View LINEARTECHNOLOGYCORP-LTC5100EUFPBF_6875643.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

ltc5100 1 5100f features applicatio s u descriptio u typical applicatio u n gigabit ethernet and fibre channel transceivers n sff and sfp transceiver modules n proprietary fiber optic links , ltc and lt are registered trademarks of linear technology corporation. n 155mbps to 3.2gbps laser diode driver for vcsels* n 60ps rise and fall times, 10ps deterministic jitter n eye diagram is stable and consistant across modulation range and temperature n 1ma to 12ma modulation current n easy board layout, laser can be remotely located if desired n no input matching or ac coupling components needed n on-chip adc for monitoring critical parameters n digital setup and control with i 2 c tm serial interface n emulation and set-up software available** n operates standalone or with a microprocessor n on-chip dacs eliminate external potentiometers n constant current or automatic power control n first and second order temperature compensation n on-chip temperature sensor n extensive eye safety features n single 3.3v supply n 4mm 4mm qfn package 3.3v, 3.2gbps vcsel driver i 2 c is a trademark of philips electronics n.v. *vertical cavity surface emitting laser **downloadable from www.linear.com figure 1. vcsel transmitter with automatic power control the ltc ? 5100 is a 3.2gbps vcsel driver offering an unprecedented level of integration and high speed perfor- mance. the part incorporates a full range of features to ensure consistently outstanding eye diagrams. the data inputs are ac coupled, eliminating the need for external capacitors. the ltc5100 has a precisely controlled 50 w output that is dc coupled to the laser, allowing arbitrary placement of the ic. no coupling capacitors, ferrite beads or external transistors are needed, simplifying layout, reducing board area and the risk of signal corruption. the unique output stage of the ltc5100 confines the modula- tion current to the ground system, isolating the high speed signal from the power supply to minimize rfi. the ltc5100 supports fully automated production with its extensive monitoring and control features. integrated 10-bit dacs eliminate the need for external potentiometers. an on- board 10-bit adc provides the laser current and voltage, as well as monitor diode current and temperature. status information is available from the i 2 c serial interface for feed- back and statistical process control. an internal digital controller compensates laser tempera- ture drift and provides extensive laser safety features. 3.2gbps electrical eye diagram 1ma/div 50ps/div 5100 ta01 + adc 3.3v scl en sda v dd v ss 24lc00 eeprom in sot-23 package md moda src modb dac dac 10nf 3.2gbps modulator fault warning: potential eye hazard. see ?ye safety information 100 in + in 50 5100 f01 digital controller arbitrary distance serializer
ltc5100 2 5100f v dd , v dd(hs) ............................................................. 4v in + , in C (cml_en = 1) (note 6) peak voltage ........... v dd(hs) C 1.2v to v dd(hs) + 0.3v average voltage...... v dd(hs) C 0.6v to v dd(hs) + 0.3v in + , in C (cml_en = 0) (note 4) .. C0.3v to v dd(hs) + 0.3v cml_en = 0 (note 4) peak difference between in + and in C .............. 2.5v average difference between in + and in C ....... 1.25v moda, modb (transmitter disabled) .... C0.3v to 2.75v moda, modb (transmitter enabled) ............ v dd(hs) C 2.75v to 2.75v en, sda, scl, fault ..................... C0.3v to v dd + 0.3v md, src ................................................... C0.3v to v dd ambient operating temperature range .. C 40 c to 85 c storage temperature range ................ C 65 c to 125 c consult ltc marketing for parts specified with wider operating temperature ranges. absolute axi u rati gs w ww u package/order i for atio uu w (note 1) electrical characteristics the l denotes the specifications which apply over the full operating temperature range, otherwise specifications are at t a = 25 c; v dd = v dd(hs) = 3.3v, i s = 24ma; i m = 12ma (i mpp = 24ma); 49.9 w , 1% resistor from src (pin 14) to moda (pin 11); 50 w , 1% load ac coupled to modb (pin 10); 10nf, 10% capacitor from src (pin 14) to v ss ; cml_en = 0, lpc_en = 1, transmitter enabled, unless otherwise noted. test circuit in figure 5. order part number uf part marking 5100 ltc5100euf t jmax = 125 c, q ja = 37 c/w exposed pad is v ss (pin 17) must be soldered to pcb ground plane 16 15 14 13 5 6 7 8 top view 17 uf package 16-lead (4mm 4mm) plastic qfn 9 10 11 12 4 3 2 1 v ss in + in v ss v ss moda modb v ss v dd en src md fault sda scl v dd(hs) 1/16 of full-scale i s current parameter conditions min typ max units power supply v dd , v dd(hs) operating voltage l 3.135 3.3 3.465 v v dd + v dd(hs) quiescent current, v dd = 3.465v excluding the src pin current (note 2) transmitter disabled, power_down_en = 1 4.5 ma transmitter enabled, is_rng = im_rng = 3 54 ma impp = 24ma high speed data inputs (in + and in C pins) (test circuit, figure 5) input signal amplitude peak-to-peak differential voltage (the single- 500 to 2400 mv p-p ended peak-to-peak voltage is one half the differential voltage) common mode input signal range (note 3) cml_en = 0 (note 4) 0 v dd(hs) v differential input resistance 80 to 120 w common mode input resistance cml_en = 0 (note 5) 50 k w open-circuit voltage cml_en = 0 (note 5) 1.65 v src pin current, i s full-scale i s current is_rng = 0 6 9 ma is_rng = 1 12 18 ma is_rng = 2 18 27 ma is_rng = 3 24 36 ma minimum operating current (note 7) resolution 10 bits src pin voltage range 1.2 v dd Cv 200mv
ltc5100 3 5100f electrical characteristics the l denotes the specifications which apply over the full operating temperature range, otherwise specifications are at t a = 25 c; v dd = v dd(hs) = 3.3v, i s = 24ma; i m = 12ma (i mpp = 24ma); 49.9 w , 1% resistor from src (pin 14) to moda (pin 11); 50 w , 1% load ac coupled to modb (pin 10); 10nf, 10% capacitor from src (pin 14) to v ss ; cml_en = 0, lpc_en = 1, transmitter enabled, unless otherwise noted. test circuit in figure 5. 1/8 of full-scale peak-to-peak modulation current parameter conditions min typ max units laser bias current, i b full-scale current (note 8) is_rng = 0 6 C i m 9 C i m ma is_rng = 1 12 C i m 18 C i m ma is_rng = 2 18 C i m 27 C i m ma is_rng = 3 24 C i m 36 C i m ma absolute accuracy src pin and moda, modb pin currents within 25 % specified voltage ranges resolution 10 bits linear tempco resolution 122 ppm/ c linear tempco range 15625 ppm/ c second order tempco resolution 3.81 ppm/ c 2 second order tempco range 488 ppm/ c 2 temperature stability ib_tc1 = 0, ib_tc2 = 0 500 ppm/ c off-state leakage transmitter disabled, v src = 1.2v 50 m a moda, modb pin current, i m full scale, peak-to-peak modulation current (note 9) im_rng = 0 6 9 ma im_rng = 1 12 18 ma im_rng = 2 18 27 ma im_rng = 3 24 36 ma minimum operating current (note 10) resolution (note 11) 9 bits current stability im_tc1 = 0, im_tc2 = 0 500 ppm/ c voltage range peak transient voltage on moda and modb 1.2 2.7 v absolute accuracy of the modulation current 25 % linear tempco resolution 122 ppm/ c linear tempco range 15625 ppm/ c second order tempco resolution 3.81 ppm/ c 2 second order tempco range 484 ppm/ c 2 maximum bit rate 3.2 gbps modulation current rise and fall times 20% to 80% measured with k28.5 pattern at 60 ps 2.5gbps deterministic jitter, peak-to-peak (note 12) measured with k28.5 pattern at 3.2gbps 10 ps random jitter, rms (note 13) 1ps rms pulse width distortion 10 ps automatic power control (note 14) minimum operating current for the monitor diode 20% of full scale (note 15) monitor diode current temperature stability imd_tc1 = 0, imd_tc2 = 0 500 ppm/ c monitor diode bias voltage (note 16) i md 1600 m a 1.45 v
