Data Plotter - Analytical Chemistry (ACS Publications)

---

tional manpower or decrease in sample input by arrangement with the heads of groups submitting excessive numbers of samples. 2 . -4 ttention to the sample submission dates provides information relative to the effectiveness of the foremen in n ork properly. Khile scheduling yiccial priorities IT ill occur, causing other samples to bp retarded, w r y feiv be allou-ed to become ~ a n i p l ~should s appreciably older than the average backlog age. Keekly throughput tabulations of all lahoratories, both by research unit and hy analyst, are extremely useful. 1. The output figure for each laboratory gives a direct measure of its efficiency. Consideration of this, along with the output of each analyst, will frequently suggest ways of improving efficiency through better scheduling or reassignment of personnel. 2 . The weekly analytical output, in terms of per cent of total, for each research unit tells directly who is getting what portion of the analytical effort. This information should be availablr to

management to aid in proper research assignments consistent \vith the analytical facilities available. CONCLUSIONS

More Effective Laboratory M a n agement. A complete record of in-

put, output, and inventory arranged by unit, laboratory, type of test, or any desired classification may be obtained daily, weekly, or monthly. Such tabulations are of great value in scheduling work mithin a laboratory, in assigning manponer nhere it is needed, and in scheduling pilot plant operation. B substantial increaqe in analytical output has been realized since the business machine system n as put into effect a t Esso Research Laboratories. Reduction in Manpower Assigned t o Data Reporting. Freedoni from

writing report sheets has saved a poition of each analyst's time. I n addition, a reduction of nianpou er assigned t o rerord keeping ha. been achieved.

Increased Speed and Accuracy.

Each answer is reported as it is run and is transcribed only from t h e analyst's work sheet onto the mark sensed card. Availability of Records. Rapid, machine sorting makes it possible t o locate any analysis quickly by any type of identification. Summaries of analytical data can be arranged and tabulated n i t h ease. ACKNOWLEDGMENT

The authors are indebted to representatives of the International Bu.'m e s s llachine Corp. for assistance in setting up the equipment involved and in training personnel in its use. The folloning items of I B l I eyuipnient have been used in the foregoing operations: 4. 26 Keypunch 5 . 82 Sorter

1. 402 Tabulator '1. 519 Reproducer

6. 77 Collator

3 . 552 Interpreter

RECEIVED for revien. May 26, 1955. Accepted 3Iairh 11, 1958.

Data Plotter RALPH H. MULLER and FRANK D. LONADIER' University o f California, Los Alamos Scientific Laboratory, Los Alamos, The incorporation of a multipoint printout recorder, a sliding translatory potentiometer, and other simple electrical equipment found in the laboratory provides a fast efficient way of plotting data received in tabular form. With simple modification of the electrical circuit, the apparatus will perform other useful functions such as squaring, taking logarithms, square rooting, or plotting hyperbolas. Use of the plotter increased speed from 1.5to 2.5-fold over manually plotting the same points.

T

HE automatic recording of analytically important information is now common practice and is used in spectrophotometers, x-ray spectrometer polarographs, thermogravimetric analysis, and vapor phase chromatography. The converse problem of plotting information If-hich is received in tabular form can become formidable. Elegant machines have been developed for this purpose nhich n-ill accept 5 and y values set in on a keyboard or received from punched-card data. Almost lvithout exception, they are 1 Present address, University of Texas, Austin, Tex.

N. M.

elaborate and costly. They are justified whenever the amount of data to be plotted is large or time is important. The multipoint recording potentionieter, n i t h a few simple accessories, can be used as a data plotter for an occasional job. In collating a large amount of nuclear data, the authors had repeated occaqion to use these simple arrangements. The immediate discussion is confined to the plotting of y (or a siniple function thereof) against x, m-here the increments of s are uniform. In this special case one depends upon the uniform motion of the recorder chart arid the precise intermittent action of the print nheel t o print the appropriate value of y a t uniform intervals of variable 5 . Any value of y set in by the operator on a voltage divider nill he

followed and duly printed by the recorder. The manner in which the dependent variable is set in by the operator determines the speed and overall coni-enience of the device. I n push-button keyboards, the scientist is likely to be lower in performance than a coniptometer clerk n h o uses the touch system. On previous occasions, a decade resistance box connected in a high resistance sn aniping circuit has been used. If the total resistance of the decade box is several thousand times snialler than the total resistance of the circuit, the potential drop across its qettings nill be subject to negligible errors. A decade box is inferior in speed of operation to a kej board selector, but it is generally a\railable in a SCALE

1

'

, ,

,,!,,,I

,,

,!

