Accelerated Life Cycle Comparison of Millenniata Archival DVD Page Two
Final Report for Millenniata DVD Testing
3. TEST RESULTS
3.1 DRIVE CHARACTERIZATION AND DISC CREATION
As can be seen, the temperature stayed relatively constant between 22.8°C and 24.6°C during the test, well within the 17°C to 30°C limits defined in the Millenniata test plan. The humidity, however, varied greatly and was dependent upon the swamp cooling systems actions. The humidity swing observed during any single day of analysis did not exceed the 30% allowable range defined in the test plan; however, it was consistently above 20%.
The resulting plots from the drive characterization are shown below in Figure 3-1.
As the first two graphs show, the Pioneer DVR-118LBK was most suitable for burning Delkin discs, and the Lite-On iHAP122 and Sony Optiarc AD-7200D were the best for MAM-A discs. The Mitsubishi graph indicates the Sony Optiarc AD-7200D is the best match, the Taiyo Yuden graph shows the Pioneer drive again and the Verbatim graph does not show a significant difference between the Lite-On and Optiarc.
All three drives from each manufacturer used were characterized to determine if they could all be used for burning. The decision based upon this information in Figure 3-2 was that all drives were suitable for use except for the Lite-On B and C drives.
Even after qualifying the drives as described above, the number of discs required to obtain a test sample of 25 discs with PI8 max values below the maximum 175 threshold was not trivial. Certain brands had a successful sample rate of less than 50%. The number of discs that were burned and analyzed in order to provide a test sample of 25 usable discs is reported in Table 3-2. (It should be noted that in order to meet the test schedule, Millenniata agreed to provide preproduction discs. These discs were supplied without the benefit of a production quality assurance program and were recorded in preproduction drives; hence, 29 discs were required to achieve the requisite sample size of 25.)
As can be seen the Delkin and MAM-A brands presented significant problems during the burning process and took more than twice as many attempts as other brands to create a usable sample.
3.2 ENVIRONMENT CONDITIONS
3.2.1 PRE TEST DATA
This irradiance map was taken at room temperature where the pyranometer should have been most accurate. It shows the irradiance levels to be within the Millenniata specification of 1075 W/m2 to 1165 W/m2. The lowest irradiance in the setup was location G1 at 1080 W/m2 and the highest was N4 at 1153 W/m2.
The initial temperature map at a chamber set point of 77.5°C is shown below in Table 3-4.
Because temperature readings were not taken at every location to save time, the map is half actual data and half interpolation. The shaded cells are an average of the surrounding cells, while the cells without shading are actual measurements. What is important to notice about the temperature map is that many of the locations are on the hot side of the 80°C to 87°C range allowed in the test plan. This was remedied by adjusting the chamber set point 2°C down to 75.5°C. The data in Table 3-5 reflects the change this had on disc temperatures.
The average difference in temperature readings as given by Table 3-5 was subtracted from the 77.5°C temperature map to create an adjusted 75.5°C temperature map. This map, Table 3-6, is the predicted temperature of every location during the test and was used to determine the placement of the test discs.
It was anticipated that the Millenniata discs might react differently to the environmental conditions than the conventional dye based media. Table 3-7 shows that this was in fact the case.
The Millenniata discs ran on average 4.4°C hotter than the other test discs. This is because the Millenniata discs are less reflective than their dye-based counterparts. They were, therefore, placed in locations predicted to be the coolest by the temperature map.
3.2.2 ACTUAL RUN CONDITIONS
Figure 3-3 shows the individual disc temperatures quickly diverge as the lamps came up to temperature about 5 minutes into the test. It also shows the humidity reach 80% at the start of dwell.
Figure 3-4 below demonstrates the 85% humidity at the end of dwell and the increase in humidity during the temperature ramp for all three test runs. It also shows the 35°C disc temperatures during the settling period.
The disc temperatures during dwell were, like the ramp conditions, close for all three runs. Figure 3-5 demonstrates the constant dwell temperatures seen in the first run. The graphs of the other runs are included in Appendix B.
The average dwell temperatures as well as the humidity and irradiance for all three runs are contained in Table 3-8.
Figure 3-6 below is a graph of average dwell temperatures for several disc locations.
The blue bars in the graph above represent locations predicted by the temperature map to be usable but on the lower limit of the allowable range. The orange data points are locations predicted to be over 85°C. The green temperatures come from a Millenniata disc that was placed in the lowest predicted temperature location and establishes a lower temperature boundary for the Millenniata discs. Finally, the grey bars are representative of temperatures seen by standard discs during the test. They are readings taken at locations which on the 75.5°C temperature map are within the target range.