ltc5100 4 5100f electrical characteristics the l denotes the specifications which apply over the full operating temperature range, otherwise specifications are at t a = 25 c; v dd = v dd(hs) = 3.3v, i s = 24ma; i m = 12ma (i mpp = 24ma); 49.9 w , 1% resistor from src (pin 14) to moda (pin 11); 50 w , 1% load ac coupled to modb (pin 10); 10nf, 10% capacitor from src (pin 14) to v ss ; cml_en = 0, lpc_en = 1, transmitter enabled, unless otherwise noted. test circuit in figure 5. parameter conditions min typ max units automatic power control (note 14) temperature compensation (note 17) linear tempco resolution 254 ? imd_nom/1024 ppm/ c linear tempco range 32300 ? imd_nom/1024 ppm/ c adc resolution 10 bits source current measurement, i s (src pin current) full scale is_rng = 0 9 ma is_rng = 1 18 ma is_rng = 2 27 ma is_rng = 3 36 ma accuracy 3% of full scale 25% of reading average modulation current measurement, i m (note 18) full scale im_rng = 0 9 ma im_rng = 1 18 ma im_rng = 2 27 ma im_rng = 3 36 ma accuracy 3% of full scale 25% of reading laser diode voltage measurement full scale 3.5 v accuracy 150mv 10% of reading monitor diode current measurement (note 19) full scale imd_rng = 0 34 m a imd_rng = 1 136 m a imd_rng = 2 544 m a imd_rng = 3 2176 m a zero scale adc code = 0 1/8 of full scale resolution relative to reading 0.2 % accuracy 25% of reading temperature measurement full scale celsius 239 c sensitivity 0.500 c/lsb termination resistor voltage measurement full scale is_rng = 0 400 mv is_rng = 1 800 mv is_rng = 2 1200 mv is_rng = 3 1600 mv accuracy 30mv 10% of reading safety shutdown, undervoltage lockout (uvlo) undervoltage detection v dd decreasing 2.8 v undervoltage detection hysteresis 150 mv
ltc5100 5 5100f electrical characteristics the l denotes the specifications which apply over the full operating temperature range, otherwise specifications are at t a = 25 c; v dd = v dd(hs) = 3.3v, i s = 24ma; i m = 12ma (i mpp = 24ma); 49.9 w , 1% resistor from src (pin 14) to moda (pin 11); 50 w , 1% load ac coupled to modb (pin 10); 10nf, 10% capacitor from src (pin 14) to v ss ; cml_en = 0, lpc_en = 1, transmitter enabled, unless otherwise noted. test circuit in figure 5. parameter conditions min typ max units bias current limit, i b(limit) set point resolution 7 bits set point range is_rng = 0 9 ma is_rng = 1 18 ma is_rng = 2 27 ma is_rng = 3 36 ma optical power limit automatic power control mode only, apc_en = 1 overpower limit expressed in % over the imd set point 50 % underpower limit expressed in % under the imd set point C50 % safety shutdown response time time from the fault occurance to reduction of 100 m s the laser bias current to 10% of nominal fault output, open-drain mode, flt_drv_mode = 0 output low voltage i ol = 3.3ma 0.4 v output high leakage current v fault = 2.4v 10 m a fault output, open-drain mode with 330 m a internal pull up, flt_drv_mode = 1 output low voltage i ol = 3.3ma 0.4 v output high current v fault = 2.4v C280 m a fault output, open-drain mode with 500 m a internal pull up, flt_drv_mode = 2 output low voltage i ol = 3.3ma 0.4 v output high current v fault = 2.4v C425 m a fault output, complementary drive mode, flt_drv_mode = 3 output high voltage i oh = C3.3ma 2.4 v output low voltage i ol = 3.3ma 0.4 v en input, ib_gain or (apc_gain in apc mode) = 16, im_gain = 4, is_rng = 0, im_rng = 0 input low voltage 0.8 v input high voltage 2v input low current en_polarity = 0 (en active low), v en = 0v C10 m a input high current en_polarity = 0 (en active low), v en = v dd C10 to 10 m a input low current en_polarity = 1 (en active high), v en = 0v C10 to 10 m a input high current en_polarity = 1 (en active high), v en = v dd 10 m a transmit enable time time from active transition on en to 95% of 100 ms nominal laser power and 95% of full modulation. first time transmission is enabled after power on or with rapid_restart_en = 0 transmit re-enable time time from active transition on en to 95% of 1 ms nominal laser power and 95% of full modulation. when transmission is re-enabled after the first time and with rapid_restart_en = 1 transmit disable time time from inactive transition on en to 5% of 10 m s nominal laser power minimum pulse width required to clear 10 m s a latched fault
ltc5100 6 5100f electrical characteristics the l denotes the specifications which apply over the full operating temperature range, otherwise specifications are at t a = 25 c. v dd = v dd(hs) = 3.3v, i s = 24ma; i m = 12ma (i mpp = 24ma); 49.9 w , 1% resistor from src (pin 14) to moda (pin 11); 50 w , 1% load ac coupled to modb (pin 10); 10nf, 10% capacitor from src (pin 14) to v ss ; cml_en = 0, lpc_en = 1, transmitter enabled, unless otherwise noted. test circuit in figure 5. note 1: absolute maximum ratings are those values beyond which the life of the device may be impaired. note 2: the quiescent v dd and v dd(hs) currents refer to the current with zero src pin current (i.e., the laser is operating with zero bias current and zero modulation current). the total power supply current is the quiescent current plus the src pin current, i s , plus any current sinked from in + and in C . note 3: the peak transient voltage at the in + and in C pins must not exceed the range of C300mv to v dd(hs) + 300mv. note 4: when cml_en = 0 (not in cml mode), the termination is 100 w differential with 50k common mode to v dd(hs) /2. note 5: the common mode input resistance is measured relative to v dd(hs) /2 with the inputs tied together. note 6: when cml_en = 1 (cml mode), the termination is nominally 50 w to v dd(hs) on each of the in + and in C pins. note 7: the src pin current can be programmed to near zero in each range, but the recommended minimum operating level is 1/16 of full scale. note 8: the laser bias current is the average current delivered to the laser. it is equal to the src pin current minus the average modulation current at the moda and modb pins, or i b = i s C i m . full scale for the bias current therefore depends on is_rng and the actual modulation current. note 9: the moda and modb pins are connected on-chip. the modulation current refers to the sum of the currents on these pins. i m refers to the total average current at the moda and modb pins. i mpp refers to the total peak-to-peak modulation current at the moda and modb pins. i mpp differs from the laser modulation current, i mod . i mpp splits between the laser and the termination resistor according to i mod = i mpp ? r t /(r t + r ld ), where r t is the value of the termination resistor and r ld is the dynamic resistance of the laser diode. parameter conditions min typ max units scl, sda scl, sda input low voltage, v il C0.5 0.3 ? v v dd scl, sda input high voltage, v ih 0.7 ? v dd +v v dd 0.5 scl, sda input low current (note 21) v sda , v scl = 0.1 ? v dd C100 m a scl, sda input high current (note 21) v sda , v scl = 0.9 ? v dd C100 m a scl, sda output low voltage i ol = 3ma 0 0.4 v hysteresis 280 mv serial interface timing (note 20) scl clock frequency 100 khz hold time (repeated) start condition. after this 4 m s period the first clock pulse is generated low period of the scl clock 4.7 m s high period of the scl clock 4 m s set-up time for a repeated start condition 4.7 m s data hold time 0 3.45 m s data set-up time 250 ns input rise time of both sda and scl signals 1000 ns output fall time of scl and sda from v ih(min) to 300 ns v il(max) with a bus capacitance from 10pf to 400pf set-up time for stop condition 4 m s bus free time between a stop and start condition 4.7 m s capacitive load for each bus line 400 pf noise margin at the low level for each connected 0.1 ? v device (including hysteresis) v dd noise margin at the high level for each connected 0.2 ? v device (including hysteresis) v dd
