fiber-optic transmitter and receiver reliability data introduction agilent technologies quality system includes an ongoing reliability monitor program to generate a data base from which this reliability data sheet is published. transmitter reliability the hfbr-0400 series fiber optic transmitters incorporate algaas emitters and double lens optical systems. this design allows the hfbr-14xx transmitters to be driven at a relatively low current level for a specified amount of coupled power into the fiber. because transmitter forward currents are proportional to failure rates, a low current translates into excellent transmitter reliability. to further improve reliability, agilent has implemented a proprietary wafer screening technology that identifies dislocations which can lead to the formation of darkline defects (the primary mechanism for transmitter light output failures). hfbr-0400 series transmitters: hfbr-1402, -1412, -1404, -1414 receivers: HFBR-2402, -2412, -2406, -2416 transmtter reliability data has been calculated using 100 ma forward current and a 100% duty factor. in many actual use conditions, a 50% duty factor and lower forward currents are highly probable. when this is true, transmitter reliability will be substantially better. the degree of improvement can be projected by examining the footnotes below the hfbr-14xx data. reliability prediction model the reliability prediction model used to project failure rate and mean time to failure (mttf) at various temperatures shown in the second table assumes an exponential cumulative failure function (constant failure rate). the arrhenius temperature derating equation is used. agilent technologies assumes no failure mechanism change between stress and use conditions. a conservative activation energy of 0.43 ev was used and is derived from mil- hdbd-217 for hybrid devices. confidence intervals are based upon the chi-squared prediction method associated with exponential distributions.
fiber-optic transmitter reliability data hfbr-1402 hfbr-1412 hfbr-1404 hfbr-1414 high temperature operating life test a. demonstrated performance test equivalent test condition samples device hours failures htol t a = 85 c, i f = 100 ma 880 units 879,500 3 b. failure criteria failure has occurred when the unit fails catastrophically, or when the light output power decreases 3 db. point typical performance performance [1] in time [2] in time (90% confidence) ambient junction mttf [1] fits [3] mttf [2] fits [3] temperature ( c) temperature ( c) (hours) (/10 9 hours) (hours) (/10 9 hours) 85 100 293,000 3411 131,000 7596 80 95 352,000 2844 158,000 6334 75 90 424,000 2360 190,000 5255 70 85 514,000 1947 231,000 4337 65 80 626,000 1598 281,000 3560 60 75 766,000 1305 344,000 2905 55 70 945,000 1059 424,000 2357 50 65 1,172,000 854 526,000 1901 45 60 1,462,000 684 657,000 1523 40 55 1,838,000 544 825,000 1212 35 50 2,326,000 430 1,044,000 958 30 45 2,965,000 337 1,331,000 751 25 40 3,810,000 263 1,711,000 585 c. failure rate prediction for random failures (i f @ 100 ma, 100% duty cycle) notes: 1. the point mttf (representing an esti- mate of the mean point mttf) is the total device hours divided by either the number of failures or unity if there are no failures. 2. 90% confident mttf and failure rate represent the minimum level of reliability performance that is expected from 90% of all samples. this confidence interval is based on the statistics of the distribution of failures. the assumed distribution of failures is exponential. this particular distribution is commonly used in describing failure rates prior to the onset of wear out. refer to mil-std-690 for details of this methodology. 3. 1 fit = 1 failure per 10 9 device hours. 2
fiber-optic link receiver reliability data HFBR-2402 hfbr-2412 high temperature operating life test a. demonstrated performance test equivalent test condition samples device hours failures htol t a = 85 c, v cc = 5.25 v 3,960 2,370,000 1 b. failure criteria failure has occurred when the unit fails catastrophically. one device failed to switch logic states. point typical performance performance [1] in time [2] in time (90% confidence) ambient junction mttf [1] fits [3] mttf [2] fits [3] temperature ( c) temperature ( c) (hours) (/10 9 hours) (hours) (/10 9 hours) 85 100 2,370,000 421 609,000 1,640 80 95 2,880,000 346 742,000 1,340 75 90 3,530,000 282 909,000 1,090 70 85 4,350,000 229 1,120,000 892 65 80 5,400,000 184 1,390,000 719 60 75 6,740,000 148 1,730,000 576 55 70 8,480,000 117 2,180,000 458 50 65 10,700,000 93 2,750,000 362 45 60 13,600,000 73 3,510,000 284 40 55 17,500,000 56 4,520,000 221 35 50 22,700,000 43 5,850,000 170 30 45 29,700,000 33 7,650,000 130 25 40 39,200,000 25 10,000,000 99 c. failure rate prediction, receiver (v cc = 5.25 v) notes: 1. the point mttf (representing an esti- mate of the mean point mttf) is the total device hours divided by either the number of failures or unity if there are no failures. 2. 90% confident mttf and failure rate represent the minimum level of reliability performance that is expected from 90% of all samples. this confidence interval is based on the statistics of the distribution of failures. the assumed distribution of failures is exponential. this particular distribution is commonly used in describing useful life failures. refer to mil-std-690 for details of this methodology. 3. 1 fit = 1 failure per 10 9 device hours. 3
