This is Electrical Photos page

* I232, E232 Communication Components for RS232 *
* Electrical Isolation, Fast Data Transfer, and Long Cables. *

Table of Contents
HOME I232 I232 Photos 1-Wire® Components 1-Wire® Photos

INFORMATION on Henry Ford Museum, Greenfield Village, & Ford Rouge Factory Tours at

* Photographs of transient current waveforms *
* from a 100 W 120 V tungsten filament bulb. *

The experimental setup:

The input voltage source is a nominal 120 V 60 Hz 20 A circuit. A Triac controlled by a reed switch is used to switch voltage to the 100 W bulb. Cascaded 555 timers are used to control the turn on phase angle from the voltage zero crossing, and the on duration. Single shot mode was used.

Instrumentation is a Fluke Hall device current probe with 2 V output at 20 A input, and a Tektronix oscilloscope.

P1 --- Photo 2742A1. The transient current from turning on a 100 W 120 V tungsten filament lamp at near the voltage peak. Time base 5 ms/cm.

Peak current is 16 A and 1 cycle later it is down to about 2.5 A. When steady state is reached the peak is 1.2 A.

The upper trace is the square wave generated from the 60 Hz voltage zero crossings and used to synchronize the turn on of the lamp.

P2 --- Photo 2744A1. Transient current from turning on a tungsten filament lamp at near a voltage zero crossing. Time base 5 ms/cm.

The peak current is 6.5 A and much lower than when turned on at the voltage peak. One cycle later it is down to about 2 A. When steady state is reached the peak is 1.2 A.

The upper trace is the square wave generated from the 60 Hz voltage zero crossings and used to synchronize the turn on of the lamp.

P3 --- Photo 2748A1. This is the same as P1 except that here the time base is increased to 20 ms/cm.

P4 --- Photo 2746A1. This is the same as P2 except that here the time based is increased to 20 ms/cm.

P5 --- Photo 2750A1. This load is an Ohmite 10 ohm 25 watt power resistor with 120 V applied. Note this is 144 W into a 25 W resistor. If the duty cycle is short this is not a problem. I did slightly cook the resistor by not allowing enough off time. However, the value of resistance was not affected.

Note: even at this overload level there was no visible time variation of resistance from self heating. There is much more thermal mass to heat in this device compared to a thin filament in a vacuum enclosure where most heat is transferred via radiation. There is little thermal conduction in a lamp bulb.

It appears I have some calibration error but that is of no importance relative to the goal of showing what happens to the current vs time. The actual voltage across the resistor might be slightly lower than 120 because of line drop. It was not measured during the pulse load.

* Photographs of transient current waveforms *
* from a Signal Transformer A41-175-16 Transformer. *

P6 --- Photo 2757A1. Display of inrush current to a small transformer, 175 VA, on application of power to the primary. The high peak inrush current is a result of the last flux state in the transformer core from the time of last turn off, and that when it is now turned on that the input polarity is such as to force the magnetic core more into saturation.

It takes a large number of input cycles, way beyond the right side of the screen, for the core to settle into a balanced state conditions. The peak inrush current here is 40 A, and when steady state is achieved it will be in the range of 250 MA, about 1/160 th.

Note: this is an unloaded transformer. With a full load the primary current would be about 1.5 A from the load and excluding the excitation current to the core of 170 MA.

P7 --- Photo 2756A1. This is the same set of conditions as photo P6 except there was a different residual flux in the core at turn on time.

P8 --- Photo 2759A1. This is the same transformer after after steady state conditions have been reached. This scale is more sensitive by a factor of 50.

The waveform is now symetrical and the current level is very much below the turn on transient.

At 120 V the unloaded power input is 3.5 W. Thus, no load losses are about 2% of full load power capability.

* Photographs of Plots for Comparison of Incandescent *
* with Compact Fluorescent Lamps. *

P9 --- Photo 2769A1. This is the reference lamp to which all the following plots are compared. It is a 75 W 120 V normal tungsten filament lamp.

All the plots have the same physical size and scaling. Thus, overlays of one plot on another provide a meaningful comparison.

All of the actual power curves, labeled Power, use the left hand scale. All normalized values use the right hand scale.