ltc5100 7 5100f electrical characteristics note 10: the modulation current can be programmed to near zero in each range, but the high speed performance is not guaranteed for currents less than the specified minimum. note 11: the effective resolution of the modulation current is 9 bits because the modulation servo system uses only one-half of the 10-bit adc range. note 12: as defined in ansi x3.230, annex a, deterministic jitter is the peak-to-peak deviation of the 50% crossings of the modulation signal when compared to the ideal time crossings. the specification for the ltc5100 pertains to the electrical modulation signal. the k28.5 pattern is the repeating sequence 00111110101100000101. note 13: random jitter is the standard deviation of the 50% crossings of the electrical modulation signal as measured by an oscilloscope. it is measured with a 1ghz square wave after quadrature subtraction of the random jitter of the pulse generator and oscilloscope. peak-to-peak random jitter is defined as 14 times the rms random jitter. note 14: the ltc5100 digitizes and servo controls the logarithm of the monitor diode current. many of the characteristics of the apc system, such as range and resolution, are determined by the adc. note 15: the minimum practical operating current for the monitor diode is determined by servo settling time considerations. note 16: i md must be less than 25 m a, 100 m a, 400 m a and 1600 m a corresponding to imd_rng = 0, 1, 2, 3. note 17: the temperature coefficients of the monitor diode current depend on the i md setting because of the logarithmic relationship between the set point and the monitor diode current. imd_nom is the digital code setting for the nominal monitor diode current. imd_nom lies between 0 and 1023. note 18: the adc digitizes the average modulation current, which is 50% of the peak-to-peak current for a 50% duty cycle signal. note 19: the ltc5100 adc digitizes the logarithm of the monitor diode current. this implies that the adc resolution is a constant percentage of reading and that the monitor diode current is non-zero when the adc reads zero. see the design notes for further information. note 20: serial interface timing is guaranteed by design from C40 c to 85 c. note 21: the ltc5100 has 100 m a nominal pull-up current sources on the scl and sda pins to eliminate the need for external pull-up resistors when connected to a single eeprom device. the ltc5100 meets the maximum rise time specification of 1000ns with external i 2 c bus capacitances up to 25pf. example: 10pf eeprom + 150mm trace ~ 25pf. note 22: v dd and v dd(hs) must be tied together on the pc board.
ltc5100 8 5100f typical perfor a ce characteristics uw v dd = v dd(hs) = 3.3v, t a = 25 c, cml_en = 0, lpc_en = 1, transmitter enabled, unless otherwise noted. test circuit shown in figure 5. optical eye diagram at 3.2gbps with 850nm vcsel 100 m w/div 50ps/div 5100 g01 emcore mode lc-tosa vcsel 2 15 prbs, 7db extinction ratio, 300 m w avg pwr, 2.4ghz 4th order bessel-thompson lowpass filter optical eye diagram at 2.5gbps with 850nm vcsel 125 m w/div 60ps/div 5100 g02 emcore mode lc-tosa vcsel 2 15 prbs, 10db extinction ratio, 300 m w avg pwr, 1.87ghz 4th order bessel-thompson lowpass filter effect of peaking control on the electrical eye diagram 3ma/div 50ps/div 5100 g03 3.2gbps, 2 23 prbs, im_rng = 2, i mpp = 12ma, peaking = 4, 8, 16, 30 electrical eye diagram at 25 c 3.2gbps, 2 23 prbs, i mpp = 3ma 0.5ma/div 49ps rising 50ps/div 5100 g04 54ps falling im_rng = 0 peaking = 16 electrical eye diagram at 25 c 3.2gbps, 2 23 prbs, i mpp = 12ma 2ma/div 51ps rising 50ps/div 5100 g05 59ps falling im_rng = 2 peaking = 16 electrical eye diagram at 25 c 3.2gbps, 2 23 prbs, i mpp = 24ma 4ma/div 50ps rising 50ps/div 5100 g06 62ps falling im_rng = 3 peaking = 16 electrical eye diagram at C40 c, 3.2gbps, 2 23 prbs, i mpp = 12ma 2ma/div 47ps rising 50ps/div 5100 g07 56ps falling im_rng = 2 peaking = 16 electrical eye diagram at 85 c, 3.2gbps, 2 23 prbs, i mpp = 12ma 2ma/div 57ps rising 50ps/div 5100 g09 67ps falling im_rng = 2 peaking = 16 2ma/div
ltc5100 9 5100f typical perfor a ce characteristics uw v dd = v dd(hs) = 3.3v, t a = 25 c, cml_en = 0, lpc_en = 1, transmitter enabled, unless otherwise noted. test circuit shown in figure 5. supply current vs modulation level (excluding laser current) supply current vs temperature modulator output resistance vs modulation level impp (ma) 0 0 i dd + i dd(hs) (ma) 10 20 30 40 im_rng 0 60 4 81216 5100 g16 20 24 50 im_rng 1 im_rng 2 im_rng 3 transmit disabled and power_down_en = 0 transmit disabled and power_down_en = 1 transmit enabled temperature ( c) ?0 48.1 48.0 47.9 47.8 47.7 47.6 47.5 47.4 20 60 5100 g17 ?0 0 40 80 i dd + i dd(hs) (ma) im_rng = 3 impp = 12ma ib = 0.0ma (includes src pin current required to supply the average modulation current impp (ma) 1 100 output resistance ( ) 1000 10000 10 100 5100 g18 laser bias current vs temperature in ccc mode monitor diode current vs temperature in apc mode temperature ( c) ?0 0.97 i b (normalized) 0.98 0.99 1.00 1.01 20 0 20 40 5100 g20 60 80 normalized to unity at 25 c ib_tc1 = ib_tc2 = 0 temperature ( c) ?0 0.97 i md (normalized) 0.98 0.99 1.00 1.01 20 0 20 40 5100 g21 60 80 normalized to unity at 25 c imd_tc1 = imd_tc2 = 0 laser modulation current vs temperature, im_rng = 1 temperature ( c) ?0 0.97 i mpp (normalized) 0.98 0.99 1.00 1.01 20 0 20 40 5100 g22 60 80 normalized to unity at 25 c im_tc1 = im_tc2 = 0 rise and fall times vs i m at the midpoint of each im_rng deterministic jitter vs i mpp for im_rng = 3 i mpp (ma) 0 50 20% to 80% rise and fall times (ps) 52 54 56 58 60 62 36912 5100 g13 15 im_rng 0 1 1 2 2 3 3 t fall t rise im_rng 0 i mpp (ma) 0 40 20% to 80% rise and fall times (ps) 45 55 60 65 6 12 15 27 5100 g14 50 39 18 21 24 t fall t rise i mpp (ma) 0 deterministic jitter (ps) 5 6 7 24 5100 g15 4 3 1 0 6 12 18 327 9 15 21 2 8 rise and fall times vs i mpp for im_rng = 3
ltc5100 10 5100f typical perfor a ce characteristics uw v dd = v dd(hs) = 3.3v, t a = 25 c, cml_en = 0, lpc_en = 1, transmitter enabled, unless otherwise noted. transmitter enable 5 s/div 5100 g26 v dd fault laser output en transmitter disable 5 s/div 5100 g27 v dd fault laser output en transmitter enable, rapid restart 10 s/div 5100 g28 v dd fault laser output en response to fault fault recovery time 10 s/div 5100 g29 fault fault laser output v md 10ms/div 5100 g30 fault laser output en 5 s wide pulse on en hot plug with en active in ccc mode 10ms/div 5100 g23 v dd scl fault laser output hot plug with en active in apc mode 10ms/div 5100 g24 v dd scl fault laser output start-up with slow ramping supply in apc mode 10ms/div 5100 g25 v dd fault laser output eeprom read scl
