Design Implementation

After deliberation and following design review, modifications were made to the carriage design and copper contacts to provide better response to the contact surfaces. Final development proceeded with the arrival of the main components and design decisions regarding improvements to the preliminary design.

Driver, Motor, and Controls
Upon arrival of the stepper motor, driver, and linear rail (comprising our linear rail system, or LRS), work began with trying  to get the components integrated and operational. Extend-retract buttons were incorporated into the system, and, ultimately, the LRS was set up to change positions linearly every 1.375" (the approximate distance from one completed tap change to the next). The maximum speed of the LRS is 2 in./sec.

Figs. A1 and A2. Integrating  IDEA drive with the stepper motor (left). Working to  get the drive and motor synchronized and operational.

LRS Operation Two Views

Figs. A3 and A4. Left image: setup showing the extend-retract buttons (left arrow)  and on/off switch  (right arrow) comprising the control system. Right image: manipulating the carriage along the track using the extend and retract buttons.

Like many things on the project, the controls underwent several design changes to reach their final state. The white breadboard was replaced with a smaller and more practical solderless unit. A 7-1/2" by 5-1/2" opening was created in the top of the cart and a box, 1-3/4" deep, was constructed to house the controls and fitted to the opening. A 1/4" thick clear acrylic panel, held down with four screws, completes the control interface. The power ON/OFF switch is located immediately below the battery pack. To the right of the power switch are the extend and retract buttons that move the tap changer in either direction along the track. 

control box
Figs. A5, A6, and A7. Control box under construction (left). Two views of completed control system for the phase-shifting transformer tap changer.

Modifications have been made to the carriage design from the previous page. The plates were laser cut in both clear and dark acrylic material, before the clear acrylic was selected for final design. Wave springs were introduced into the brush holder to provide reaction for each resistor bridge. The carriage was then assembled with four long machine screws (3-1/2" by 1/4") comprising the four upright posts connecting the top and bottom plates. The carriage was then mated to the LRS for testing of the mechanical aspects of the system. The resistor bridge was redesigned to allow for improved contact between the brushes and copper contacts. 

Brush in new plates.
Fig. B1. Set of three brushes incorporated into carriage bottom plate by way of an earlier cartridge design.

wave spring
Fig. B2. Wave spring used to generate contact force between brushes and contacts.

carriage three views
Figs. B3, B4, and B5. Assembled carriage (inverted) mated to LRS (left). Close up of carriage from side and above at an angle (right).

Carriage Two Views
Figs. B6 and B7. Assembled carriage traveling along LRS: side and top views.

close up
Figs. B8 and B9. Side (left) and head-on close views of carriage. 

Resistor Bridge

The resistor bridge, the central component of tap changer, was designed for optimum contact between the brushes and copper contacts. The constituent parts of the resistor bridge--the brushes and resistors--were selected based on the design specifications of the tap changer and the operating parameters of the transformers.

brush and resistor
Figs. C1, C2, and C3. Carbon brush (left), resistor (with wires attached), and brush and resistor joined (second from left). The carriage will hold nine carbon brushes and six resistors. The brush dimensions (not including the wire) are 1-1/4" long by 6/10" by 5/10". The size of the brushes is important for meeting our 80 A/in-squared current density goal. Each of the 2-ohm, wire wound resistors is approximately 1-1/3" long and 4/10" wide (not including the connecting wires). The resistor rating was dictated by the maximum voltage and amperage flowing through the tap changer, as well as the expected speed of carriage movement. 

Two views of cartridge

Figs. C4 and C5. Exploded (left) and intact views of a later design of the resistor bridge cartridge.

assembled two views
Figs. C6 and C7.
Assembled cartridge top view (left) and side view.

carriage bottom plate
Figs. C8 and C9. Three cartridges inserted into carriage bottom plate: top view (left) and bottom view. One wave spring (Fig. B2) is inserted under the head of each of the 1/4-28 x 1/2 socket head cap screws that hold the cartridge to the acrylic bottom plate to provide downward contact force.

carriage front and back
Figs. C10 and C11. Views front and back of cartridges inserted into the carriage bottom plate.

bottom carriage
Figs. C12 and C13. Views top (left) and underside (bottom) of cartridges fully assembled around carbon brushes and integrated into carriage bottom plate. All three cartridges were manufactured from ABS plastic in the University of Idaho's 3-d printer. Note the 2-ohm resistors (two per cartridge), completing the resistor bridge assemblies.

