Power line voltage

"How large should my power line wiring be?", "What outlet size do I need?", and "Do I need 240 volts?" are commonly asked questions. There seem like simple questions, but the simple answers are usually wrong. Even technical editors at the ARRL have had problems understanding power line systems and how to evaluate them! The early AL1200 review was flawed because the ARRL Lab used a defective or inadequate power line source in the review. They were tripped-up by the very same thing that commonly fools other people.

Single family residential power lines in USA homes and most apartments are 120/240 volt single-phase 60 Hz systems. These lines have a center-tapped 240 volt winding. They use a common neutral and ground at the distribution transformer and at the dwelling entrance, with two 120-volt lines of opposite polarity. One could also say the "hot" conductors of 240 volt systems are 180 degrees out-of-phase with reference to ground, or to neutral, or to the supply center tap. Note that this does NOT make it a two-phase system! It is a single-phase 240 volt system with a center tap. The "hot" lines are opposite polarity and (essentially) equal voltages, but they are not a different "phase". It is a simple grounded center tap system.

The neutral is bonded or connected at the breaker panel to a safety ground. The safety ground is the round "third pin" of the outlet. The breaker or distribution panel is the only point where the safety ground should be grounded in house wiring, although a second ground point to the safety ground is generally unavoidable in amateur radio installations. The second ground point in amateur radio installations will render any GFI outlets or breakers for that outlet string unusable or unreliable. This ground loop is caused by the third "safety ground" wire that connects to equipment cabinets or chassis.

Most modern wiring provides 15 amperes at any normal 120 volt outlet. If the wire to the breaker or distribution panel is not too long, a normal 120 Vac outlet will often be adequate for amplifiers up to 1200 watts PEP voice SSB output, or 600 watts CW output. The key here is that the power mains wiring to the breaker or distribution panel must be reasonably short, and must not be a shared outlet with critical loads. Typically for standard 14 AWG copper wiring:

Power line voltage is always specified in RMS (root mean square) voltage. RMS is the quadratic mean, or square root of the mean of the squares of waveform values. RMS relates to perfectly shaped sine wave voltage or current waveforms by the special case of 2, where it is the square root of 2 (1.414) or inverse of 1.414 (.707). With a square wave, RMS voltage is equal to peak voltage, and RMS current is equal to peak current.

RMS quantifies voltage or current in a way useful for determining the work that can be done, such as when heating something. At the same time, there is no such thing as RMS power, although audio people often use the nonsensical term “RMS power” to describe sine-wave power. The consumer audio confusion probably comes from use of a sine wave’s RMS current and voltage to calculate power.

With a perfect sine wave, peak voltage is 1.414 times RMS voltage. In other words, sine wave RMS voltage is 1/1.414, or .707, times peak voltage under perfect undistorted sine wave conditions.

Although voltage has changed over the years, residential USA power mains have maintained a single phase 60 Hz frequency. Standardized power mains voltages started as 110 and 220 VAC systems. Around the end of WWII, standardized mains voltage increased from 110/220 to 117/234 VAC.

117/234 VAC remained the residential voltage standard for a few decades, changing to 120/240 VAC in the 1960’s.

In the 1970’s, the American National Standards Institute (ANSI) set current 120/240 specifications through ANSI Standard C84.1-1970. This standard specifies two voltage ranges, Range A and Range B, which included both service voltage and utilization voltage. Service voltage for this situation was usually interpreted to be at the meter, and utilization voltage was at the terminals of the utilization equipment. Electric supply systems were to be so designed and operated that most service voltages would be within the limits specified in Range A. The occurrence of service voltages outside Range A was to be infrequent. Range A service voltage was to be 120 V (+/-5 percent), which would be 114 to 126 V. Range A utilization voltage was specified to be between 110 and 126 V.