ltc5100 11 5100f pi fu ctio s uuu v ss (pins 1, 4, 9, 12, 17): ground for digital, analog and high speed circuitry. these pins are internally connected. connect pins 1, 4, 9 and 12 to the ground plane with minimal trace lengths. place a minimum of four vias (preferably nine vias) to the ground plane in the exposed pad area. most of the high speed modulation current is returned through the exposed pad (pin 17). in + , in C (pins 2, 3): high speed laser modulation inputs. the inputs are differential with internal termination resis- tors. the input amplifier is internally ac coupled. with current mode logic (cml) enabled, the inputs are indepen- dently terminated to v dd(hs) with 50 w resistors. with cml disabled, the inputs provide 100 w differential termi- nation and permit rail-to-rail common mode range. the input pins can be ac coupled with external capacitors. when externally ac coupled, the input pins self-bias to v dd(hs) /2. the cml_en bit selects the termination mode. fault (pin 5): signals one of five safety fault con- ditions: laser overcurrent, overpower, underpower, power supply undervoltage and memory load error. the pin can be programmed active high or active low with the flt_pin_polarity bit. the fault pin can be programmed to four different drive modes with the flt_drv_mode bits. sda, scl (pins 6, 7): serial interface data and clock signals. the pins are open drain with a 100 m a internal pull- up current. an external pull-up resistor can be added to drive larger capacitive loads. v dd(hs) (pin 8): power input for the high speed laser modulation circuitry. filter this pin with a ferrite bead and bypass the pin directly to the ground plane with a 10nf ceramic capacitor. moda, modb (pins 11, 10): high speed laser modula- tion outputs. moda and modb are connected on-chip and driven by an open-drain output transistor. one of these pins should be connected to the laser. the other should be connected to a termination resistor. see the applications information section for details. md (pin 13): monitor diode input for automatic power control of the laser bias current. the md pin allows connection to the cathode or anode of the monitor diode. the md_polarity bit selects the polarity of the monitor diode. src (pin 14): current source for biasing the laser. see the applications information section for details. en (pin 15): transmitter enable and disable input. this input is ttl compatible and can be programmed for active high or active low operation with the en_polarity bit. an internal 10 m a current source disables the transmitter if the en pin becomes disconnected. this safety feature oper- ates whether the en pin is active high or active low. v dd (pin 16): power input for digital and low speed analog circuitry. connect this pin to v dd(hs) (pin 8) with a short trace. no bypassing is needed at the v dd pin if the trace length to the v dd(hs) bypass capacitor is less than 10mm long.
ltc5100 12 5100f block diagra w + + i m(mon) i m v term(mon) v ld moda modb transmit_en transmit_en im_rng is_rng is_rng peaking dac 5 bits i m dac 10 bits i s dac 10 bits i s v dd im_rng is_rng i s(mon) i md(mon) over_current over_pwr under_pwr en_polarity over_pwr under_pwr i m(mon) 11 10 v ss 5100 f02 9 v ss 12 src 14 i s(mon) over_current + + imd_rng md_polarity logarithmic amplifier flt_pin_polarity transmit_en flt_drv_mode fault pin driver ib limit dac 7 bits 10-bit adc i s(mon) i m(mon) v ld i md(mon) v term(mon) temp sensor sel laser power controller register set serial digital interface undervoltage detection under_voltage power limit comparators transmit and fault controller current attenuator 6 7 sda en 16 v dd scl 2 in+ 3 in v ss 100 a data bus 100 a md 13 fault 5 4 17 v ss 1 v dd(hs) 8 50 50 cml_en v dd(hs) 20pf 15 v ss (exposed pad) mem_load_errorr fault figure 2. block diagram
ltc5100 13 5100f figure 3. functional diagramautomatic power control mode fu ctio al diagra s u u w + dac adc + + im set imd set im error peaking im nom im tc2 im tc1 imd tc2 imd tc1 ext temp en t nom t int adc t ext im gain im dac im adc in + in v ss 100 v ss 8 v dd(hs) 2 3 4 1 sda en 16 v dd 6 scl 7 15 11 10 9 12 dac + imd error imd nom apc gain is dac imd adc dac adc adc v dd adc adc v ss 5100 f03 v ss v ld v term i s i s modb moda i m user_adc. data user_adc. data user_adc. data adc + + + 14 src 13 md 5 fault im rng is rng imd rng md polarity temp comp temp comp minimum gain ctrl current attenuator temp sensor i md log amp + 17 v ss (exposed pad)
ltc5100 14 5100f fu ctio al diagra s u u w figure 4. functional diagramconstant current control mode user_adc. data user_adc. data + dac + ib_set im nom im tc2 im tc1 ib tc2 ib tc1 ext temp en t nom t int adc t ext im gain im dac im adc in + in v ss 100 v ss 8 v dd(hs) 2 3 4 1 sda en 16 v dd 6 scl 7 15 11 10 9 12 dac + ib_error im_error ib nom ib gain is rng im rng is dac is adc dac adc v dd adc adc v ss 5100 f04 v ss v ld v term i s i s modb moda i m adc + + + 14 src 13 md 5 fault im rng is rng temp comp temp comp temp sensor (bias current) + peaking + 17 v ss (exposed pad) adc +
ltc5100 15 5100f test circuit figure 5. test circuit equivale t i put a d output circuits u u u 15 en v dd en_polarity 0 en active low 10 a v dd v ss 10 a 5100 f06 1 en active high ltc5100 100 a 5100 f07 sda, scl 6, 7 v dd ltc5100 v dd figure 6. equivalent circuit for the en pin figure 7. equivalent circuit for the sda and scl pins 6 3.3ma driver complementary drive fault 5100 f08 250 a pull-up 250 a 3.3ma driver 400 a pull-up 400 a v dd ltc5100 v dd 13 v dd m2 md_polarity md 5100 f09 imd imd m1 1 0 ltc5100 v dd figure 8. equivalent circuit for the fault pin. all switches are open in open-drain mode figure 9. equivalent circuit for the md pin fault sda scl v dd(hs) v dd v dd v ss in + in v ss v ss moda modb v ss en src 1.8v power source ltc5100 md 13 12 10nf z o = 50 microwave blocking capacitor to scope 50 resistors: 0402 surface mount capacitors: 0402 surface mount, x7r dielectric 11 10 9 1 2 3 4 14 15 16 8 10nf 5100 f05 7 6 5 10nf z o = 50 z o = 50 from bert
ltc5100 16 5100f equivale t i put a d output circuits u u u 2 in + 3 in 50 50k v dd(hs) to input amplifier 50k 5100 f10 25k 20pf r on @ 3 cml_en 50 ltc5100 v dd(hs) v dd(hs) 14 v dd m2 25k src 11 moda 10 modb 5100 f11 1m m1 ltc5100 v dd(hs) v dd(hs) figure 10. equivalent circuit for the in + and in C pins figure 11. equivalent circuit for the src, moda and modb pins operatio u overview (refer to figure 1 and the block diagram in figure 2) the ltc5100 is optimized to drive common cathode vcsels in high speed fiber optic transceivers. the chip incorporates several features that make it very compact and easy-to-use while delivering exceptional high speed performance. only a capacitor, a resistor and a small eeprom (excluding laser diode and power supply filter- ing) are needed to build a complete fiber optic transmitter. digital control over the i 2 c serial interface allows fully automated laser setup to improve manufacturing effi- ciency. the ltc5100s extensive set of eye safety features meet gbic and sff requirements but go beyond the standards with open-pin protection, redundant transmit- ter enable controls and other interlocks. 10-bit integrated dacs set laser bias and modulation levels, eliminating the cost and space of digital potentiom- eters. a multiplexed adc allows monitoring of tempera- ture and laser operating conditions in production or field operation. laser bias and modulation currents are digi- tally temperature compensated to second order for tight control of average power and extinction ratio. the ltc5100 provides both constant current and automatic power control of the laser bias current. in automatic power control mode, special circuitry maintains constant set- tling time in spite of variations in the laser slope efficiency and monitor diode response characteristics. the high speed inputs of the ltc5100 are internally terminated in 50 w and internally ac coupled, eliminating all external components at the inputs. the modulation output is dc coupled to the laser and presents a high quality resistive drive impedance to deliver very fast and clean eye diagrams in spite of laser impedance variations. the modulation output is capable of driving significant lengths of transmission line, allowing the ltc5100 to be placed at an arbitrary distance from the laser. this feature allows for packaging flexibility within the module. the ltc5100 minimizes electromagnetic interference (emi) with several architectural features. the unique design of the driver output forces the high speed modulation current to circulate only in the laser and ground system. the high speed amplifier chain and the digital circuitry are internally filtered and decoupled to further reduce power supply noise generation.