, ,

r-7

,~

I

POINTER

I TRANSLPTORY POTENTiOHETER

!I

13

&

4

POINTER

V GROOVE

Figure 1.

Mechanical arrangement of plotter VOL. 30, NO. 5, MAY 1958

891

selecting any desired number of channels the interval at which points are printed can be set to suit the needs and capability of the operator. After a scale has been selected to suit the range of the data a t hand, the edge of the scale is positioned so that its zero corresponds to the electrical zero of the translatory potentiometer. The shaft and its attached pointer is then drawn out to the maximum value of y which occurs in the data. The Helipot is adjusted to produce full scale deflection on the recorder chart or any desired maximum value. The recorder will not print values as fast as they can be set in by the operator. As the print wheel operates a t stated intervals, it is necessary to have the next value set in before the printing event occurs, after which the pointer can be set to the succeeding value. The audible click in the printwheel mechanism has provided ample

TR4NSLATORY POTENTIOMETER

TO RECORDER

Figure 2.

Schematic of plotter circuit

ABSCISSA

Figure 3. pendence

indication of the event and enables the operator to set in the next value promptly after this event. It is fairly simple to install microswitches in the print wheel system to flash a signal lamp or two if the audible warning is not preferred. USE OF THE PLOTTER

The scheme is superior to moving a pencil over a sheet of graph paper and marking the point directly for two reasons. The data plotter is definitely faster and much less fatiguing: Khen several observers plotted the same data manually and with the plotter, the increase in speed was between 1.5and 2.5-fold. The greatest asset lies in the inherent advantage of performing several functional operations on the

I

Linear plots showing de-

of slope on 300-kw. Helipot

research laboratory whereas the keyboard selector is a specialized machine. APPARATUS

The present scheme uses translatory precision potentiometers (Technology Instrument Corp., Acton, Mass.), in which the uniform winding is mounted in a rectangular housing from one end of which a shaft projects. This shaft can be withdrawn from the housing, thus drawing an internal slider along the winding. The 1200-ohm slider used has a maximum stroke of 11.5 inches. The mechanical arrangement of the plotter is shown in Figure 1. The threaded end of the shaft is fitted to a rectangular block of Lucite. This block is provided with a knob for positioning and a sharp pointer moving over a scale. The lower surface of the block is drilled to receive a steel ball ( 3 / 8 inch in diameter), which rides in a Vgroove of polished brass. After careful alignment, this arrangement permits smooth and fast settings of the slider. The ball and S'-groove arrangement is necessary because the long stroke of almost 1 foot would permit lateral displacement by the operator and interfere with smooth motion. It is convenient to provide removable scales of various ranges to suit particular limits of the dependent variable y. The potentiometer is connected in the ciicuit shown in Figure 2. E is a 1.5volt dry cell and R is a 300-kw. Helipot which provides the scale factor. The translatory potentiometer output is connected to a Leeds Bi: Northrup 10point, 10-mv. recording potentiometer. With a choice of chart speed and by 892

ANALYTICAL CHEMISTRY

Figure 4. system

Example of a special electrical characteristic

of

n I

'

3150

2800

24BO

5W

2100

z 1750

\

cn z

1400

3

0 1050 0 700

350

10

20

30

40

50

BO

7o

CHANNEL NUMBER Figure 5.

---.*.-.--.so

'0

100

Pulse height analysis of a cerium-1 44-praseodymium-1 4 4 source Gamma peaks at 33.7,80.7 and 134 k.e.v.

dependent variable y by modifications of the electrical circuit, such as squaring, square rooting, taking logs, or plotting hyperbolas. Figure 3 is an example of several linear plots. I n each case the slope factor has been set in by the Helipot. I n this and all other examples, the printed points have been enlarged and intensified for photographic reproduction. Figure 4 shows the use of special electrical characteristics of the system. Here the translatory potentiometer has been shorted and the potential between slider and one end of the winding is recorded. The circuit has been described ( I ) . The output is a quadratic function of slider displacement measured from the midpoint. The abscissas represent integral values of n set in by the operator, and the data plotter has printed the corresponding values of the area of the corresponding circles of diameter n. I n this example, the translatory potentiometer squared the values of n and the Helipot in the swamping circuit adjusted the scale factor of 4/n, hence the printed value 4n2 is area, or

-.