What is a normalized curve? These curves have the actual measurement scaled to the measurement of the 75 W reference bulb at 120 V input. In other words the value is represented as a percentage of the reference device. On our normalized scale 100% equals 1.00 .

Note: the tungsten filament light intensity drops off more rapidly than the input power. Also the power curve for the tungsten filament is closer to a straight line than it would be for a invariant resistor. A tungsten filament tends toward being a constant current load with respect to voltage variations.

The horizontal axis is in volts measured with a Fluke 87 true RMS meter across the lamp. Voltage was a sine wave from a Variac for all plots except the two plots of a phase shift dimmer control of the GE dimmable CFL.

Light intensity was measured with a Clairex CL505L cadmium sulfide photocell 5 feet from the lamp. The light to the photocell was from the side of the lamp. The photocell resistance was measured with a Fluke 27 in resistance position. Cell resistance was 1040 ohms for the reference lamp at 120 V. Light intensity is inversely proportional cell resistance. Dark resistance (leakage resistance in total darkness) is over 10 megohms for this individual cell, manufactured in the early 1960s, and is glass enclosed.

A two coil Simpson 75 W full scale wattmeter was used for power measurement.

P10 --- Photo 2770A1. Plot for a 100 W 120 V tungsten filament lamp. I only had a 75 W meter and thus these curves terminate near 100 V. Note: the similar shape to the curves in P9.

P11 --- Photo 2771A1. This 75 W carbon filament lamp was made by General Electric as a replica of an early Edison lamp for the 1979 100th anniversary of Edison's first successful operation of a lamp.

Back to 1879. Sunday October 19, 1879 the the latest experimental lamp was finished and the life test started. On the 21st the lamp was still burning and then Edison gradually increased the voltage until the filament failed. At this point the bulb was broken to allow examination of the filament. See Chapter XLVI page 351 of "Menlo Park Reminiscences", Vol 1, by Francis Jehl.

Notice how much less efficient this lamp is than a modren tungsten filament. This also means the color temperature is much lower.

P12 --- Photo 1373A. Front cover of "Menlo Park Reminiscences", 1937. Thomas A. Edison, Henry Ford, and Francis Jehl circa 1929 in the restored Menlo Park Lab.

P13 --- Photo 1372A. The Edison lamp used in the P11 experiment operating at a low supply voltage to make an interesting picture.

P14 --- Photo 2772A1. A GE FLE26HT 3/2/DV/SW Compact Fluorescent Lamp that is designed for use with phase shift dimmers. In this plot the input is a sine wave of adjustable amplitude. Cutout occurs just below 100 V. Note: the relatively constant light output in comparison with a tungsten filament lamp.

This 26 W lamp produces about the same light output as the 75 W tungsten incandescent lamp, is more stable with voltage variation, and uses about 37% of the power of the incandescent. This lamp is not dimmable with a sine wave dimmer.

Might be useful for someone that has light flicker problems from air conditioner cycling.

P15 --- Photo 2773A1. A Nivision SKU 599-526 5500 K CFL rated at 27 W. This is approximately comparable in light output to the reference lamp at 120 V. Dimming with a sine wave voltage source is quite good. Does not work well with phase shift dimming. About 40% of the reference lamp's power is required for the same light output at 120 V. The dimming curve is quite different than that of an incandescent.

P16 --- Photo 2774A1. A Nivision SKU 159-452 3500 K CFL rated at 30 W.

P17 --- Photo 2775A1. A GE FLE26HT 3/2/DV/SW Compact Fluorescent Lamp that is designed for use with phase shift dimmers. In this plot the input to the lamp is a low cost two wire phase shift dimmer. I believe the dimmer is more of the range problem than the lamp. The input excitation to the dimmer was a 120 V sine wave,

Note: the stark contrast in the operation of this lamp with the phase shift dimmer compared with the operation from a sine wave dimmer.

P18 --- Photo 2776A1. Same lamp and operating conditions as P17 except here is shown the dimmer turn on phase angle vs intensity.