ltc5100 17 5100f laser bias and modulation modulator architecture the ltc5100 drives common cathode lasers using a method called shunt switching. as shown in figure 12, shunt switching involves sourcing dc current into the laser diode and shunting part of that current with a high speed current switch to produce the required modulation. the src pin provides the dc current and the moda, modb pins (which are connected on chip) provide the high speed modulation current. this technique results in a very fast, single-ended driver that confines the high speed modulation current to the laser and ground system. the ltc5100 actually uses a modified shunt switching scheme in which the source current is delivered through a termination resistor, r t , that is bypassed to ground with a large capacitor. the resistor brings three advan- tages to the modulation stage. first, it gives the modulator a precise resistive output impedance to damp ringing and absorb reflections from the laser. second, the resistor isolates the capacitance of the src pin from the high speed signal path, further improving modulation speed. third, the resistor and capacitor heavily filter the high speed output signal so that it does not modulate the power supply and cause radiation or interference. on-chip decoupling of the high speed amplifiers further reduces power supply noise generation. operatio u terminology and basic calculations figure 12 through figure 16 define terminology that is used throughout this data sheet. the current delivered by the src pin is called i s . the average modulation current delivered by the chip at the moda, modb pins is called i m . the laser bias current, i b , is defined as the average current in the laser. i b is the difference between the source current and average modulation current. the peak-to-peak modulation current delivered by the chip is called i mpp . i mpp is twice the value of i m because the high speed data is assumed to have a 50% duty cycle. the peak-to-peak modulation current is divided between the termination resistor and the laser. the peak-to-peak modulation amplitude in the laser is called i mod . the relationship between i mpp and i mod depends on the relative values of the termination resistor and the laser dynamic resistance. figure 12. simplified laser bias and modulation circuit v ss i m i s = i b + i m i b i mod i s v dd moda, modb 11, 10 src ltc5100 r t 50 typ c t 10nf i mpp = 2 ?i m 3.2gbps modulator 14 figure 13. components of the ltc5100 source and modulation currents (the laser bias current is also shown) 0 i m i s i m i b i mpp 5100 f13 the relationships between the source, bias, and modula- tion currents are as follows. i b = i s C i m (1) i mpp = 2 ? i m (2) i r rr i mod t tld mpp = + () (3) where r t is the termination resistor value. r ld is the dynamic resistance of the laser, defined in figure 15. the expression for i b in equation 1 shows that the maxi- mum achievable laser bias current is a function of the maximum source current, i s , and the average modulation
ltc5100 18 5100f current, i m . the maximum value of i s is given in the electrical characteristics and the value of i m depends on the laser characteristics and the termination resistor value. the logic 1 and 0 current levels in the laser are given by: ii i b mod 1 2 =+ (4) ii i b mod 0 2 = (5) operatio u figure 14. components of the laser bias and modulation currents 0 i b i th i mod 5100 f14 the power levels corresponding to i1 and i0 are p1 and p0, as shown in figure 16. p1 = h (i1 C i th ) (6) p0 = h (i0 C i th ) (7) where h is the slope efficiency and ith is the laser threshold current, defined in figure 16. the average optical power and extinction ratio are given by: p pp avg = + 10 2 (8) er p p = 1 0 (9) the average voltage on the laser diode relative to ground is v ld (see figure 12 and figure 15). the voltage on the src pin is: v s = v ld + i s ? r t (10) = v ld + (i b + i m ) ? r t the value v s is important because v s must not exceed the limits given in the electrical characteristics. v ld v i0 i b i ld 5100 f15 i1 r ld figure 15. approximate vi curve for a laser diode figure 16. approximate li curve for a laser diode p1 p avg p0 l i0 i th i b i ld i mod 5100 f15 i1 h the voltage across the termination resistor is: v term = v src C v moda (11) = is ? r t the ltc5100 can digitize the voltage across the termina- tion resistor using the on-chip adc, which can give a more accurate measurement of is than that given by digitizing the current internally. see the electrical characteristics for details. temperature compensation the ltc5100 digitally compensates the temperature drift of the laser bias current, laser modulation current and
ltc5100 19 5100f monitor photodiode current. in each case the fundamental calculation is the same. the ltc5100s digital controller multiplies the nominal value of the quantity (i b , i m or i md ) by a quadratic function of temperature. temperature mea- surements are supplied either by an on-chip temperature sensor or by an external microprocessor, according to the setting of ext_temp_en. the general temperature com- pensation formula is: i = i_nom ? (tc2 ? 2 C18 ? d t 2 + tc1 ? 2 C13 ? d t + 1) (12) where i is the digital representation of the laser bias current, modulation current or monitor diode current (i b , i m or i md ). when using the internal temperature sensor (ext_temp_en = 0), the temperature measurements are taken by the on- chip adc, and d t is the change in the ltc5100 die tem- perature relative to a user defined nominal temperature: d t = t_int_adc C t_nom (13) when using an external temperature source (ext_temp_en = 1), the temperature measurements are provided in digital form by a microprocessor or host computer and d t is the change in temperature relative to a user defined nominal temperature: d t = t_ext Ct_nom (14) t_int_adc, t_ext, and t_nom are 10-bit, unsigned numbers scaled at 0.5k/lsb. the maximum temperature that can be represented is therefore 2 10 ? 