General Guidelines for Writing

Electrical Engineering Laboratory Reports

Edited and Adapted for Use by Andrews University

Engineering Students

2007

David M. Beams

Department of Electrical Engineering

University of Texas at Tyler

Lucas P. Niiler

Department of Languages and Literature

University of Texas at Tyler

July 6, 2004

Introduction

Engineers often find it necessary to write technical reports to describe their work to management, patent attorneys, clients, or colleagues. Technical communication skills are crucial to good engineering practice and professional achievement. Experience in writing formal laboratory reports is therefore a part of the electrical engineering curriculum at the University of Texas at Tyler.

The goal of writing a laboratory report is to convey clearly and accurately what was done in the laboratory. Style, format, and organization of the laboratory report are thus as important to this end as is technical content, but weighting of each of these factors in grading is left to the discretion of the instructor.

General style guidelines

· Engineering laboratory reports must be prepared with a word processor.

· The engineering laboratory report should be readable by a person who is technically trained but not necessarily familiar with the experiment. That person should be able to replicate the experiment after reading the laboratory report.

· Written communications should reflect a professional approach to technical content and style. Avoid colloquial expressions; for example, “The prototype operational amplifier had much greater gain-bandwidth product and slew rate than an LM741” is acceptable; “Compared to a 741, this circuit was a screamer” is not.

· Write in third person. “I,” “we,” and “you” are not to be used in a laboratory report. Nor is the second-person imperative mood to be used (e.g., “Measure the voltage gain of the common-emitter amplifier”).

· A title page must be included. All pages except the title page must be numbered sequentially, beginning with "1." Page numbers are to be at the bottom of the page, centered or right-justified.

· All diagrams, illustrations, and graphs except tables are called "figures." Graph axes must be labeled legibly and units must be specified (e.g., mV, A, W, kHz). Each figure must be cited in the report text. References to figures in text are of the form "Fig. 1," but the word “Figure” is not abbreviated when it is the first word of a sentence. All figures are to be numbered sequentially. Figures will be computer-generated and embedded in the report text whenever possible. Figures must be centered and must be contained on a single page. Figures also must have captions that are left-justified or full width (not centered). The entire caption must be contained with the figure on the same page. See the example below:

Fig. 2. Test circuit for measuring the I_C vs. V_CE characteristics of a 2N3906 PNP transistor. All components except the 2N3906 and collector resistor R_C form a voltage-controlled current source to supply user-controlled base currents to the 2N3906. Indicated supply voltages correspond to the outputs of the HP E3631A DC power supply. The HP34401A DMM is configured as a dc voltmeter by LabVIEW virtual instrument program BB_QPNP.vi. Values of emitter resistors R_E and collector resistor R_C are given in the text. Currents I_B and I_C are respectively the base and collector currents of the 2N3906 transistor under test. Voltage V_ctrl (supplied by the 0- –25V output of the HP E3631A) controls I_B. V_ref is the voltage at the anode terminal of the LM336 voltage-reference device.

· Equations are to be numbered sequentially using a right-justified number enclosed in parentheses. Reference to equations in your text is made by the equation number enclosed in parentheses. Variables used in equations must be defined. See the following example:

The control voltage V_ctrl required to produce a specified base current I_B to the 2N3906 under test in Fig. 2 above is given by:

(1)

where R_E is the value of the emitter resistors and V_ref is the voltage at the anode of the LM336 voltage reference. The derivation of (1) assumes that the input-bias current of the LM324 operational amplifier is negligible and that the emitter-to-collector current-transfer ratio α of the 2N4401 is 1.0.

· Tables are to be numbered sequentially. The word "Table" is not abbreviated. All tables must have legends or captions and must be numbered sequentially (e.g., “Table 1,” “Table 2,” etc). Table legends are placed at the top of their respective tables.

It is a common stylistic practice (but not required) to write captions and legends in a smaller type font than the body of the main text. This text, for example, is in 12-point Times New Roman font while the legend of Table 1 below is in 10-point Times New Roman font with the table number in boldface. The same consideration applies to captions for figures.

Table 1. Experimental apparatus used for the measurement of I_C vs. V_CE characteristics of the 2N3906 PNP transistor.

