Part 1 showed us how to use unconventional transformer configurations to resolve equipment availability problems and unusual circumstances. Part 2 provides more examples, and a look at the tradeoffs involved. Let's begin with Case Five.

Case Five: Open-delta/open-wye connection (OD-OY). A fabrication shop acquired a lathe with a European programmable logic controller (PLC), powered by a 480V/240V autotransformer. Shortly after energizing, the controller experienced two catastrophic and expensive failures. The failures incinerated the circuit board and two metal-oxide varistors. Considering their connection to ground and rating of the onboard MOVs, the PLC designers probably did not intend the user to power the PLC from an ungrounded circuit.

A 12kV/480V unit substation served the shop as an ungrounded-delta system. Informal "corporate memory" revealed cases of phase-to-ground transient-overvoltage insulation damage in the high-bay areas. There were simultaneous fuse/breaker operations, too: symptoms typical of ungrounded low-voltage distribution systems.

A spare electrical parts crib in a shop-stock area carried epoxy-encapsulated machine tool control power transformers (CPTs): most with 120V secondaries, a few with 24V. The site's elevation made its low-voltage systems prone to electrical-insulation damage from lightning.

Our answer was powering the controller card from a grounded, separately derived source. The lathe's control console had room to replace the existing control power autotransformer with two dual-winding 150VA CPTs. We installed them with 480V primaries in parallel and 120V secondaries in series.

Integral type-M fuses protected the transformer secondaries. Class-J fuses in a 3-pole "pullout" block the size of an E-frame circuit breaker provided upstream protection.

After energizing this reworked installation, new symptoms arose. The controller's four electrolytic capacitors split open (detonated!) at the ends. This time, the MOVs kept their bright-red color. We then installed a digital multimeter with min-/max-/avg-logging functions; one night's readings displayed 1-sec sample voltages averaging 242V. The minimum was 239V, while the maximum was 260V.

ANSI standard C84.1 ["...Voltage Ratings (60 Hz)"] permits a worst-case excursion to 254V on a 240V circuit. Low utility system loading at night probably aggravated things. The PLC called for a 200-220V, 50/60-Hz supply. However, from day one its supply ordinarily didn't drop below 240V!

An unconventional application of 3-phase power solved the single-phase problem. The top diagram of Fig. 1, original article, shows the "fix attempt" (the first fix used separately derived control power). The bottom diagram shows the "fix success" (OD-OY arrangement to use the system and lower the control voltage to better match to PLC nameplate).

We rewired the control power transformer primaries in open-delta, reconnecting one transformer-primary terminal to the console's 480V B-phase bus. This put the formerly series-connected secondaries in open-wye. Now, instead of the two 120V secondaries producing 240V (22 each winding), the 120 degree phase shift between the secondary windings provided closer to 208V (1.732 each winding). After changing a 21/2% tap at the unit sub, a week's worth of 24-hr logs indicated voltage to the PLC at less than 220V, sparing another expensive board.

Case Six: Tee to Tee Connection (T-T). A plant had a 2000kVA 480Y/277V unit sub feeding process motor and lighting loads. It also had scattered single-phase 480V-to-120/240V dry-type transformers of limited capacity for convenience outlet circuits. The plant needed a small packaged HVAC unit. The unit had a 3-phase motor and two single-phase fractional-hp motors. Its size dictated lower-voltage motors and control power transformer. Unable to procure a 460V model at good price with delivery, we got an in-stock 230V unit locally.

Two 5kVA transformers hung adjacent to the lab (see Fig. 2 and Table 2, original article). With a short gutter section, GRC nipples, and a fused NEMA type-3R disconnect switch, the T-T connection provided a 240V 3-phase circuit with minimal effort.

