DE2359143A1 - METHOD OF LINEARIZING THE SIGNALS OF A PHOTODETECTOR - Google Patents

Description

Patent attorneys.,. . ,

51 Aachen, the y 7

DIPL-ING. BRUNO SCHMETZ Augustastraße 14-16-fei.

DiPL-ING. WERNER KÖNIG

Eirma SELOOM AB Selective Electronic Co, Lerum, Sweden

Description of patent application '

Method for linearizing the signals of a photodetector

The invention relates to a method for linearizing the signals a I'otodetektor, which the position of an illuminated point a semiconductor plate built into the photodetector indicates the signals being formed by one of the light energy generated in the semiconductor plate] through a resistive layer on one side of the semiconductor plate divided and two pairs of elongated electrodes facing each other in a right-angled square. \ - is conducted and the currents from the electrodes via current (voltage) Measuring devices and a fifth, the semiconductor plate on the side opposite the resistance layer covering translucent electrode, fed back to the resistive layer, with the difference between the measured Currents (voltages) of the two electrodes of an electrode pair to obtain the output signals is formed and a measure for the position of the illuminated point along a right angle represents the axis extending to the electrode in question.

Subject to the discovery of the lateral I ¹ Oto effect and the explanation of the theory of this effect by Professor JT, - Wallmark, Ohalmers Institute of Technology, Gothenburg, Sweden, and by G. Lueovsky, the iOtodetectors of this type have been further developed, and es has come to practical applications, eg in servo systems and for "the purpose of determining the position of an object, a point, in a two-dimensional and right-angled coordinate system in the photocell,

A09823 / Q852

" ² " 2359H3.

However, lateral photocells of this type have the disadvantage. that the relationship between the output signals and the location indicating coordinates is not linear, so that it is often only possible to have only a small area in the center of the To use photocell in which the nicbtlinearity is not too is disturbing. Exceptions are, however, applications in servo systems, where the control is in seal on a particular one Fixed point takes place, and the non-linear dependencies are mostly irrelevant. Attempts have been made linearize the photocell with the help of various correction grids. However, the results were not encouraging. Among other things, this type of correction lowers the sensitivity of the cell. It will refer to publications in this regard by D. Allen, I. Weimar and J. Winslow (Transactions on instruments IRE 1-9, page 336 ff (I960)), which in iDiagraramforai the non-linearity of a photocell of the type affected here with and without a correction grid.

The object of the present invention is now to provide a Specify a method with which it is possible to essentially determine the output signals of a photodetector of the type in question here design linear.

According to the invention, this object is achieved in a method as described at the outset mentioned type solved in that each electrode pair corresponding current (voltage) difference by the sum the same stream is divided and the quotient the linearized Output signal results.

In this way, the high linearity can be reduced over the entire surface of the detector down to about 0.1 fa, whereby it should be noted that in the edge areas of a detector without correction the non-linearity is about 80%. Previous attempts to achieve linearity have nowhere near as good results.

In the following part of the description, an embodiment of the

* (Voltages)

409823/0852

Subject of the invention described with reference to drawings. It shows:· ·:;

! 'ig. 1 a lateral photodetector of the type in which the "inventive linearization method applied should be, and

Pig. 2 is a greatly simplified model of the invention Current distribution in a photodetector according to Fig. 1.

To facilitate understanding of the invention should be next the structure and mode of operation of a lateral photo detector to be discribed. As Fig. 1 shows, 'is a late -. real photodetector built around a rectangular plate 1, which consists of a semiconducting doping material. The plate 1 is transparent on its upper side (oxide layer 2, which forms a fifth electrode, and on its The underside of a thin, homogeneous resistance layer 3 covered by a resistor material. Along the edge of the resistance layer 3 near the corners of the plate -1 are two pairs of each other opposite he diverter electrodes 4 arranged.

If a point 5 on the plate is illuminated, then, generated the Iricütstrom to the. Plate a voltage / current wave 6 on the p-n junction of the plate 1. The voltage / current source 6 can over the G-oldschicbt 2 and the · resistance s layer with the diverter electrodes 4 a closed circuit with ' Close external measuring devices, not shown, whereby four partial currents, one for each electrode 4, are registered en can .- ■ -. "-

An important feature that can be used in the present case is that the current depends on the position of the illuminated point "5 with respect to the four electrodes 4- themselves unevenly distributed over the electrodes. The one on each electrode 4 measured currents thus provide information about the

40982 3/0852

Position of the illuminated point 5. According to previously known procedures, the difference between the currents measured in each pair of electrodes has been calculated and this difference has been regarded as essentially proportional to the position of the point 5 within a limited usable area in the center of the detector.