The current line voltage, since the late 1960’s in most locations and since the 1970’s by written uniform standard ANSI Standard C84.1, is 120/240 + - 5%. After 40 years or more of being 120/240, it is probably time we stopped calling it “220”, “117”, “115”, or “110”. In the USA, it is 240 volts or 120 volts.

USA residential lines employ a common-ground center tap at the transformer, where primary feeder neutral connects to the residential secondary center tap. A small generally-poor earth ground connects to each supply pole, and occasionally along long primary runs without transformers.

At the dwelling entrance, as required by national safety codes, all cables entering the building must share a common ground point. This is also the common bonding point for the safety ground wire to the neutral. This common ground point prevents significant voltage differences between grounds on cables or wiring inside the dwelling. There is a small earth ground rod or earthing system required. Generally this system has an earthing resistance well over 30 ohms, so it isn’t really much of a ground. It is, however, better than no ground at all.

By legal requirement, all cables entering the Hamshack, including the Hamshack ground rod or ground system, must be bonded to the power mains entrance ground. Again, as with the power, CATV, and Telco grounds, this is to prevent ground potential differences inside the dwelling.

The pole ground and the house ground help protect against voltage rise in the event of lightning strikes, power line ground faults, or open neutrals. While there should not be much potential difference, there is always some current flowing into these grounds to earth. This current flows because there is always some voltage drop along neutrals. That voltage drop excites the grounds rods with respect to earth and other grounds distributed along the power grid. As a matter of fact, if we drive two ground rods into the earth some distance apart, even somewhat far from power mains, 60 Hz voltage can detected! This voltage is excited by the earthing currents in our mains system.

As described above, power line voltage is specified in RMS voltage based on a pure, ideal, sine waveform.

Most amplifiers and power supplies, including switching supplies, use capacitor input filters. While most meters respond something around average or RMS voltage, capacitor input supplies operate from peak voltage. Peak voltage of a perfect sine wave is 1.414 times RMS, so our 120 VAC mains (without harmonics or clipping) crest at 169.68 volts peak. If we rectified the power line and filtered the DC with a perfect capacitor input supply, like most conventional and switch mode supplies do, we would have about 170 volts DC. This happens because the capacitor charges to the power line’s peak voltage at the crest of the sine wave.

When near full voltage, the power supply draws current only at peaks. If the supply is delivering 1 ampere DC near 170 volts, all of the energy would be supplied during a very short period at sine wave crest. Current from the power line would be many amperes, but for extremely brief periods of time.

Since load power is drawn only at sine wave voltage peaks, there is a high ratio of peak current to average or “heating” current. In the example above, while average current might be near 1.4 amperes, peak current would be several amperes. This ratio of peak to average current gives rise to something called apparent power factor (APF). APF is based on peak to average current, and is not same as standard power line phase-shift power factor caused by inductive loads, such as motors.

Nearly all radio and amplifier power supplies, since they almost always employ capacitor input filters, have a very high apparent power factor. The more robust we make power supply components, and the stiffer we design a supply in an attempt to stay near 1.414 times AC RMS voltage, the greater APF becomes. The stiffest, largest, most oversized supplies have the highest APF’s, demanding greatest attention to power mains’ equivalent series resistance (ESR) if we want to maintain that regulation.

Because the typical supply mainly works from peaks, average or RMS voltage has little practical use (aside from calculating heat). While this may seem complex, power supply regulation has to be calculated using peak current and/or peak voltage. It is common to see 5% voltage drop on an average or RMS meter, while power line peak voltages are dropping 15% or more. This can mislead us into thinking the power mains regulation is good and a power supply bad, even when the bulk of the problem is actually in the mains.

In an AL1200 amplifier operating from very stiff power lines, peak-to-average current ratio is about 4:1.

With 12 amperes RMS (heating) current, peak current will be about 48 amperes. While power line heating is calculated at 12 amperes, voltage drop is calculated at 48 amperes. A power line resistance of one ohm, for example, would produce only 12 watts of heat, while the same one-ohm resistance would reduce peak line voltage by 48 volts if the power supply system maintained the same APF! With a nominal 240 V RMS line, peak line voltage would drop from 339 volts down to 291 volts.

