The method consists of a sample collecting system, a micro positioning display system, a spectrum automatic acquisition and calculation system. The principle of colorimetry of the CIE Lab. was applied to collect, analyze and calculate the information of tested diamond. At the same time, a continuous sequence color with different classes was adopted and used as the standard for reducing the system error. The design of the instrument structure was shown in figure 1, and the experimental apparatus was shown in figure 2. D65 light source with the color temperature of (6 500± 200) K was used. D65 light source is the most commonly used artificial sunlight as standard light sources. In order to meet the test conditions of vertical pavilion facets, a V type groove fixed angle was applied. A powder standard white board specified by standard GB/T 9086 was used and spectral reflectance ratio of which was above 95%, and the absorption of visible light was neutral^[10]. Two standard white boards were placed above a fixed angle of the V type slot, which could be relatively movable, so that each surface of the test sample could be rotated to observe easily. This study used the VC development integration tool of Microsoft Windows desktop application. The color signal acquisition was realized by CCD optical fiber spectrometer with high speed, high resolution and large pixels. The technical parameters of the spectrometer were shown in table 1.

Fig.1

	Figure Option ViewDownloadNew Window
	Fig.1 Instrument working principle diagram 1: CCD; 2: Zoom lens; 3: Spectrometer; 4: Semi transparent and semi reflective mirror; 5: Sample; 6: Light; 7: V type sample slot; 8: Adjustable two dimensional platform; 9: Positioning display; 10: The signal line; 11: Signal processing computer

Fig.2

	Figure Option ViewDownloadNew Window
	Fig.2 Experimental device diagram 1: CCD; 2: Zoom lens; 3: Light source; 4: White board and V type slot; 5: Adjustable two dimensional platform

Table 1

Table 1

Table 1 Technology parameters of the spectrometer Optical platform Ultra low stray light type symmetrical optical path, 75 mm focal length Wavelength range 2001 100 nm Optical resolution 0.0520 nm Sensitivity 470 000 Detector type 2 048 pixel CCD array Signal to noise ratio 300:1

Table 1 Technology parameters of the spectrometer

In order to ensure the same collection of diamond samples with different sizes and the results would not be affected by the size of the diamond, a measurement area was precisely positioned in the same characteristics of the diamond pavilion area, as shown in figure 3. Through a software analysis and calculation system, it could obtain the color measurement coordinates (x, y) value and diamond color level automatic conversion.

Fig.3

	Figure Option ViewDownloadNew Window
	Fig.3 Sample test area (Features of different sizes of diamond in the same area of the Pavilion)

To ensure traceability, 5 sets of master diamonds with the national(China) standard certificate were tested. In which, the first set of stones included D, E, F, …, L color level, total of 9 stones, whose weight were 0.230.26 ct, and the 24 sets had similar size with the first set of master stones. The fifth set of master stones were larger, whose weight were 0.300.38 ct.

The 5th set of master stones were measured, and the diamond color coordinates (x, y) were shown in figure 5. The difference between the x and y index of the adjacent color grade were calculated, and the results were shown in table 2. The digital precision of the X and Y index were 0.000 2.

Fig.5

	Figure Option ViewDownloadNew Window
	Fig.5 Color coordinate chart

Table 2

Table 2

Table 2 Difference between adjacent color levels x y index (D, E) 0.002 3 0.003 2 0.003 95 (E, F) 0.002 1 0.003 2 0.003 77 (F, G) 0.002 1 0.003 3 0.003 77 (G, H) 0.000 9 0.001 7 0.001 90 (H, I) 0.003 4 0.004 7 0.005 82 (I, J) 0.001 2 0.001 6 0.001 89 (J, K) 0.002 1 0.003 0 0.003 66 (K, L) 0.002 4 0.003 6 0.004 31 (L, M) 0.002 8 0.003 4 0.004 33 (M, N) 0.003 1 0.004 4 0.005 42

Table 2 Difference between adjacent color levels

As shown in Figure 5, the method could also be used to test the color span levels of diamonds. The projection point of the test result basically showed linear distribution, and the whole uniformity was good. The method could easily distinguish the diamonds with different color gradings.