Contacts and Base Plate
The final design of the copper contacts specifies a flat contact surface. A prototype was made of each design: one with a threaded exterior and another with a threaded interior. The two prototypes were then test-fitted on one of the test pieces of acrylic. The decision was then made to go with the contacts with the threaded exterior. Each contact will be fastened to the acrylic contact board using 5/16" brass hex machine nuts holding a 5/16" ring terminal in place.

copper contacts
Figs. D1 and D2. Two contact prototype models (left), and top of copper contact used in the final design.

Shaping base plate
Figs. D3
and D4.
Shaping the base plate on the manual mill (left). Drilling holes in side of base plate.

Drilling the board
Figs. D5 and D6. Drilling the center holes (left) and the half-inch holes (right) on the Haas mill.

The base plate began as a piece of clear acrylic with dimensions 18.150" x 6" x 0.708". It is important that the plate holes and the contacts be precise. In that regard, the base plate drawing and dimensions were entered into MasterCam and the board drilled on a Haas mill. Center holes were drilled first, followed by the 5/16" thru-holes, and then the 1/2" counterbores. The fluid running over the board during the process is a  mixture of water and coolant. Prior to drilling, the board was shaped on a manual mill by having its edges faced with a 1" end mill.

Base plate.
Fig. D7. Base plate with holes drilled.

The first step in shaping the thirty-three copper contacts  was to write a program in MasterCam. That program was entered into the Haas CNC lathe. Five separate CNC lathe tools were used to cut and shape each contact. After machining was completed on the Haas lathe, the contact was taken to the Logan lathe where shaping of the contact face was completed using an aluminum holder made specifically for the purpose of holding each contact in the same position with respect to the Logan facing tool. The finished contacts extrude 0.025" above the surface of the acrylic base plate.

haas and logan lathe work
Figs. D8 and D9. Shaping a copper contact on the Haas CNC lathe (left) and the Logan lathe.

finished base plate
Fig. D10. Base plate with copper contacts in place.

To ensure all contacts were the same height with regards the carriage, the base plate with copper contacts attached was mounted on the Haas mill. A 1/2" end mill traveled up and down each row, removing 0.002" with each pass, except for the final pass when only 0.001" was removed (0.013" removed in all) until all contact surfaces were on the same plane with respect to the z-axis. As a final touch, the copper contacts were removed one at a time from the board to a Logan lathe to have residual burrs removed and the contact surfaces made smooth.

Base plate after end mill
Figs. D11 and D12. Base plate contacts undergoing end milling process on the Haas mill (left). Base plate after finishing work on the Logan lathe. Brass hex nuts underneath the board hold the contacts to the board.

After end milling, the copper contacts were of varying heights with respect to each other, owing to slight variance in the thickness of the acrylic base plate. As a precaution, each contact was chemically etched with its own identification number so that  in the event more than one contact was separated from the base plate at the same time they could be restored to their original positions, thereby maintaining an even contact surface across the base plate.

etched contact
Fig. D13. Bottom of copper contact chemically etched with a unique identification number.

Cart Preparation
Four hex-head cap screws, 2-1/2" by 3/8", were used to secure each of the shelves. The shelves are 3/4" thick MDF. Sharp corners were rounded using a router with a 3/16" radius corner bit. 

A decision was made to wire the tap changer directly to the transformer through the top shelf.  To that end a 5" by 15-1/4" opening was made in the top shelf to accommodate the wires for all thirty-three contacts. The hole is centered 23-1/2 inches from the leading edge of the table (toward the back of the top shelf) to create a workspace around the front of the tap changer.

Figs. E1 and E2. Installing supporting platforms on the cart (left), and rectangular opening created to accommodate wires.

After cutting the opening through the top shelf the next task was to construct the support structure for the LRS and carriage. Two pieces of acrylic, each approximately 11/16" thick, 6" wide, and 5-1/2" tall, comprise the main support structure.  Two more acrylic pieces, each 1-1/4" tall, cap off these structures and  form a clamp on either end of the LRS.Triangular-shaped reinforcement pieces (two per end) will be fitted to each end piece,  completing the support structure for the tap changer.

Figs. E3 and E4. Base plate (left) positioned over hole, and partially-completed support system holding up the LRS and carriage.

Two views present setup
Figs. E5 and E6. Two views of tap changer (still under construction) as of March 27, 2012.

second opening for wire
Fig. E7. A second rectangular opening, 10" by 1", was cut and centered behind the tap changer to accommodate the wires coming out of the top of the carriage. A wire leading from each of the three resistor bridges passes down through this aperture and connects with the exciter transformer below. A 1" dia. hole was drilled in cart top near the stepper motor to accommodate the wires between the motor and the controls.

cart finish
Figs. E8 and E9. Left: finishing touches on the cart included installing 1/8" thick clear acrylic panels to the sides and front and adding a hinged door to the back. Right: after coats of wood stain and polyurethane seal the cart is ready to have the transformers and tap changer re-installed for display at Expo.