?r

For more extensive variations in the functional response, the manufacturers of this translatory potentiometer supply it in the following forms. Up to three separate resistance elements can be installed within a single unit with

individual voltage take-offs actuated by the common shaft. The potentiometers can also be furnished with a wide variety of nonlinear functions and as many as three independent resistance functions can be contained in a single unit. These features afford a wide variety of functions which can be produced as a consequence of the simple motion of the slider and its pointer. The final example of performance of this plotting aid has offered much in saving time and energy. Figure 5 shows a plot of a tape which is printed out by a 100-channel pulse height analyzer. It represents the respective count rate a t each of the 100 channels representing discrete energies. The species lvere the isotopes cerium-144 and praseodymium144. The entire scan of 100 channels was 1 minute and the respective pulses were stored in a magnetic-core memory matrix. The latter is read out automatically and printed on tape. Although the spectrum is presented oscillographically for visual inspection, the printed record is more useful as a permanent record and for more detailed and precise analysis. The plot in Figure 5 was made with the data plotter in 300 seconds with a plotting interval of 3 seconds per point. This speed could be increased somewhat without too much strain on the operator. DISCUSSION

The authors have no illusions about

the superiority of this scheme for setting in data a t high speed. From some experience they are inclined to believe it better than rotary dials, decade boxes, or push-button calculator-type assemblies, unless the operator is skilled in touch typing. The ideal is a voiceactuated selector, which has interested communication experts for a long time. The problem of treating two variables, neither of which is time, can be solved by duplicating the translatory potentiometer systems and presenting the data on an x-y recorder. Although this has been done satisfactorily, no examples are given because the x-y recorder was not equipped with an intermittent print wheel, but with an ordinary pen. Consequently, the traces are continuous rectangular lines, The scheme would be eminently practical if a solenoid pen-lift attachment were added and if this could be actuated momentarily by a foot switch. ACKNOWLEDGMENT

The authors are indebted to Bruce J. Dropesky of this laboratory, for numerous pulse height analyses. LITERATURE CITED

(1) Muller, R. H.,

-4NAL. CHEM.23, 1491 (1951). RECEIVED for revierv August 27, 1956. Accepted January 18, 1958. Work performed under the auspices of the U. S. Atomic Energy Commission.

Punched Card Storage of Gas Chromatographic Data CHARLES F. SPENCER and JULIAN F. JOHNSON California Research Corp., Richmond, Calif.

b The use of IBM punched cards provides a convenient w a y to store, report, and revise chromatographic data, and it eliminates much duplication. Only data for a single compound are punched on a card. To correlate data, standard equipment is used to print tables of information obtained from the cards. gas chromatographic data, accumulated in more or less haphazard fashion, would be of greater value to the analyst if it were available in some systematic form. The problem of recording such data has been discussed by several authors ( I , 2, 6). Standard IBM punched cards provide a method whereby chromatographic data can be conveniently stored and exchanged, and made available in the ANY

form of charts for use in the laboratory. Tables can be easily brought up to date as new information is obtained. DESCRIPTION OF SYSTEM

Each I B M card is punched with data for a single compound. Tables such as Table I can be obtained from the cards with standard equipment which prints the data for each gas chromatographic column. The first three spaces show the column operating temperature. The next four spaces are for the column description code, which serves as an index to a separate record of the complete column description. Three spaces are reserved for the serial number of a specific column. The next column of seven spaces is for boiling point. If desirable, some other item of gas chromatographic data-such as calibra-

tion response, resolution, or relative band width-could replace the boiling point column. The corrected retention volume per gram (2, 4, 5 ) is tabulated in the next five-space column. Partition coefficients by Porter, Deal, and Stross (7), are listed in a column of five spaces. Six spaces contain the code for the compounds listed by the American Petroleum Institute Project 44. This code applies to a large number of hydrocarbons and a limited number of related compounds. I t s use is convenient because it permits machine tabulation of any gas chromatographic data with the physical properties of the compounds available on Project 44 punched cards. Thirty-seven spaces are retained for compound names; in most cases this will permit the complete name to be spelled out. The relative retention VOL. 30, NO. 5, MAY 1958

893