fiber-optic link receiver reliability data hfbr-2406 hfbr-2416 high temperature operating life test a. demonstrated performance test equivalent test condition samples device hours failures htol t a = 85 c, v cc = 5.25 v 2,250 2,250,000 0 b. failure criteria failure has occurred when the unit fails catastrophically. point typical performance performance [1] in time [2] in time (90% confidence) ambient junction mttf [1] fits [3] mttf [2] fits [3] temperature ( c) temperature ( c) (hours) (/10 9 hours) (hours) (/10 9 hours) 85 100 2,250,000 444 977,164 1023 80 95 2,698,516 371 1,171,953 853 75 90 3,252,687 307 1,412,627 708 70 85 3,941,173 254 1,711,633 584 65 80 4,801,432 208 2,085,240 480 60 75 5,882,744 170 2,554,849 391 55 70 7,250,382 138 3,148,808 318 50 65 8,991,407 111 3,904,927 256 45 60 11,222,799 89 4,874,010 205 40 55 14,102,949 71 6,124,846 163 35 50 17,848,023 56 7,751,315 129 30 45 22,755,516 44 9,882,616 101 25 40 29,238,409 34 12,698,107 79 c. failure rate prediction, receiver (v cc = 5.25 v) notes: 1. the point mttf (representing an esti- mate of the mean point mttf) is the total device hours divided by either the number of failures or unity if there are no failures. 2. 90% confident mttf and failure rate represent the minimum level of reliability performance that is expected from 90% of all samples. this confidence interval is based on the statistics of the distribution of failures. the assumed distribution of failures is exponential. this particular distribution is commonly used in describing useful life failures. refer to mil-std-690 for details of this methodology. 3. 1 fit = 1 failure per 10 9 device hours. 4
hfbr-0400 mechanical and environmental test data [1] mil-std-883d units total test name reference test conditions tested failed temperature cycle 1010 500 cycles from -55 to +125 c, 15 minutes 2020 1 at extremes, 5 minutes transfer. [1] hfbr-1414 500 cycles from -55 to +125 c, 15 minutes 2090 0 at extremes, 5 minutes transfer. [1] hfbr-2416 85/85 t a = 85 c, 85% relative humidity, 2140 7 no bias, duration = 1,000 hours. [1] hfbr-1414 t a = 85 c, 85% relative humidity, 2220 6 v cc = 5 volts, duration = 1,000 hours [1] hfbr-2416 high temperature 1008 t a = 125 c800 storage condition b 1000 hours resistance to 2015 three 1 minute immersions. 20 0 solvents brush after solvent immersion. chemical resistance 5 minutes in acetone, methanol, freon tf 20 0 and boiling water vibration variable 2007, 20 g min., 20 to 2000 hz. 20 0 frequency condition b 4, 4 minute cycles each x, y, and z. thermal shock 1011 -55 c to +125 c, 15 cycles 60 0 condition b 5 min. dwell / 10 sec. transfer mechanical shock 2002, 5 blows each x1, x2, y1, y2, z1, z2 60 0 condition b 1500 g, 0.5 msec. pulse. port wear test [2] t a = 25 c 500 connectorings 20 0 less than 1 dbm variation connector side t a = 25 c 1 kg side load 10 0 load [3] less than 1 dbm variation port strength [4] t a = 25 c 6 kg-cm (5.21 inch-lbs), no port damage 10 0 seal-dye penetrant 1014 45 psi, 10 hours 20 0 (zyglo) condition d no leakage into microelectronic cavity solderability 2003 245 c300 esd method 3015 human body model @ 10,000 v 5 0 hfbr-1414 human body model @ 2,000 v 5 0 HFBR-2402 human body model @ 1000 v 5 0 hfbr-2416 notes: see following page. 5
notes: 1. devices were preconditioned with 10 second, 260 c solder dip and 20 cycles, -40 c to 85 c, temperature cycle. 2. coupled power measurements were maximized before and after stress in determining the 1 dbm variation for sma hfbr-0400 products. hfbr-0400 st products do not require this due to the improved coupling design. 3. the connector side load test was only applied to hfbr-0400 sma products. the connector side load testing required that the housing be held so to prevent the leads from yielding. the load was applied through a sma connectored fiber optic cable, perpendicular to the port. the product family is designed to limit cable and ferrule damage due to cable loading. the support and active leads should yield before damage to the cable or connector occurs. if extreme mechanical abuse of the cable/ connector is anticipated please contact agilents application department for suggestions about mechanical strain relief. due to the spring loaded feature of the st connector, hfbr-0400 st products will experience 1 dbm coupled power variation at a side load of less than 1 kg. 4. the port strength test was designed to gauge the concerns with hand tightening the connector to the fiber optic port. the limit is set to a level beyond most reasonable hand fastening loading. 5. package tests are defined as stresses that indicate the environmental strength of the package. units tested indicate the total number of devices taken from the product family. while not all part numbers have been subjected to each stress, worst case products have been included. www.semiconductor.agilent.com data subject to change. copyright ? 1999 agilent technologies, inc. 5968-6620e (11/99)
|