* Photographs Relating to Four Terminal Resistors. *

P19 --- Photo 2778A1. Photo of a Weston Shunt. This is an application of a 4 terminal resistor. The purpose is to eliminate errors in the voltage measurement across the resistor from variation in the contact resistance of the current conductors to the resistor.

P20 --- Photo 2766A1. My technique for a 4 terminal resistance to measure wire resistance.

P21 --- Photo 2768A1. Four terminal connection for measuring resistance of a 10 ft length of EMT conduit. The voltage measurement point must be inboard of the current injection point.

* Plots of Voltage and Power at my Home *
* Using the TED (Energy, Inc.) Monitor System. *
* Unexpected Sq-Wave Modulation of the Voltage. *
* Uncorrelated with my load or my neighbor's load. *

P22 --- Photo 2834A1. Plot of the measured voltage and power to my home when an unusual repetitive voltage variation occurred. This plot is 33 minutes starting at 1250 EST on 4-4-09, a Saturday, The unusual voltage variation started at 1259:25. Note that my variations in power load do not correlate with the voltage variations. The voltage rise periods are about 45 to 70 seconds long, and the drops about 70 to 80 seconds.

The "TED" monitoring system measures voltage and current, and calculates power each 1 second. This data is output to a display and to a collection computer. My bench experiments indicated that they actually calculate power and not just VA. Each of the plots here have 2000 time data points, or 2000/60 = 33.3 minutes per plot. The start of each plot is about 30 minutes from the start of the previous plot.

P23 --- Photo 2835A1. The variations terminated at 1328.

P24 --- Photo 2836A1. The 1 minute on and off variations restart about 1420.

P25 --- Photo 28377A1.

P26 --- Plot 1010031TA-150. My whole house power consumption 10-3-2010 for 24 hours
with 1 second resolution. Spikes are motor inrush.The resolution in the .JPG
is better than what you see on the screen. The screen is limited to 640 pixels.

P27 --- Plot 1010031TB-150. My whole house power consumption 10-3-2010 for 4 hours
with 1 second resolution from 0000 to 0400. Spikes are motor inrush. Two freezers
one refrigerator, and possibly a furnace blower.

P28 --- Plot 110814HRB-150. My whole house power consumption from
11-28-2011 thru 11-14-2011.

The plot time axis is from 0 to 25 hours. 0 is midnight. Each day's data is superimposed on all the other days.

The power line voltage becomes less stable centered about mid afternoon. Voltage is more stable through the night.

The power points data shows the variation through the day. Power measurements are made every second. The 3600 measurements made in an hour period are added together, and then divided by 3600 to obtain the average power for the previous hour. Note: this value for average power for the last hour is numerically equal to the energy in KWH used in the last hour.

P29 --- Plot 091031A1-150. Admiral freezer, unopened for 24 hours.

The plot time axis is from 0 to 24 hours. 0 is midnight.

The voltage plot shows the voltage drop from the freezer load with about 100 ft of branch circuit Romex.

The power plot shows a running power of about 300 watts, a large startup transient, an initial higher current of about 400 watts just after the inrush peak, and a gradual drop from the 400 watts to a final value of about 280 watts. The duty cycle is about 46 % with a typical on time of 1.03 hours and an off time of 1.20 hours. Note the perodic consistency.

P30 --- Plot 091031L1-STEP-150. Same as P29, except only a very short time segment.

The plot time axis is from 2.1265 to 2.1400 hours, 48.6 seconds (2 in the morning). One second is approximately 0.0003 hours. Our conversion from hours, minutes, and second to decimal hours at a resolution of 0.0001 hours provides a small jitter in data points, but no cumulative error. More exactly 1 second equals 0.0002778 hours.

The TED system has a 1 second offset of power relative to voltage. This results from the way TED processes the data. In reality the voltage and power changes are in sync. TED has power change lagging voltage change.

A step curve is plotted to provide a useful interpretation of the data. Note the power value from TED is the average for the last 1 second period. Thus, changes within the 1 second period are averaged and no detail exists as to what actually occurred in that period.


PO Box 1288, Ann Arbor, MIchigan 48106-1288
voice 734-930-6136
This web site is      Our e-mail address is
Copyright©   2003-2014   Gordon A. Roberts    All rights reserved.     141201-2338