Description	Manufacturer	Model / Rev.	Notes
Triple-output DC power supply	Hewlett-Packard	E3631A
Computer, NT workstation	Compaq	Professional workstation	GPIB interface card installed
Digital multimeter	Hewlett-Packard	34401A
Test circuit	N/A	N/A	Described below
GPIB cables	(unknown)	(unknown)
LabVIEW	National Instruments	V5.1
LabVIEW virtual instrument program BB_QPNP.vi	Allen Barger and David Beams, Department of Electrical Engineering, University of Texas at Tyler	Version date: 10 April 1998	Virtual instrument for measurement of common-emitter PNP characteristics

· SI units are to be used in laboratory reports unless otherwise specified. These units are the meter (m), kilogram (kg), and second (s), and their derivatives such as the Newton (N), Joule (J), and Watt (W). Electric units include the volt (V), ampere (A), Coulomb (C), ohm (Ω), and Siemens (S). Prefixes commonly used with these units are pico- (p), nano- (n), micro- (m), milli (m), kilo- (k), mega- (M), and giga- (G). Frequencies are measured in Hertz (Hz).

·

Text Box: vo=zeros(1,length(t));
for k=1:length(frequency_vector)
vo=vo+A(k)*cos(2*pi*k*t)+B(k)*sin(2*pi*k*t);
end

Short sections of computer code written for the experiment may be incorporated directly into the text or attached as appendices. Long listings (of length greater than one-half page) should be attached as appendices. A listing incorporated into the text should be enclosed in text box and include a caption and figure number.

Fig. 9. Matlab code written to compute the time-domain output voltage from the Fourier-series components of the original periodic waveform. Vectors A and B contain respectively the amplitudes of the cosine and sine terms.

· Reports written with word-processing software should use 14-point font for the title page and 12- or 10-point font for text. Legends and captions may use smaller fonts, but fonts smaller than 8-point should not be used. Times New Roman is preferred but not mandatory. Serif fonts (e.g., Times New Roman) are preferred to sans serif fonts (e.g., Arial). The chosen font should be consistent throughout the entire report, including text appearing in figures or tables.

· Variables used in text and equations should appear in italics (e.g., ”Q”). Subscripts are generally not italicized (e.g., “Z_in,” “A_vol”). Subscripts are italicized when they are variables in their own right (e.g., “A_j,k” refers to the element in the j^th row and k^th column of a matrix named A).

· Reports written with a word processor may use either left-aligned or full-width (aligned left and right) text.

· Materials incorporated into the laboratory report from textbooks or journal articles must be referenced. Numbers enclosed in square brackets (e.g., “[1]”) are used to denote a reference in the body of the text. A “References” section must be included in a laboratory report which contains cited materials. References should follow the formats below:

[8] G. F. Klir and T. A. Folger, Fuzzy Sets, Uncertainty and Information. Englewood Cliffs, NJ: Prentice-Hall, 1988. (book)

[27] H. R. Berenji, “Fuzzy logic controllers,” in An Introduction to Fuzzy Logic Applications in Intelligent Systems, R. R. Yager and L. A. Zadeh, Eds. Boston, MA: Kluwer, 1992, pp. 69-96. (reference to a specific chapter in a work compiled by editors)

[29] S. K. Pal and S. Mitra, “Multilayer perceptron, fuzzy sets, and classification,” IEEE Trans. Neural Networks, vol. 4, pp. 759-771, 1991. (journal article)

· The following format will be used for references to on-line sources. The citation will include the name(s) of the author(s); the title of the work in quotations; the title of the complete work (if applicable) in italics; the date of the document or latest revision (if known) in parentheses; the full URL; and the date of access.

[6] Anonymous, “Electrical engineering program educational objectives,” Electrical Engineering Student Handbook, (2003). http:/ee.uttyler.edu/EEStudentHandbook.pdf (28 June 2004).

Presentation of data with tables and graphs

Small amounts of numerical data may be presented in tabular form, but most data sets are best presented in the form of graphs. There is no hard and fast rule dictating when to use a graph in place of a table; the answer depends on the type of information to be conveyed and how readily apparent will be the meaning of the data to your reader.

Consider the first example below (Table 2). This summarizes the results that might be obtained from measuring the values of samples of 1000Ω resistors from five manufacturers. These data are appropriately displayed in a table since the data set is relatively small and there is no clearly-defined independent variable with numerical values.

Table 2. A sample data set appropriate for display in tabular form.