This is a 14% reduction in high voltage, while meter measured RMS or average line voltage would typically only change about 5%. (The exact amount would depend on waveform distortion and the meter.)

Power mains problems can be tricky, escaping even the most experienced amateurs. Even someone as experienced as the ARRL Lab can miss problems like this. The ARRL lab, reviewing an AL1200 amplifier, measured operating DC plate voltage of an AL1200 amplifier as 2900 volts under load, while the average-based RMS line voltage was fairly steady near 240 VAC.

Despite emphatic warnings the ARRL Lab had something wrong in the lab power mains system, the ARRL failed to investigate properly before releasing the review. The ARRL finally realized the error when, after installing the amplifier at W1AW, loaded voltage was suddenly a normal 3300-3400 volts.

The ARRL Lab’s problem was an expensive voltage regulator that held power mains’ average or RMS voltage steady, while allowing line peaks to sag over 15%. This caused the normal 3400 volts full load voltage of an AL1200 to drop to 2900 volts, while the metered voltage at the socket barely changed. The Lab measured good, stable, line voltage on a typical meter, but peak regulation was terrible because their expensive voltage regulator could not handle the amplifier’s APF.

While the realization they had a problem came too late to prevent false review data, at least it is a good learning tool for others. Good stable voltage on a traditional meter does not mean the power line system is problem free. The Lab missed a simple, easy, observation. It is electrically impossible to significantly decreased dynamic regulation inside a supply with accompanying heating or ripple.

Apparent power factor (APF) is high in capacitor input supplies, with peak currents 2-5 times average currents. The better the power transformer, the more disproportionate peak current becomes when referenced to average current. Because of high APF, voltage regulation in capacitor input supplies is largely a function of series impedances from the filter capacitors back to the power source. This undesired series impedance is normally dominated by resistance in the wiring back to the pole transformer, and by the amplifier’s power transformer.

Evaluating regulation with voltage measurements requires thought and care. A capacitor input supply works from the peak line voltage. Peak voltage does not change in proportion to average or RMS voltage. As a matter of fact, average voltage often barely changes when peak voltages drop a very noticeable amount.

Virtually all multimeters do not detect true peak voltage, and they also do not read RMS or average voltage. Most multimeters detect something around average AC voltage, ranging up toward peak voltage. Whatever they happen to read is corrected or adjusted to provide a pseudo-RMS voltage on the display. Unfortunately, this only works well with a sine wave. Since the supply only loads the peaks, the waveform squares. The average voltage hardly changes even with significant fractional-cycle peak clipping, which means significant DC voltage loss without a similar change on the power line meter.

To actually determine power line regulation when feeding a capacitor input supply, the multimeter should be a true peak reading meter.

In almost every tube- type amplifier, the high voltage meter provides a good way to determine power line quality. If plate voltage runs normal at idle, but falls well below manufacturer’s rated specifications under full load without undue carrier hum or power supply components heating, chances are good power line equivalent series resistance (ESR) is too high. It is electrically impossible to significantly decreased dynamic regulation inside a supply with accompanying heating or ripple.

Normally, losses inside an amplifier do not change much with power line voltage changes. Changing from 120 volts to 240 volts might increase or decrease life of some components, such as switches and relays, but overall dynamic regulation is generally not changed much. Operating voltage is not changed at all, provided the primary system is wired to exactly double voltage. This happens because most systems employ identical dual primaries, which paralleled for 120 volts and in series on 240 volts. With dual primaries, current in each primary and voltage across each primary remains the same regardless of 120 or 240 volt wiring, causing transformer losses and ESR to remain exactly the same.