Transformer Preparation and Integration
Preparation for testing required wiring the transformers to the new tap changer. To minimize "wire clutter," the wires leading to the exciter transformer were trimmed and new ring terminals fitted to the trimmed ends. Also a linkage connecting the windings on the exciter transformer was replaced with a solid bar, for use when the transformer isn't connected to the tap changer. Finally, four 1/2" holes were drilled in the bottom plank so that the transformers could be securely bolted to the cart.

fcopper bar four views
Figs. F1 (TL), F2 (TR), F3 (BL), and F4 (BR). Top left: previous connecting linkage atop the exciter transformer. Top right: drilling the replacement bar. Bottom left: comparisons between old and new connecting pieces. Bottom right: new bar in place on the exciter transformer. The dimensions of the bar are 14" by 1" by 1/3". 

Wiring to base plate
Figs. F5 and F6. Wires leading to the partially-wired base plate as seen from below the base plate (left) and above the base plate looking down.

transformer wiring
Figs. F7 and F8. Transformer wiring after wires had been trimmed.

Testing the Tap Changer
As refinements and finishing touches were being applied to various components, testing began on the mechanical and electrical properties of the partially-integrated tap changer.

Testing April 5
Figs. G1 and G2. April 5, 2012: testing of the partially-assembled tap changer using a digital multi meter (DMM).

On April 10, 2012, all of the components of the tap changer were fully integrated in preparation for a test of the complete system. The tap changer was initially tested for its mechanical capabilities in the Gauss Johnson Senior Design Suite. This test consisted of using the two control buttons to repeatedly move the tap changer back and forth across the contacts of the base plate. Adjustments were made until the team was satisfied as to the tap changer's mechanical capabilities.

Click on the link below to see a demonstration of the tap changer's mechanical motion:

Mechanical Test

Later that afternoon the tap changer was moved to the University of Idaho's Power Laboratory in the same building for the testing of its electrical phase-changing capabilities. SEL engineer Doug Taylor was present to supervise and direct this test and to oversee the proper wiring of the transformers.

Figs. G3 and G4. April 10, 2012: wiring and power connections for testing the tap changer in the University of Idaho's power laboratory.

Fig. G5. Another angle of test setup. The power analyzer is in the lower left quadrant of the picture.

A Fluke 41B Power Harmonics Analyzer was used to test the output voltage on one line during each successive tap change. The testing started at the 230 V tap and moved across the board in successive steps all the way to the 23 V tap, then back again, under a 1.8 amp load current. The power analyzer measured the voltage added in the series transformer between the input (source) side and the output side. Results were called out during the test from visual reading of the analyzer and recorded by hand contemporaneous with the testing.

Click on the links below to view the power lab tests:

Electrical Test Part One
Electrical Test Part Two
Electrical Test Part Three

The test results, showing approximately 25 volts at maximum phase shift (25 V corresponding to 10 degrees of phase change), appeared to demonstrate the phase-changing capability of the tap changer. Also, the results were consistent in both directions of travel.

April 10 results
Fig. G6. Results from first electrical test of tap changer.

On April 20, 2012, SEL engineer Doug Taylor performed a second test on the tap changer in the UI Power Laboratory. This time the measuring instrument was an SEL 487-E Station Phasor Measurement Unit. The tap changer was having difficulty moving from tap to tap, a problem which was remedied in time for Expo. Because of this latter problem only a few of the taps rendered useful data on this second electrical test. However, the data collected showed a voltage phase shift as well as a smooth current transition when moving from the seventh to eighth tap windings, providing further evidence that the tap changer works as intended.

2nd test
Figs. G7 and G8. Relay used to gather data (left). Connections between relay and the transformers beneath the tap changer.

April 20 results
Figs. G9 and G10. April 20 test results showing phase shift (left) and smooth current transition (right) (test results courtesy of SEL).

Snapshot Day (March 6, 2012)

Snapshot March 6
Figs. H1
,H2, and H3
Carriage and LRS mated to mock up of support structure (left), front view, and side view (right).

Expo (April 27, 2012)

expo I
Figs. I1, and I2 PSTTC booth at the University of Idaho Engineering Expo in the SUB Ballroom.

expo II 

Figs. I3 and I4 Two more views of Expo display.

expo side view
Fig. I5.  Side view of tap changer.