Dynamic linearization system for a radiation gauge

• US 4,064,396 A
• Filed: 12/13/1976
• Issued: 12/20/1977
• Est. Priority Date: 12/13/1976
• Status: Expired due to Term

• Current Assignee: Sangamo Weston, Inc. (Atos SE)
• Original Assignee: Sangamo Weston, Inc. (Atos SE)
• Inventors: Panarello, Joseph A.
• Primary Examiner(s): Morrison, Malcolm A.
• Assistant Examiner(s): Krass, Errol A.
• Application Number:
  US05/750,029
• Time in Patent Office: 372 Days
• Field of Search: 235/151.3, 235/150.5, 235/150.52, 250/252, 250/272, 250/393, 250/308, 340/347 NT, 340/347 AD
• US Class Current: 702/86
• CPC Class Codes: G01B 15/025 by measuring absorption

Abstract

The disclosed linearization system and process converts a high resolution non-linear analog input signal, representative of the thickness of an object, into a high resolution linear analog output signal suitable for use in driving a variety of output devices. The system requires only a small amount of memory for storing pre-calculated non-linear correction coefficients. Prior art linearization systems typically require large memory configurations and/or powerful computers to develop the output signal from the non-linear input. The known systems do not take advantage of the high resolution inherent in the input signal. The disclosed system "channels" the input signal to separate circuit paths so that it may be used directly to; (1) locate an appropriate correction coefficient; (2) develop a correction term after an appropriate correction coefficient is located; and (3) develop a linearized signal having the same high resolution inherent in the input signal. Finally, the disclosed system processes the linearized signal to compensate for the possible errors introduced by radiation source noise. The processed linearized signal is the high resolution linear analog output signal which accurately represents the thickness of the object being gauged.

Claims

1. In a radiation gauge for measuring the thickness of objects, apparatus for converting a thickness representative non-linear analog signal to a thickness representative linear analog signal exhibiting the same degree of resolution inherent in said non-linear signal, comprising:

a. memory means for storing a set of precalculated correction coefficient signals each of which is associated with a preselected range of possible values for said non-linear signal;

b. addressing means responsive to said non-linear signal to address said memory and output the stored coefficient signal associated therewith;

c. correction term generator means, coupled to said memory means, responsive to both said non-linear signal and said output coefficient signal to develop a correction term signal exhibiting the same degree of resolution inherent in said non-linear signal; and

d. summing means, coupled to said generator means, responsive to said non-linear signal and said correction term signal to develop a linearized signal exhibiting the same resolution inherent in said non-linear signal.

2. Apparatus as set forth in claim 1 wherein said memory means comprise a programmable read only memory in which said coefficient signals are stored in digital format.

3. Apparatus as set forth in claim 2 wherein said correction term generator means comprises a digital to analog multiplier operative to convert a digital format coefficient signal to an analog coefficient signal and form the product of said analog coefficient signal with its associated non-linear analog signal.

4. Apparatus as set forth in claim 3 wherein said correction term generator means further comprises scaling circuitry for varying the range of possible correction term signals.

5. Apparatus as set forth in claim 1 wherein said addressing means comprises an analog to digital converter responsive to said non-linear signal to develop a digital memory address signal.

6. Apparatus as set forth in claim 5 wherein said addressing means further comprises an address register coupled to said analog to digital converter for storing said digital address signal and for maintaining a constant memory output over a predetermined time period.

7. Apparatus as set forth in claim 1 further comprising an averaging circuit, coupled to said summing means, for averaging said linearization signal over a period of time to develop an accurate linear analog output signal.

8. In a radiation gauge for measuring the thickness of objects, apparatus for converting a thickness representative non-linear analog signal to a thickness representative linear analog signal exhibiting the same degree of resolution inherent in said non-linear signal, comprising:

a. a programmable read only memory for storing a set of precalculated correction coefficient signals each of which is associated with a preselected range of possible values for said non-linear signal;

b. a digital to analog converter responsive to said non-linear signal to address said memory and output the stored coefficient signal associated therewith;

c. a digital to analog multiplier, coupled to said memory, responsive to both said non-linear signal and said output coefficient signal to develop a correction term signal exhibiting the same degree of resolution inherent in said non-linear signal;

d. scaling circuitry, coupled to said multiplier, for varying the range of possible correction term signals by generating a scaled correction term signal;

e. summing means, coupled to said scaling circuitry, responsive to said non-linear signal and said scaled correction term signal to develop a linearized signal exhibiting the same resolution inherent in said non-linear signal; and

f. averaging circuitry, coupled to said summing means, for averaging said linearized signal over a period of time to develop an accurate linear analog output signal.

9. In a radiation gauge for measuring the thickness of objects a method for converting a thickness representative non-linear analog signal to a thickness representative linear analog signal exhibiting the same degree of resolution inherent in said non-linear signal, comprising the steps of:

a. storing a set of precalculated correction coefficient signals, each of which is associated with a preselected range of possible values for said non-linear signal, in a memory;

b. converting said non-linear signal into an address signal representing the location in said memory of the coefficient signal associated therewith;

c. outputting the coefficient signal stored at the memory location represented by said address signal;

d. developing a correction term signal exhibiting the same degree of resolution inherent in said non-linear signal by forming the product of said non-linear signal and the output coefficient signal; and

e. summing said non-linear signal and said correction term signal to develop a linearized signal exhibiting the same degree of resolution inherent in said non-linear signal.

10. The method as set forth in claim 9 which includes the further step of averaging said linearized signals over a period of time to develop an accurate linear analog output signal.

11. The method as set forth in claim 9 further comprising the step of scaling the correction term signal as a function of known distortion characteristics of the thickness curve for the object being gauged.