An average humidity graph is not shown because, unlike the disc temperatures, the humidity did not reach equilibrium by the beginning of the dwell period. Figure 3-7 shows the humidity during the first run. It demonstrates the same trends as the graphs for the 2nd and 3rd runs included in Appendix B.
As can be seen, the humidity started close to 80% and rose during the first 12 hours of the test until it reached a steady average of 84%. The many spikes and dips seen in the graph have time frames in the 2-4 minute range and were likely caused by the opening and closing of the steam addition solenoid. As a result of the ramp time and humidity dips, the average humidity during the dwell period for all three runs was 83%.
While the temperature and humidity data was very similar for each cycle, the irradiance data demonstrates differences. One element that was present in all three cases was the irradiance reading shooting up to a peak quickly during the temperature ramp up and then dropping down below the target irradiance before the start of the dwell period. Figure 3-8 shows this for the first test run.
This peak and then drop is likely due to the effects of the temperature ramp upon the sensor. Likewise, the average irradiance at dwell is likely different from the irradiance measured at room temperature due to a temperature dependence of the CMP3. What is important to note, and can be seen in Figure 3-9 and Figure 3-10 as well, is that the irradiance levels appeared to vary during each test as much as ±30 W/m2. While this data was taken at the edge of the pyranometer temperature range and may not be an accurate representation of the actual conditions, it indicates that the irradiance levels were not maintained within the Millenniata boundaries of 1075 W/m2 and 1165 W/m2.
In addition to this within test irradiance fluctuation, it is also important to note that while the average irradiance during dwell for the first two runs were close at 1095 W/m2 and 1094 W/m2 the third run had a recorded average below the first two at 1056 W/m2. It is not known whether this was a result of error from the pyranometer or an actual change in the output of the array but it did not appear to have an impact on the disc results.
The actual disc locations during each test run are included in Appendix B. These locations and the 75.5°C temperature were used to calculate the average disc temperature by manufacturer shown in Table 3-9.
As the table demonstrates, random placement of discs resulted in uniform temperatures across the disc brands with the exception of the Millenniata media which ran more than 1°C higher.
The average irradiance was closer than the average temperature. Table 3-10 shows that the irradiance had a spread of only 6 W/m2 between brand averages.
3.3 POST STRESS ANALYSIS
The Delkin discs showed very little change, only a little fading around the inner ring.
The MAM-A disks exhibited hardly any change. No patterns emerged. The only difference was a barely perceptible universal fade.
None of the Millenniata discs experienced any visible change during the test.
The greatest difference was visible on the Mitsubishi discs. Every disc exhibited obvious light and dark rings in an alternating pattern with the greatest fading on the inside radius.
The Taiyo Yuden disks were visibly the least affected of the dye based media. Most discs failed to exhibit any perceptible difference.
The Verbatim discs were second only to the Mitsubishi ones in visible damage. Every disc exhibited an obvious light ring on the inner radius.
The discs which sustained the heaviest visible damage, the Mitsubishi and Verbatim, all failed to be read by the analyzer. Many of the discs were not recognized at all by the drive. Those that were recognized and started to run quickly dropped down to read speeds approaching zero and eventually terminated unsuccessfully. Figure 3-17 and Figure 3-18 show combined baseline and post stress PI8 max graphs for the Mitsubishi and Verbatim discs.
The Taiyo Yuden discs were the only non-archival grade media tested. Most of the discs were completely dead, but 5 of 25 were analyzable as Figure 3-19 shows.
The Delkin discs faired better though one half of them still failed to read. None of these discs were able to meet the ECMA failure criteria as can be seen in Figure 3-20.
Less than a third of the MAM-A discs failed to read, though none of the discs were able to meet the Ecma failure criteria.
Finally, the Millenniata discs did not suffer any damage due to the exposure cycle. As shown in Figure 3-22, there was no data degradation. Every disc performed within the ECMA PI8 Max limit of 280.
Figure 3-23 below compares the 6 disc brands by the average PI8 max error with dead discs included as a 5000 PI8 max reading (the maximum number the analyzer records).
This result is mirrored by the number of dead discs after the test as shown in Figure 3-24.
4. CONCLUSIONS & RECOMMENDATIONS
Copyright© 1996-2017 American-Digital, LLC d/b/a American-Digital.Com and Am-Dig.Com
All Rights Reserved. Designated trademarks and brands are the property of their respective owners.