
The Power Electronics consists of the electronic components that directly control the power delivered to the motor, and the associated mounting, support, and cooling hardware for components. This section does not cover the MOSFET drivers.
|
The bus bars are 1.0" wide by 0.25" thick, and 14" long, made of solid electrical grade copper and are silver plated. There is 12.5" available to mount power devices to.
For my implementation, the MOSFETs are TO-220 while the diodes are TO-247. The MOSFET pin pattern (left to right) is Gate, Drain, Source, while the diode pattern is Anode, Cathode, Anode. The hole patterns represented in the "Power Section" CAD drawing reflect this.
I used a metal-cutting bandsaw, though a hacksaw works just as well.
|
|
All of the holes are drilled and tapped for 2-56 x 0.25" brass machine screws. I printed the hole pattern using a CAD program (QCAD) and a laser printer, using legal size paper so that I could print the entire pattern on one sheet of paper. MAKE SURE that each pattern aligns the same way on each bar; the marked end of the bar on the paper should like up with the end of the real bar.
I used three pieces of 1.0" x 0.125" aluminum bar stock as drill templates. I sprayed them with a thin layer of 3M Super 77 glue, then carefully laid the patterned pieces of paper on them. I carefully center-punched each mark and drilled all the holes on a drill press. I then tapped dthreads in the first few holes of each aluminum bar to test the spacing. I actually used these to mock up the parts layout before drilling the expensive copper bars.
I then carefully laid and aligned the aluminum bar on top of the copper bar and drilled the first holes into the copper. After that, you can tap the first hole in the bus bar, and screw the aluminum bar on top of the copper bar to keep them aligned for subsequent holes.
This is a long and tedious process. Note that it's not necessary to drill all of the holes in the aluminum. You can drill 10 or so, use them to drill the copper, then slide the aluminum bar down so that only two or three holes overlap, screw the copper down, and continue making holes.
|
|
The bus bars are mounted into a cast plastic spacer piece that supports and separates the bus bars and heat sink.
To make the casting, I first constructed a pattern using bass wood. Bass wood can be found in your local hobby / model shop near the balsa wood. It's very good for making precision patterns because it's dimensions are very accurate. A 3" by 0.25" piece of bass wood is very close to being exactly that size.
The pattern was glued together using wood glue, then sanded and sprayed with 3 coats of acrylic gloss lacquer, to make it smooth.
I glued the pattern to a piece of foamcore (again from the hobby shop), ensuring that there was no gap beneath it. I then built a dam around the pattern using more foamcore and masking tape. Ensure that the masking tape makes a good seal against leaks.
I mixed up a quantity of OOMOO 30 RTV silicon mold making rubber and poured it into the pattern box. Pour slowly in a thin stream in one corner to avoid trapped air bubbles. After 24 hours, I de-molded the pattern by removing the side pieces of foamcore and gently lifting the new mold off of the pattern.
When making the mounts themselves, I sprayed the interior of the mold with silicone lube, then mixed and poured some Instacast with some milled fibers mixed in for strength. Make sure there is no moisture in your mold. Also, I baked the milled fibers at 250 degrees for 15 minutes to ensure they were dry. The resin is very sensitive to moisture. Additionally, I used "wine preserver" gas from the supermarket that I sprayed into the resin containers to displace moisture-laden air.
|
|
The capacitor bank is comprised of twelve 250V 1000uF Cornell Dubillier (P/N: 381LQ102M250K042) electrolytic capacitors, and twenty 400V 0.47uF EPCOS (P/N: B32652A4474J) polypropylene film capacitors. The are mounted on a 5oz FR-4, .060" thick copper-clad PCB. This PCB is bolted to the underside of the bus bars using .25" x .25" x #6 screw size brass standoffs and brass 6-32 x .5" screws, with #6 bronze lock washers.
The copper pattern is found in the Power Section CAD file.
Again, I used a laser printer to print the copper mask and screw hole locations. Unfortunately, my PCB blanks are only about 6 3/4" long, so I had to use two boards soldered & glued together. Using Super 77 glue, I sprayed the copper board surface, and carefully placed the boards against the backside of the paper patterns, holding it up to the light so that I could see the proper placment. Both boards were placed on the paper, then I cut the paper at the point where the boards joined using a scalpel (xacto knife would also work). This ensured that the pattern was properly aligned across both boards so that when they were joined, everything would line up.
I took my scalpel and used it to cut through the paper, tracing along the serpentine path that snakes between the capacitor leads to electrically isolate the positive and negative potentials. I then peeled up the paper exposing the copper. The scratches left in the copper by the scalpel I later used as a guide for etching
I then used a scratch awl to make a small indent in the exact centers of the holes for the capacitor leads and mounting screws. The holes for the small capacitors are 1mm (.039") in diameter, and the holes for the larger electrolytics are 2mm (.078") in diameter. I then used a drill press to drill the 76 holes in the board, using the appropriately sized drill bit of the 1 and 2mm holes, but I pre-drilled the 6-32 screw holes using the 2mm drill bit, which I then later drilled to the proper size (about .143"). Note that I simply used imperial drill bits that were near to the metric 1 and 2mm sizes. The hole size isn't as critical as getting in the correct location.
Finally, I covered the front and back (to prevent etchant from entering the holes) of the board with clear acrylic USPS packing tape. Where the tape edges met, I covered with scotch tape. I thoroughly burnished the tape down. I then used a scalpel to re-trace the scratches from earlier, and peeled the tape away from the serpentine path that I was going to etch. Again, I made sure the edges of the tape were well burnished. I then etched the two board halves in a solution of Ferric Chloride. This took over 3 hours at room temperature with occasional manual agitation of the solution.
|
|