Manufacturer	n	Mean resistance, Ω	Standard deviation, Ω
A	36	1001.4	22.4
B	29	986.5	19.4
C	41	1024.9	31.5
D	27	1012.6	16.2
E	53	979.3	24.0

The second example is a data set that is not appropriate for display as a table. The large amount of data presented in this table is overwhelming to the reader, and the presentation of these data in tabular form does not help the reader visualize the relationship of the dependent variables (gain and phase) with respect to the independent variable (frequency). The table is also rather long and it is likely that it will extend across a page boundary (as it does here). A table and its caption must not extend across a page boundary.

Table 3. Frequency-response data for a single-pole low-pass filter with cutoff frequency of 1kHz presented in tabular form.

Frequency, Hz	Gain, dB	Phase, deg
100.0	0.0	-5.7
125.9	-0.1	-7.2
158.5	-0.1	-9.0
199.5	-0.2	-11.3
251.2	-0.3	-14.1
316.2	-0.4	-17.5
398.1	-0.6	-21.7
501.2	-1.0	-26.6
631.0	-1.5	-32.3
794.3	-2.1	-38.5
1000	-3.0	-45.0
1259	-4.1	-51.5
1585	-5.5	-57.7
1995	-7.0	-63.4
2512	-8.6	-68.3
3162	-10.4	-72.5
3981	-12.3	-75.9
5012	-14.2	-78.7
6310	-16.1	-81.0
7943	-18.1	-82.8
10000	-20.0	-84.3
12589	-22.0	-85.5
15849	-24.0	-86.4
19953	-26.0	-87.1
25119	-28.0	-87.7
31623	-30.0	-88.2
39811	-32.0	-88.6
50119	-34.0	-88.9
63096	-36.0	-89.1
79433	-38.0	-89.3
100000	-40.0	-89.4

Figure 10 below was prepared with the data of Table 3. This graph has a number of formatting problems:

The independent variable covers a large range and has non-uniform increments. A logarithmic scale for the independent variable would therefore be much more appropriate than a linear scale.
The labels for the horizontal axis are in the plot area itself.
The size of the text is disproportionately large relative to the graph itself.
Two dependent variables representing different quantities are plotted using the same y-axis labels.
The two curves depicted would be indistinguishable when the figure was printed on a monochrome printer.

Fig. 10. An example of a poorly-formatted graph. This graph is typical of an x-y plot produced by Excel using its default settings.

Figure 11 below represents the same data but with much-improved formatting:

A logarithmic scale was used for the abscissa (x-axis).
The x-axis has been moved to the bottom of the graph to keep the labels out of the plot area.
Text has been re-sized and the font changed to match this text.
A secondary y-axis has been added with appropriate independent labels for each axis. Note that the scales of the primary and secondary y-axes have been chosen to utilize the same gridlines.
The two curves are distinct even without the use of color and can be distinguished when printed on a monochrome printer.
Other stylistic improvements have been made:

Gridlines have been added for both axes and are formatted differently (dashed lines) than the axes themselves (solid lines).
The plot area has no colored fill.
The border of the graph has been eliminated.
The legend has been moved from the side of the graph to permit greater width.
The units of the x-axis have been changed from Hz to kHz. This improves the readability of the labels by reducing the number of zeros (e.g., “100” is more-readily understood than “100000”).

Fig. 11. Frequency response plot of a single-pole low-pass filter with cutoff frequency of 1 kHz. The formatting of this figure is much improved compared to Fig. 10.

Figure 11 is visually appealing and fits well into the style of the report. It also conveys its information to the reader in a much clearer fashion than either the raw data (Table 2) or the poorly-formatted graph (Fig. 10).

Significant whitespace often occurs at the bottom of a page when a figure is too large to fit the remaining space on the page. Text following the figure may often be moved to precede the figure to reduce this whitespace. In general, no more than ¼ page of unbroken whitespace should appear on any page.

Guidelines for drawing schematic diagrams

Drawing schematics is an integral part of writing an electrical engineering laboratory report. The preferred tool for drawing schematic diagrams Multisim. Figure 2 above is an example of a schematic diagram drawn with Multisim.

Circuit simulation with electrical simulator software (e.g., PSpice) will frequently be required in laboratory procedures. Schematic diagrams must be included of circuits drawn for simulation as part of a laboratory procedure. Circuit simulation software often includes schematic-capture utilities; schematic diagrams of circuits for simulation drawn with these utilities may be acceptable with proper formatting. Figure 12 was drawn with the schematic-capture package of MicroSim PSpice 8.1 with default display options. This particular figure has problems that make it unacceptable; the most obvious are the visible grid points and the dc voltage annotations automatically added to the display after simulation. The font is also not consistent with the body of this text.