A full analysis of the type of power distribution in the plate 1 made of semiconducting doping material and in the resistance tube 3 is very complicated. However, it is it is clear that both output signals also change when the light point 5 moves only in one coordinate direction, whereby the already mentioned considerable non-linearity is justified.

Referring to that in Pig. 2, an attempt is now made to describe the linearization method according to the invention, in the greatly simplified model shown. A coordinate system xy, the axial directions of which run normal to the longitudinal directions of the electrodes 4, is arranged above the resistance layer 3. The currents measured in the electrodes 4 intersecting the coordinate axes are _denoted by 1 χ1 and 1 _χ2 ^tz «i _y -j and i». If a point (x; y) is illuminated, then the generated current can be viewed in such a way that it divides itself into these currents which _{flow via the resistors R x1} and R _x2 or R ₁ and R ₂ to the relevant electrode 4 . These resistors are proportional to the distance between the illuminated point 5 and the respective electrode, it is now _ϋ χ1 + R ^ ₂ = R _x and R. + R- = R. where R _x and R are constants, which is easy to see from the drawing. Here i _x1 + i _x2 = i _x> i _y1 + i _y2 = i _y and 1 _χ + i _y = i, ie the total current in x-direction, y-direction and through the pn connection.

From the above-mentioned relationships it follows that the equivalent resistance, e.g. for the parallel connection of the resistors R _x - and R _x2, changes in value considerably if the illuminated point 5 moves from a point near one of the x -electrodes.

409823/08

troden moved to a location near the center of the detector.
lOlglioh i also changes, whereby the difference i ".. - i" does not change linearly with the change in position of the coordinates.

The contradictions R ₁ and R_ ₂ or R ^ and R " ₂ , however, are proportional to the distance of the illuminated point 5 from the respective electrode 4- · The expression (^ _x -J ~ ^ x2 ^ *" ^e * ^som ^ · ^ proportional to the χ-coordinate of the point 5 if the origin of the co-ordinates lies in the center of the square formed by the electrodes 4, which in a corresponding application also applies to the y-coordinate to find the measurable values. According to known laws,

^{Ä 1} X

^B x2

Xt

Since R _x1 + R ^ ₂ = R _x = is constant, one obtains

where this expression is e.g. independent of the y-coordinate.

By calculating the difference between the measured currents
a pair of electrodes and by dividing this difference by the sum of the same currents, a quotient is thus obtained which has a linear relationship to the desired position of the point 5. The above derivation can be applied analogously with regard to the y-coordinate:

3/0852

It should be pointed out here that the above derivation was made on the basis of a very simple model of the actual circumstances. The general applicability of this model can be questioned, but in the present case the approximation must be described as very good, especially since the result actually has an accuracy of $ 0.1.

The applicability of the invention to other similar types of circuits has not been tested; but there are good reasons to believe that the linearization method is applicable in a wide range of systems which make use of current sharing in a conductive sohiobt. The electronic components, which are required to carry out the mathematical processes, are easily found by a supervisor, so that for this reason a description of the elements for office management of the linearization process was not considered necessary .

4 0 9 8 2 3/0852

Claims (1)

2359U3 Claim? Method for linearizing the signals of a photodetector, which indicates the position of an illuminated point of a semiconductor plate built into the photodetector, the Signals are formed in that a lateral current generated by the light energy in the semiconductor plate through a Resistance layer on one side of the semiconductor plate divided and fed to two pairs of elongated, mutually opposite electrodes in a right-angled square and the currents from the electrodes via current. (Tension) lifting devices and a fifth the semiconductor plate on the side opposite the resistance layer covering the transparent electrode of. Resistive layer be fed back, the difference between the measured Strumen (voltages) of the two electrodes of one Pair of electrodes formed to obtain the output signals and a measure of the position of the illuminated point lengthways one running at right angles to the electrode in question Represents the axis, thereby g e k e η η ζ e i c h η e t, that each current (voltage) difference corresponding to an electrode pair is caused by the sum of the same currents (voltages) is divided and the quotient is the linearized output signal results. . 40982 3/08 52