0.5 k = 512 k or 239 c. tc1 and tc2 are the first and second order temperature coefficients. they correspond to the registers im_tc1 and im_tc2 for modulation current, ib_tc1 and ib_tc2 for bias current and imd_tc1 and imd_tc2 for monitor diode current. in each case tc1 and tc2 are 8-bit signed numbers in twos complement format. the range of the temperature coefficients is therefore C128 to +127. when tc1 is multiplied by its weighting coefficient of 2 C13 in equation 12, the effective value of the first order tempera- ture coefficient is 122ppm/ c per lsb. the full-scale range is approximately 15500 ppm/ c. when tc2 is multiplied by its weighting coefficient of 2 C18 in equation 12, the effective value of the second order temperature coefficient is 3.81ppm/ c 2 per lsb. the full-scale range is approximately 484 ppm/ c 2 . note that equation 12 is applied to the digital representa- tion of the currents, not the physical current themselves. this is a particularly important point where monitor diode current is concerned, because the digital representation of the monitor diode current is the logarithm of the current. thus the temperature compensation is of the logarithm of the monitor diode current and not the current itself. notation used for registers and bit fields the ltc5100 has a large set of registers, many of which are subdivided into fields of bits. register names are given in all capitals (sys_config) and bit fields are given in mixed case (apc_en). for example, the bit that enables automatic power control mode is contained in the system configuration register. this bit is denoted by: sys_config.apc_en in many cases this bit field will simply be referred to as apc_en. the functional diagrams of figure 3 and figure 4 show registers and bit fields within registers between horizontal bars. for example, the data field in the adc register is shown as: user_adc.data a write operation to this register is shown as: user_adc.data a register read operation is shown as: peaking range selection for the source and modulation currents the source and modulation currents each have four ranges of operation to optimize adc and dac resolution as well as high frequency performance. the source current range is controlled by two bits called is_rng. similarly, the modu- lation current range is controlled by two bits called im_rng. the maximum current that can be delivered is proportional to the range, so the current output is 1, 2, 3 or 4 times the typical base value of 9ma for the source current and 4.5ma for the average modulation current or 9ma peak- to-peak. operatio u
ltc5100 20 5100f figure 17 depicts the current ranges for the source cur- rent. the guaranteed full scale is 6ma per range. the minimum operating level should be limited to 1/16 of full scale to avoid the coarse relative quantization seen in any adc or dac when operated at low levels. the source range, is_rng, should be selected as low as possible such that the source current, i s , stays within the guaranteed current limits over temperature, considering the laser temperature characteristics. from equation 1 we can see that the source current is the sum of the laser bias and the average modulation currents: i s = i b + i m (15) is_rng should be chosen to support the total current required for laser bias and modulation, taking temperature changes in i b and i m into account. figure 18 depicts the current ranges for the average modulation current. this is the average modulation cur- rent at the moda and modb pins of the chip (recall that the moda and modb pins are connected on-chip). the peak- to-peak modulation at the pins of the chip is twice the average. guaranteed full scale is 3ma average or 6ma pp per range. the minimum operating level should be limited to 1/8 of full scale to preserve the quality of the eye diagram. operating below 1/8 full scale causes increased overshoot and undershoot. the modulation range, im_rng, should be selected as low as possible such that the modulation current, i m , stays within the guaranteed cur- rent limits over temperature. the modulation current varies over temperature to compensate the loss in slope efficiency typical of most vcsels. therefore, the choice of im_rng should take temperature changes into account. high speed aspects of the modulation output the ltc5100 modulation output presents a resistive drive impedance with very low reflection coefficient. this output design suppresses ringing and reflections to maintain the quality of the eye diagrams in spite of laser impedance variations. the reflection coefficient is sufficiently low that the ltc5100 can drive the laser over an arbitrary length of transmission line, as shown in figure 19. a well designed transmission line stretching the entire length of a typical transceiver module goes virtually unnoticed in this sys- tem. the only practical limitation on interconnect length to the laser is high frequency line loss. operatio u figure 17. ranges for the source current figure 18. ranges for the modulation current figure 19. high speed details of the modulation output 0 9 is_rng = 0 5100 f17 4.5 18 27 36 is_rng = 1 recommended minimum is 1/16 of full scale is_rng = 2 is_rng = 3 i s (ma) 0 4.5 im_rng = 0 5100 f18 9 13.5 18 im_rng = 1 recommended minimum is 1/8 of full scale im_rng = 2 im_rng = 3 i m (ma) v ss i m l bwb m1 i s = i b + i m z o = r t i b i s v dd src moda modb ltc5100 r t 50 typ c t 10nf 3.2gbps modulator transmission line 14 11 10 c1 l bwa
ltc5100 21 5100f figure 19 shows how the ltc5100 achieves a low reflec- tion coefficient. the unavoidable capacitance of the high speed driver transistor, bond pads and esd protection circuitry (c1) is compensated by the inductance of the bond wires (l bwa and l bwb ). the high speed behavior of the circuit in figure 19 can be understood in greater detail by examining the simplified circuit in figure 20. in figure 20 the switched current source (m1 in figure 19) launches a current step (1) toward the termination resistor (2a) and toward the trans- mission line (2b) connected to the laser. the laser is typically mismatched to the line impedance and reflects a portion of the incident wave (3) back toward the modb pin. there it encounters an l-c-l structure composed of the bond wires and driver capacitance. this structure is carefully designed as a lumped element approximation to the transmission line impedance. it therefore transmits wave (3) through the ic package without reflecting energy back toward the laser. the traveling wave passes through the chip largely unimpeded (4) and is absorbed by the matched termination resistor, r t . the matched termination is provided by the termination resistor, r t , decoupled by capacitor c t . c t forms an ac short across the entire frequency range contained in the modulation data. the termination resistor, r t , need not be 50 w . 