Fig. 12. An example of an unacceptable schematic diagram drawn with a schematic-capture program.

Figure 13 shows an improved version of the same schematic. The grid points and the dc voltage annotations have been removed and the font has been changed to match the surrounding text.

Fig. 13. An improved version of the circuit of Fig. 12.

Incorporation of PSpice simulation results

Simulation results from PSpice should be copied and graphed with Excel. This provides greater flexibility in controlling the format of the graph and also permits graphs to be readily embedded in text. Figure 14 below was drawn from PSpice results copied into Excel. Markers to distinguish the traces when printed on a monochrome printer were not feasible in this graph due to a large number of data points; different weights were assigned to each line to differentiate them.

Fig. 14. Graph produced from PSpice simulation results copied into Excel.

Incorporation of other simulation or analysis results as appendices

Simulation or analysis results will ordinarily be included in the word processor file of the report itself. They may be attached to the report as appendices only if they cannot be embedded in the word processor file. Appendix pages must be numbered consecutively with the main body of the report in any case.

Significant figures and formatting of numerical results

Measurements always involve limited accuracy. Computations with measurement data cannot produce results of greater accuracy than the original data. In general, the measurement with the lesser accuracy will determine the accuracy of the computations:

· In addition and subtraction, the result’s last significant digit will be in the same place as the last significant digit in the less-accurate number. For example, the addition of 4.34 (accurate to the hundredths place) and 0.0446 (accurate to the ten-thousandths place) is 4.38 (accurate to the hundredths place), not 4.3846.

· In multiplication and division, the result will have as many significant digits as the measurement with the lesser number of significant digits. The quotient of 2.2 (two significant digits) and 1.976 (four significant digits) is 1.1 (two significant digits).

Some simple rules for significant digits are given below:

· Non-zero digits are significant, and zeros between non-zero digits are significant. The number 719.3 has four significant digits; 0.10045 has five significant digits.

· Leadings zeros for values less than 0.1 are not significant. For example, there are three significant digits in 0.000309.

· Trailing zeros to the right of the decimal point are significant. A reading of 4.300 has four significant digits.

· Trailing zeros to the left of the decimal point may or may not be significant. For example, a reading of 330W on a wattmeter that resolves readings with an accuracy of ±1W would have three significant digits because the uncertain digit is in the units place. However, a reading of 330W on a wattmeter that resolves readings with an accuracy of ±10W would have two significant digits because the uncertainty is in the tens place.

· Defined constants whose values are known exactly have unlimited significance. For example, if a reading is 4.752 (four significant digits) is multiplied by 2, the result is 9.504 (four significant digits).

Formats of numerical results given in tabular form must be consistent with the number of significant figures in the result.

Some simple rules for expressing numerical quantities are given below:

· Scientific notation is to be written in a format like the following examples: 1.602´10^–19, 6.02´10²³. The “´” symbol is not the letter “x.” The exponent of 10 is expressed as a superscript; the use of a caret (^) to denote exponentiation is not acceptable. For example, 1.602´10^–19 is correct; 1.602´10^–19 is not.

· Exponentiation of the Naperian base e is expressed as a superscript; the use of a caret (^) to denote exponentiation is not acceptable. For example, e^1.632 is acceptable; e^1.632 is not.

· A leading zero is used for decimal numbers less than 1 (e.g., 0.125, –0.0314).

· Negative signs are to be represented by the special En Dash character (–) instead of hyphens (-).

· Values whose units are ratios of other units should be expressed like the following examples. Either form may be used, but the usage should be consistent throughout the report.

o V·s^–1 (volts per second), H·m^–1 (henries per meter), Wb·m^–2 (webers per square meter)

o V/s (volts per second), H/m (henries per meter), Wb/m² (webers per square meter)

· Quantities involving units raised to a power (e.g., cubic meters) are expressed as superscripts; the use of a caret (^) to denote exponentiation is not. For example, 4.03 m³ is acceptable while 4.03 m^3 is not.

Additional resources

A template laboratory report is available. Laboratory reports may be composed by replacing the text of the template with the students’ own text, tables, and figures. A model laboratory report is also available.