For testing the stability and repeatability of this method, we took a diamond of grading D as the example, fixed the same test surface, ten times in a row, and the results were shown in table 3.

Table 3

Table 3

Table 3 D color diamond test data on the same test surface No x y Y index 1 0.317 5 0.335 1 42.5 0.461 63 2 0.317 6 0.335 1 42.6 0.461 68 3 0.317 6 0.335 1 42.5 0.461 73 4 0.317 6 0.335 1 42.5 0.461 68 5 0.317 6 0.335 2 42.5 0.461 81 6 0.317 6 0.335 3 42.5 0.461 84 7 0.317 7 0.335 3 42.5 0.461 91 8 0.317 7 0.335 4 42.6 0.461 98 9 0.317 5 0.335 2 42.9 0.461 69 10 0.317 5 0.335 2 42.9 0.461 72 average 0.317 6 0.335 2 42.6 0.461 77

Table 3 D color diamond test data on the same test surface

The data in Table 3 indicated that the stability and reproducibility of this method were good, and the average value method could further reduce the error.

Selecting No.1 and No.2 sets of mater stones to test, according to the same method as the above, and the results were shown in figure 6.

Fig.6

	Figure Option ViewDownloadNew Window
	Fig.6 2 sets of different master stones test results

From Figure 6, the test results indicated that it was dramatically different from a set of master diamonds to anther set. Although the numerical results of each set were linear, there was a big difference in series of colorimetric stones, especially, the result of the same color level of different sets and uniformity of each set of stones. The same grading stones with different colors like this maybe will bring a big problem to test laboratories. In order to solve the problem, the method provided a suitable idea to avoid the problem, which was to make the numerical value more uniform, as shown in table 4 and figure 7, before and after simulation. Linear fitting formula was y=1.439 2X-0.121 7. This simulation had a guiding significance for the choice of the diamond master set, through the color data, it could be more accurate to select the more uniform diamond master set.

Table 4

Table 4

Table 4 The actual measurement results and the simulation results of equal distance between different colors Actual measurement results Equal interval analog processing data x Difference between upper and lower levels y Difference between upper and lower levels x Difference between upper and lower levels y Difference between upper and lower levels D 0.317 6 / 0.335 1 / D 0.317 6 / 0.335 4 / E 0.319 9 0.002 3 0.338 3 0.003 2 E 0.319 8 0.002 2 0.338 6 0.003 2 F 0.322 0 0.002 1 0.341 5 0.003 2 F 0.322 1 0.002 3 0.341 8 0.003 2 G 0.324 1 0.002 1 0.344 8 0.003 3 G 0.324 3 0.002 2 0.345 1 0.003 3 H 0.325 0 0.000 9 0.346 5 0.001 7 H 0.326 6 0.002 3 0.348 3 0.003 2 I 0.328 4 0.003 4 0.351 2 0.004 7 I 0.328 8 0.002 2 0.351 5 0.003 2 J 0.329 6 0.001 2 0.352 8 0.001 6 J 0.331 0 0.002 2 0.354 7 0.003 2 K 0.331 7 0.002 1 0.355 8 0.003 0 K 0.333 3 0.002 3 0.358 0 0.003 3 L 0.334 1 0.002 4 0.359 4 0.003 6 L 0.335 5 0.002 2 0.361 2 0.003 2 M 0.336 9 0.002 8 0.362 8 0.003 4 M 0.337 8 0.002 3 0.364 4 0.003 2 N 0.340 0 0.003 1 0.367 2 0.004 4 N 0.340 0 0.002 2 0.367 6 0.003 2

Table 4 The actual measurement results and the simulation results of equal distance between different colors

Fig.7

	Figure Option ViewDownloadNew Window
	Fig.7 The results of the actual measurement of the mater stones and the simulation results after equal distance treated