50 w is best for electrical testing because it matches the impedance of most high frequency instruments. r t can be made smaller, 35 w , for example, to more closely match a laser with low dynamic impedance or to allow more voltage headroom at the src pin. this may be necessary for lasers that run at high voltages or high bias currents. r t can be made larger, 70 w for example, to more closely match a laser with high dynamic impedance or if a narrow, high impedance pc board trace is needed to connect to the laser. figure 21 shows that the high speed modulation current is confined to the ground system, laser and back termination network. no high speed current circulates in the power supply where it could cause radiation and interference problems. high speed data inputs the high speed data inputs, in + and in C , are internally terminated in 50 w and internally ac coupled, eliminating the need for external termination resistors and ac cou- pling capacitors. figure 10 shows the equivalent circuit for the high speed data pins. by default, the high speed data inputs are terminated differentially with 100 w for compatibility with lvds, pecl and similar differential signaling standards (cml_en = 0). alternately, the inputs can be programmed for 50 w single-ended termination to the power supply for biasing a current mode logic (cml) driver. to select cml compatibility, program cml_en to 1. although internally ac coupled, the inputs are biased with high valued resistors (50k equivalent) to v dd(hs) /2, so the ltc5100 remains compatible with external ac coupling capacitors. when externally ac coupled, the inputs self- bias to approximately v dd(hs) /2. internal ac coupling gives the ltc5100 rail-to-rail input common mode capability. the inputs can be driven as much as 300mv beyond the rail during peak excursions. the ac coupling circuit is a distributed highpass filter with operatio u figure 21. high speed current flow in the modulation output 11 moda 10 modb transmission line z o = r t 5100 f20 c1 r t 50 typ l bwa l bwb 1 2b 2a 4 3 v ss m1 exposed pad v dd src moda modb 5100 f21 ltc5100 10nf no high speed current 50 3.2gbps modulator 14 11 12 10 9 figure 20. wave propagation in the laser interconnect
ltc5100 22 5100f approximately second order characteristics. the design maximizes the flatness of the step response over extended periods, giving optimal performance during long strings of ones or zeros in the data. modulation current control in apc and ccc modes the ltc5100 controls the modulation current with a digital servo control loop using feedback from the on-chip adc. figure 3 and figure 4 are functional diagrams of the ltc5100 operating in automatic power control (apc) mode and constant current control (ccc) modes, respec- tively. these diagrams show the organization and opera- tion of the servo control loops for laser bias and laser modulation. either diagram can be used to understand the modulation current control loop. servo control the average modulation current is controlled by a digital servo loop (shown in the lower half of figure 3). the nominal modulation current, im_nom, is multiplied by a temperature compensation factor, producing a 10-bit digital set point value, im_set. im_set is the target value for average modulation current. the adc digitizes the average modulation current, producing a 10-bit value im_adc. the difference between the target value and the actual value produces the servo loop error signal, im_error. im_error is multiplied by a constant, im_gain, to set the loop gain. the result is integrated in a digital accumulator and applied to a 10-bit dac, increasing or decreasing the modulation amplitude as required to drive the loop error to zero. the servo loop adjusts the modulation amplitude every four milliseconds, producing 250 servo iterations per second. the modulation servo loop operates on the average modu- lation current, which is one-half of the peak-to-peak value for a 50% duty cycle signal. the analog electronics in the high speed modulator ensure that controlling the average modulation current is equivalent to controlling the peak- to-peak current. the adc input for average modulation current is scaled such that code 512 is the nominal full-scale value, corre- sponding to 4.5ma per range. thus, if im_rng = 0 and im = 4.5ma, the adc digitizes code 512. the control system for the modulation current effectively has 9-bit resolution, because at most one-half of the 10-bit adc range is utilized. this provision maximizes the compliance voltage range of the modulation output. the difference equation for the modulation servo loop is: im_ im_ im_ im_ ( ) im_ im_ im_ im_ adc adc gain error adc gain set adc nn nn =+ =+ () -- 1 11 8 16 8 im_gain is a 3-bit digital value, so the scaling factor, im_gain/8, takes on the discrete values 0, 1/8, 2/8, , 7/8. if im_gain = 4, then im_gain/8 = 0.5 and the error in the control loop is cut in half with each servo iteration. in this case the step response of the loop is given by: im_ im_ im_ adc set gain n n = ? ? ? ? ? ? 11 8 (17) the step response has the familiar exponential settling characteristic of a first order system. the step response is shown in figure 22 for im_gain = 4. the remaining error is reduced by one-half with each servo iteration. in seven iterations, or about 28ms, the modulation current settles to under 1% in this example. the measured step response, including the modulation envelope, is shown in the typical performance characteristics. operatio u figure 22. step response of the average modulation current for im_gain = 4 im_adc 0 4 8 12 16 20 24 28 32 5100 f22 2 1 345678 time (ms) servo iterations im_set
ltc5100 23 5100f reducing im_gain slows the settling time and increasing im_gain speeds the settling time. for example, with im_gain = 1, the residual loop error is cut by 1/8 with each servo iteration. in this case it would take 35 servo iterations (about 140ms) to settle to 1%. with im_gain = 7, the residual servo loop error is cut by 7/8 with each servo iteration. in this case it would take only three servo iterations (about 12ms) to settle to 1%, but the servo loop will tend to hunt or oscillate at a low level with such a high loop gain. temperature compensation the set point value for the modulation current, im_set in figure 3 and figure 4, changes with temperature to com- pensate the temperature dependence of the laser diodes slope efficiency. temperature measurements are supplied either by an on-chip temperature sensor or by an external microprocessor, according to the setting of ext_temp_en. the temperature compensated expression for im_set is given by: im_ im_ im_ im_ set nom tc t tc t = d +d+ ? ? ? ? 22 12 1 18 2 13 (18) im_tc1 and im_tc2 are the first and second order tempera- ture coefficients for the modulation current. laser bias current control in apc mode figure 3 is a functional diagram of the ltc5100 operating in automatic power control (apc) mode. in apc mode, the ltc5100 servo controls the average optical power with feedback from a monitor photodiode. setting apc_en = 1 selects this mode. in apc mode the monitor diode current can be temperature compensated with first and second order temperature coefficients. figure 9 shows an equivalent circuit for the md pin and figure 23 shows details of the monitor diode circuit. the md_polarity bit selects whether the monitor diode sources or sinks current from the md pin. a programmable attenu- ator and logarithmic amplifier permit a very wide range of monitor diode currents spanning 4.25 m a to 2176 m a (typi- cal) with constant 0.2% set point resolution. the attenu- ator divides the monitor diode current by 1, 4, 16 or 64 depending on the value of imd_rng. two bits called imd_rng control the attenuator setting, selecting a full scale current range of 34, 136, 544 or 2176 m a typical. a 5khz lowpass filter provides antialiasing and limits noise. the logarithmic amplifier compresses the dynamic range of the monitor diode current and plays a role in maintaining constant and predictable settling times regardless of the photodiode characteristics. range selection f igure 24 depicts the current ranges for the monitor diode current. the full-scale range of the monitor diode current is 34 m a ? 4 imd_rng typical where imd_rng = 0, 1, 2 or 3. the minimum operating level should be limited to 20% of full scale to ensure adequate settling time of the optical power output of the laser. the range should be selected so that the monitor diode current stays within the guaranteed current limits over temperature. operatio u figure 23. detail of the monitor photodiode circuit 13 log amp 10-bit adc ltc5100 md v dd 5khz lowpass filter attenuator 1, 4, 16, 64 polarity control md_polarity md_polarity = 1 md_polarity = 0 5100 f32 imd_rng imd_adc
