Differential phototransducer for automatic reading of information from X-ray films
Abstract Generated abstract
The paper addresses the design of a lightweight photoconverter for automatic scanning and reading of information from X-ray films at relatively high mechanical sweep speeds. It compares photoresistors and photodiodes with respect to sensitivity, inertia, and linearity, and argues for a differential photodiode-based converter because of its linear illuminance-current characteristic. Using an FDK-1 photodiode model, the authors derive load-resistance conditions needed to preserve linear conversion over an optical density range of approximately 0 to 2. The calculation shows that an 11 kOhm load can provide about 5 percent conversion nonlinearity with a maximum output of about 1.1 V, while nonlinearity is substantially lower in the low-brightness region.
Full Text
Proceedings
of the Tomsk Polytechnic Institute named after S. M. Kirov,
holder of the Order of the October Revolution and the Order of the Red Banner of Labor
Vol. 213 1972
DIFFERENTIAL PHOTOCONVERTER
FOR AUTOMATIC READING OF INFORMATION
FROM X-RAY FILMS
A. N. LAFEROV, Yu. V. PILYUGIN
(Presented by the joint scientific seminar of the DSM and MRD sectors).
In systems for automatic reading of information from X-ray films, as the scanning speed increases, with mechanical sweeps the inertia of the entire system increases; therefore the weight of the photodetector becomes one of the principal characteristics of the photoconverter, while the electrical characteristics must still satisfy the assigned tasks. Thus, for a scanning speed
$V = 250 \ \dfrac{\text{cm}}{\text{s}}$
and an entrance-aperture diameter of the collimator $d_o = 1 \ \text{mm}$, the maximum conversion frequency is determined from the following approximate expression
\[ f_{\max} \ll \frac{2v}{d_0} = 5 \ \text{kHz}. \tag{1} \]
For direct methods of measuring blackening density, the system must have a linear conversion characteristic over a wide dynamic range, exceeding values $D > 2$, and, at the same time, the sensitivity of the photodetector must make it possible, without significant amplification, to drive the recorder being used.
Among photoresistors produced by industry, sensitivities on the order of amperes per lumen are not exceptional (SF3-1 has $9 \ \text{A/lm}$), but their great inertia limits their possibilities.
The main drawback of all photoresistors is the nonlinearity of the lux-ampere characteristic over a large part of the conversion range.
Photodiodes, as conversion elements, have a linear lux-ampere characteristic over the entire conversion range; however, since their sensitivity is on the order of $3 \div 20 \ \text{mA/lm}$, additional amplification of the photodiode signals is required.
It is characteristic of photoconverters that the current is a function of illuminance:
\[ I = kE, \tag{2} \]
where $E$ is the illuminance in lm,
$J$ is the current in mA,
$k$ is the sensitivity of the photoconverter.
It follows from this that the photoconverter operates in the current-generator mode; therefore, to preserve the linearity of conversion, the condition must be satisfied
\[ \frac{R_i}{R_{\text{n}}} \gg 1, \tag{3} \]
where \(R\) is the internal resistance of the photoconverter, and \(R_{\mathrm{n}}\) is the load resistance (Fig. 1b).
Let us determine the value \(U_{\text{out}}\) and the ratio \(\dfrac{R_i}{R_{\mathrm{n}}}\) for a differential nonlinearity coefficient \(K_{\mathrm{n}} = 5\%\). From Fig. 1 it is seen that
\[ U_{\text{out}} = U_0 \frac{R_{\mathrm{n}}}{R_{\mathrm{n}} + R_i}. \tag{4} \]
Dividing the numerator and denominator by \(R_{\mathrm{n}}\), we obtain:
whence, for \(\dfrac{R_i}{R_{\mathrm{n}}} \gg 1\), we see that the value \(U_{\text{out}}\) decreases, and the more so the larger \(R_i\) is.
Fig. 1
\[ U_{\text{out}} = U_0 \frac{1}{1 + \dfrac{R_i}{R_{\mathrm{n}}}}, \tag{5} \]
Since
\[ K_{\mathrm{n}}=\frac{U'_{\mathrm{out}}}{(U'_{\mathrm{out}})_{R_{\mathrm{n}}=0}} =\frac{R_{\mathrm{n}}}{R_{\mathrm{n}}+R_i}=0.05, \tag{6} \]
we have
\[ \frac{R_i}{R_{\mathrm{n}}}=17. \]
Using an FDK-1 type diode as the photoconverter (in the photodiode mode) with parameters \(U_0=20\ \mathrm{V}\), \(J=1\ \mu\mathrm{A}\), we obtain the converter current, when the intensity changes by two orders of magnitude, equal to \(1\ \mu\mathrm{A}\cdot 100=0.1\ \mathrm{mA}\); therefore
\[ R_i=\frac{20}{0.1}=200\ \mathrm{k}\Omega. \]
Then from (2)
\[ R_{\mathrm{n}}=\frac{200}{17}=11\ \mathrm{k}\Omega. \]
Thus, a load resistance of \(11\ \mathrm{k}\Omega\) provides a conversion nonlinearity of 5% in the illumination range interval \(D=0\div 2\); in this case the maximum voltage across \(R_{\mathrm{n}}\) is determined from (5):
\[ U_{\mathrm{out}}=U_0\frac{1}{1+\dfrac{R_i}{R_{\mathrm{n}}}} =\frac{20}{18}=1.1\ \mathrm{V}. \]
Substituting into expression (6) the value \(R_{\mathrm{n}}=11\ \mathrm{k}\Omega\),
\[ R_i=\frac{U_0}{I_T}=\frac{20\ \mathrm{V}}{1\ \mu\mathrm{A}}=20\ \mathrm{M}\Omega, \]
we obtain, for the region of low brightnesses (maximum blackening density), the value of the differential nonlinearity coefficient
\[ K_{\mathrm{n}}=0.05\%. \]
The complete differential circuit of the photoconverter using the FDK-1 photodiode is shown in Fig. 1a. Fig. 1b presents the characteristic of the entire converter.