Because of high APF, ESR which causes noticeable regulation issues can be surprisingly low. Wiring that normally handles a 1500-watt resistive load with minimal drop can have much worse regulation with a 1500-watt power supply load. Worse, a conventional multimeter might not show the line voltage loss.

This is because APF causes peak current demand to be high, which in turn clips the sine wave into a flat-topped waveform.

Normally, changes in performance come from changes in power line loading outside the amplifier. Performance changes do not come from efficiency changes inside the amplifier. By doubling voltage from 120V to 240V, we halve current. All things equal, the system has half the voltage drop at twice the line voltage. This results in four times better regulation, when expressed as a percentage, with no change in wire size.

Keep in mind this is four times improvement. A 0.1 ohm ESR line with 40 amps peak current would drop 4 volts out of 170 volts peak. This translates to a 2.4% loss in voltage. A system change to 240-volts results in 2-volts drop out of 340 peak volts. This is about 0.6% regulation loss. With a 3000 volt supply, we can expect about 50 volts more high voltage under load from power line changes.

This is, of course, a fictional case with 0.1 ohms loop resistance. This would be typical for a 25-foot run of #12 AWG (.05 ohms) to a good 200-ampere breaker box system with nearby pole transformer (typically around .05 ohms). My shop workbench feed measures ~0.1 ohms ESR, including the line transformer. If the power line has significant ESR, a change from 120 to 240 volts can greatly improve dynamic regulation. Whatever portion of total sag is caused by a 120V power line, that sag will be ¼ the amount. Voltage sag inside the amplifier will not change very much.

APF effect on dynamic regulation is less problematic on voice SSB. The power supply filter capacitors supply energy for voice peaks; the power line never seeing the full peak power input demand used on SSB voice.

If your amplifier uses a capacitor input power supply (including most domestic high power switching supplies), do not rely on normal ac voltmeters for power line stability measurements. Normal ac meters are usually fine for choke input supplies, or waveform distortion corrected power supplies.

Most amateur power supplies are capacitor input filter systems. When measuring power line voltage or current for high power amplifiers, almost every meter measures the wrong thing! The power supply runs of a very small portion of the sine wave near the top of each half, especially on the rising edge of the waveform. Virtually all meters measure average or pseudo-average voltage and current, so they do not measure the voltage the amplifier power supply requires. Meters are often calibrated in peak or RMS, but they often just apply a correction factor to the average voltage that is actually measured.

We might do the same thing manually, assuming peak voltage is 1.414 times indicated RMS voltage. This method is correct, and meters are often very close, when the measured waveform is a perfect sine wave. This is not the case with capacitor input power supplies. Conventional meters cannot be trusted to reliably evaluate power line health when the powerline is loaded by a capacitor input power supply. With a non-peak meter, a power line voltage reading can show no significant measured voltage drop, yet the powerline can be causing terrible voltage regulation and performance in a capacitor input supply. The ARRL case was an almost perfect example of a meter indicating a stable mains source, while the mains source was almost useless.

A capacitor input supply amplifier develops high voltage on sine wave crests or peaks. Because of that, the power line current to operate capacitor input supplies comes on peaks.

Here is an example of 811H secondary current and voltage for a plate current of 750 mA:

Secondary current is just over 4 amperes, while secondary voltage is just over 1500 volts peak. Transformer and power line loading is on the rising waveform edge, with a load duration of about 2.5 mS during each 8.3 mS power line half-cycle.

Measured transformer current roughly agrees with secondary current. At 120 volts, measured primary peak current is 32-amperes. One-cycle average current with a steady 750 watt carrier is 11-amperes.

This is why voltage drop needs to be measured with a true peak reading AC power line meter, or roughly estimated by watching the HV meter inside the amplifier. Abnormal voltage sag under load is almost always caused by inadequate power line regulation.

Power lines and circuit breakers should be minimally sized for heating currents, which are average currents. When using capacitor input supplies, voltage supply stability calculations should use about three times average current for maximum power amplifier dc input power.