Case Seven: Xi quasi-autotransformer connection (XQA). This is a variation of the OD-LM connection shown in part 1, case Two. See Fig. 3 (on page 59) and Table 3 (on page 60) of the October issue. You can use the Xi (Greek letter J, pronounced "zy") configuration to derive 208Y/120V power from a 240/120V, 3-phase, 4-wire (mid-point-grounded) delta source. The task needs only one single-phase transformer.

A shop leased a high-vacuum helium mass-spectrometer cart requiring a voltage unavailable in the building. A utility-owned, pole-mounted 240/120V, 3-phase, 4-wire delta bank served the neighborhood. The cart had a 5-hp, 200V, 3-phase motor; a 3kW, 208V, 3-phase heater; some 115V controls; and miscellaneous 1-phase loads powered through a 5-wire, 30A, 208Y/120V cord cap.

As an alternative to three 5kVA units, we used a 71/2kVA 240x480-to-120/240V single-phase transformer-See the sidebar (on page 58) for other options. A visual inspection, ohmmeter readings, 1kV insulation checks and a 120V excitation test verified the transformer as salvageable. The transformer lead j-box made a convenient place to secure a NEMA L21-30 receptacle in a 411/16-sq-in box. Five No. 10AWG XHHW conductors powered the assembly. A delta-fed 30A, 240V, 3-pole, 4-wire fusible disconnect switch supplied the transformer/receptacle.

We considered a pair of 240V/32V buck-connected transformers. Connecting this transformer set for proper phase-to-phase voltages would not have provided the correct wye-system phase-to-neutral voltages. This hookup could have damaged the cart's 115V components. The single-transformer configuration allowed the shop to complete the job as scheduled. It was less expensive and easier to obtain and install than a temporary 15kVA delta-wye bank.

Transformer application comments. In using nonsymmetrical 3-phase configurations, voltage balance is more difficult than with symmetrical transformations (delta-wye, wye-autotransformer, or delta-delta).

Though we had almost 10% imbalance in the phase currents of the OY-OD-fed dough-mixer motor (Case One, Part 1), during normal operation currents didn't exceed nameplate full-load current. For OD-LM (Case Two, Part 1), OY-LM (Case Three, Part 1), LQA (Case Four, Part 1) and XQA (Case Seven, Part 2) connections, an identical load applied A-phase to neutral (a-n) with more drop than if applied b-n or c-n. With the OY-OD interconnection, identical loads connected A- to C-phase produces greater drop than if connected A-B or B-C.

With the exception of the LQA and XQA connections, there is an inherent reduction in allowable load from total kVA transformer capacity, as compared to the usual three-transformer arrangement. The OY-OD, OD-LM, OY-LM and T-T (case six, part two) configurations can carry a continuous load to 87% of their combined ratings. The XQA connection method can supply a 3-phase load at 150% of its kVA rating. The LQA arrangement also serves a balanced load 150% of combined-transformer ratings.

In these first seven cases, the transformer banks must have primary and secondary overcurrent protection. You must ground OD-LM, OY-LM, and OD-OY secondary windings because they form separately derived "wye-voltage" systems. You can't ground LQA and XQA configurations at the transformer because windings connect to the source of supply. They do not constitute separately derived systems, similar to other autotransformer applications.

Usually, 3-phase motors are desirable over single-phase because they have three (or six) similar wye-connected windings of about equal thermal mass. Compare this to one (or two) heavier-gauge main windings and one lighter-gauge auxiliary or starting winding typical of a single-phasedevice. This difference makes 3-phase motors easier to protect from winding-insu lation failure by thermal overload.

The cost breakpoint for "plain-vanilla" 1-phase capacitor-start versus 3-phase induction motors is about a quarter hp. The comparison includes several attributes:

  • NEMA frame

  • Rigid base

  • Mechanical-output characteristics

  • Enclosure (e.g., open-drip or totally enclosed fan-cooled)

  • Relative prices

In Part 3, we'll look at more conventional configurations that do not fit the "rules" imposed on the first seven cases. They merit a close look because of their cost-savings and analytical applications.