1. CALIBRATION OF CURRENT INTEGRATORS USED WITH IONIZATION CHAMBERS V. Spasić Jokić, I. Župunski, B. Vujičić, Z. Mitrović, V. Vujičić, Lj.Župunski Faculty of Technical Sciences, University of Novi Sad

2. S PECIFIC AIMS Purpose : trace the harmonization of uncertainty evaluation within accreditation framework Uncertainty estimation in accordance with the GUM but it is necessary to establish the method more suitable for the measurements in calibration laboratories Good metrology practice : evaluation of Type B uncertainty is particularly important and requires proper use of the available information is based on experience and skill.

3. D OSIEMETERS BASED ON I ONIZATION CHAMBERS Reading device The typical order of magnitude of ion currents: (10 -6 to 10 -14 ) A

4. R EADOUT CONSIDERATION Voltmeter function: The input resistance of an integrator is greater than 100 TΩ, the input offset current is less than 3fA. Ammeter function: can detect currents as low as 1fA Coulombmeter Function: Current integration and measurement as low as 10fC, has low voltage burden, (less than100μV).Currents as low as 1fA may be detected using this function Low current source + HV IC IC self capacitance = 100 pF

5. C URRENT INTEGRATOR For charge measurement For current measurement Capacitor in the feedback: (10 -5 - 10 -11 ) F (calibrated within 0.1%) Conventional carbon resistors are available in values up to 10 8 

8. C ALIBRATION : WHICH SOLUTION IS THE ‘ BEST ’ Ionization chambers are used together with current integrators and they should be calibrated together Chamber is standard instrument Integrator is standard instrument Calibrated together as the same rank instruments Good reason for separate calibrations is that, one integrator is used with a number of chambers, so it would be inconvenient to calibrate it with every chamber.

9. Assumes user has a calibration factor for exposure N D for the ion chamber/ integrator combination in use But allows IAEA 398 SOLUTION D w,Q = M Q N D,w,Q o k Q,Q o beam quality factor calibration coefficient at Q o corrected instrument reading at Q P elec a factor allowing for separate calibration of the integrtor - here 1

10. Calibration method for a current-measuring feedback-controlled integrator The output impedance of the current source must be large compared to R.

11. Verification of dosimeters used in health care and radiation protection is a legal requirement in Serbia Verification is a subject of accreditation according to SRPS/ISO 17025

12. ISO 17025: G ENERAL REQUIREMENTS FOR THE COMPETENCE OF TESTING AND CALIBRATION LABORATORIES EUROMET Project n. 830, “Comparison of small current sources” Laboratory for metrology at the Faculty of Technical Sciences, University of Novi Sad is accredited in terms of SRPS/ISO 17025 for verification of current integrators

13. C ALIBRATION OF CURRENT INTEGRATORS Suitable direct current source that simulate the output from ionizing radiation detectors. Range: 100 fA - 100 mA (uncertainty better than 0.05 %), depending of chamber type IEC 60731 Calibration: using method of direct measurement

14. S IMPLIFIED CALIBRATION SETUP - standard high impedance DC source Keithley 6220, - various standard resistors and capacitors and - digital multi-meter HP 3450 B

15. A CCREDITED METROLOGICAL LABORATORY FTN UNS

17. C ONCEPT OF UNCERTAINTY ESTIMATION Model function for uncertainty estimation in the calibration procedure for current integrator can be expressed as I x - current read by integrator under the test; δI x – error of reading obtained by integrator under the test due to final resolution; I e – preset current (on current source) derived from the declaration of the manufacturer or calibration certificate

18. C ONCEPT OF UNCERTAINTY ESTIMATION Sensitivity coefficient is derived from expressions Calibration uncertainty for current integrator can be expressed as

19. RESULTS The main part of each calibration procedure is uncertainty estimation and design of uncertainty budget Uncertainty budget obtined during calibration procedure of current integrator type NP 2000 manufactured in OMH, Hungary Preset value: 2 nA Rectangular probability distribution was assumed

20. T HE UNCERTAINTY OF THE CURRENT SOURCE ITSELF Comes from several contributions: Capacitance calibration (5 ppm) Temperature coefficient (4 ppm/K) ac-dc difference Voltage reading (35 ppm) Triggering timing (1 ppm) Leak current compensation (2.10 -5 I + 10 aA)

21. Preliminary uncertainty assessment for the current generated by the source Only type B evaluation has been considered IU B (I)( k=2) 100 fA13 aA 1 pA48 aA 10 pA420 aA 100 pA4.2 fA

22. E STIMATION OF TYPE B UNCERTAINTY ASSUMPTIONS I = 2 nA Lower and Upper limit values: ( I - =I – Δ, I + =I+Δ ) Rectangular distribution: there is 100 % probability that the true value is found in the interval 2 nA I-I- I+I+

23. Step 1. Probability density p(x) for the distribution of current values as p(x)=C for I- Δ  x  I +Δ p(x)= 0 in all other cases E STIMATION OF TYPE B UNCERTAINTY

24. Step 2: Calculation of the best estimated value and variance

25. U NCERTAINTY BUDGET QuantityValue Uncertainty (Type) cici Integrator under the test 2,004 nA 0,00802 nA (A) Uncertainty due to final resolution of reading 100 fA 28,9 fA (B) 1 Preset current on DC source 2 nA 0,001 nA (B) Uncertainty of integrator 0,0081 nA

26. U NCERTAINTY BUDGET OF THE CURRENT TO VOLTAGE CALIBRATION FOR THE 100 P A, 10 P A AND 1 P A Uncertainty component 1 pA [ppm] 10 pA [ppm] 100 pA [ppm] Voltage measurement5010 Resistor value135019070 1/f noise130013013 Current source10 Combined standard uncertainty 190023075 Expanded uncertainty (k=2)3800460150

27. M EASUREMENT CAPABILITIES WITH UNCERTAINTY BUDGET IdV/dt Reference capacitor u 95 source u 95 integrator calibration 100 fA10 mV/s1 pF400 μA/A2 % 100 pA100 mV/s1 pF100 μA/A1 mA/A 1 pA100 mV/s10 pF20 μA/A500 μA/A 10 pA100 mV/s100 pF18 μA/A90 μA/A 100 pA100 mV/s1 nF10 μA/A70 μA/A The expanded uncertainty U with the coverage factor k = 2, corresponding to the 95% confidence level, is often used to represent the overall uncertainty, which relates to the accuracy of the measurement of the quantity Q.

28. CONCLUSION The current uncertainty permits the calibration of even the most accurate commercial meters present on the market. The source is simple, portable and based on low- cost electronics and equipment typically present in most electrical metrology calibration laboratory, where it could be efficiently employed.