ltc5100 24 5100f the src pin current range, is_rng, should be chosen so that the src pin can supply the required bias current over temperature. see the section titled range selection for the source and modulation currents. servo control the average optical power is controlled by a digital servo loop shown in the upper half of figure 3. the loop sets and controls the logarithm of the monitor diode current. the logarithm of the nominal monitor diode current, imd_nom, is multiplied by a temperature compensation factor, pro- ducing a 10-bit digital set point value, imd_set. imd_set is therefore the temperature compensated logarithm of the target value for monitor diode current. the adc digitizes the logarithm of the monitor diode current, producing a 10-bit value called imd_adc. the difference between the target value and the actual value produces the servo loop error signal, imd_error. imd_error is multiplied by a constant, apc_gain, to set the loop gain. imd_error is also multiplied by the set point value of the modulation current to further stabilize the servo dynamics, as explained below. the result is integrated in a digital accumulator and applied to a 10-bit dac, increasing or decreasing the src pin current (and consequently the laser bias current) as required to drive the loop error to zero. the servo loop adjusts the laser bias current every four milliseconds, producing 250 servo iterations per second. the open-loop gain of the apc loop is proportional to the laser slope efficiency, h (watts/amp), and monitor diode response, g (amps/watt). these parameters vary widely from laser to laser. if nothing is done to compensate the variations in h and g , the settling time of the optical power output will vary over an unacceptably wide range. for example, a 4:1 variation in slope efficiency and a 5:1 variation in monitor diode response could create a 20:1 variation in settling time. the ltc5100 uses two techniques to fully compensate for variations in the laser and monitor diode characteristics, achieving constant settling times under all conditions. first, taking the logarithm of the monitor diode current precisely compensates variations in the monitor diode response. second, multiplying the error signal by the modulation current precisely compensates for variations in laser slope efficiency. the difference equation for the apc loop is: im_ im _ im _ ( ) im _ im _ im _ adc d adc a d error d adc a d set d adc nn nn =+ =+ () - -- 1 11 19 where a is the small-signal loop gain, given by: a apc_gain 32 = + + + + _ im_ ln( ) () 1 1 1 8 20 1 1 is rng rng er er rr r tld t where: ln(8) = 2.079 is the natural logarithm of 8 er is the extinction ratio r t is the termination resistance r ld is the dynamic resistance of the laser diode apc_gain is a 5-bit digital value, so the scaling factor, apc_gain/32, takes on the discrete values 0, 1/32, 2/32, , 31/32. in practice, the extinction ratio is usually high (er >> 1), and r t ~ r ld , so equation 20 simplifies to: a apc gain is rng rng ? + + _ _ im_ 32 1 1 (21) operatio u figure 24. operating ranges for the monitor diode current 0 34 imd_rng = 0 5100 f24 7 136 544 2176 imd_rng = 1 recommended minimum is 20% of full scale i md ( a) (log scale) imd_rng = 2 imd_rng = 3
ltc5100 25 5100f equation 20 shows that the loop gain is completely inde- pendent of the slope efficiency and monitor diode re- sponse. consequently the servo dynamics and settling time are independent of these highly varying quantities. the apc_gain quantity can be set to compensate for the selected values of is_rng and im_rng as well as the extinction ratio, termination resistance and laser dynamic resistance. the step response of the apc loop is: imd_adc n = imd_set ? [1 C (1 C a) n ] (22) the step response given in equation 22 has the familiar exponential settling characteristic of a first order system. the step response is shown in figure 25 for a = 0.5. the remaining error is reduced by one-half with each servo iteration. in seven iterations, or about 28ms, the modula- tion current settles to under 1% in this example. the measured step response, including the modulation enve- lope, is shown in the typical performance characteristics. choosing a = 0.5 is nearly optimal because it results in smooth, exponential settling. a = 1 will settle in about two servo iterations or 8ms, but hunting or low level oscil- lation will be seen in the laser bias current. a > 1 results in overshoot and a > 2 results in sustained high level oscillation. microprocessor, according to the setting of ext_temp_en. the temperature compensated expression for imd_set is given by: im _ im _ im _ im _ d set d nom dtc t dtc t = d +d+ ? ? ? ? 22 12 1 18 2 13 (23) imd_tc1 and imd_tc2 are the first and second order temperature coefficients for the monitor diode current. equation 23 applies to the digital representation of the monitor diode current. recall that imd_set is the digital set point for the logarithm of the monitor diode current. this fact has two important implications. first, the first order temperature coefficient in equation 23 (imd_tc1) results in an exponential change in the physical monitor diode current with temperature. however, the monitor diode temperature drift is usually very small, and the exponential is well approximated as linear. second, if imd_tc2 = 0, the relative temperature sensitivity of the physical current is given by: di md dt i dtc d nom md ln( ) im _ im _ 1 82 1 1024 13 = (24) where i md is the physical monitor diode current in amps. equation 24 shows that the temperature coefficient of the physical current depends on the nominal monitor diode current. for example, if imd_nom = 512 and imd_tc1 = 4, the physical temperature compensation would be: di md dt i ppm c md ln( ) / 1 82 4 512 1024 508 13 == (25) the effect of imd_tc2 on the physical monitor diode current has no simple physical interpretation. in most cases it will be sufficient to set imd_tc2 to zero and use the first order temperature coefficient, imd_tc1 to correct monitor diode drift. laser bias current control in ccc mode figure 4 is a functional diagram of the ltc5100 operating in constant current control (ccc) mode. in ccc mode, the ltc5100 sets the laser bias current directly. setting apc_en = 0 selects this mode. in ccc mode the laser bias operatio u figure 25. step response of the monitor diode current for a total loop gain of 0.5 im_adc 0 4 8 12 16 20 24 28 32 5100 f25 2 1 345678 time (ms) servo iterations imd_set temperature compensation the set point value for the monitor diode current, imd_set in figure 3, can be changed with temperature to compen- sate the temperature dependence of the monitor diode response. temperature measurements are supplied either by an on-chip temperature sensor or by an external
ltc5100 26 5100f current can be temperature compensated with first and second order temperature coefficients. servo control the laser bias current is controlled by a digital servo loop (shown in the upper half of figure 4) and can be under- stood as follows. the nominal bias current, ib_nom, is multiplied by a temperature compensation factor, produc- ing a 10-bit digital set point value, ib_set. ib_set is the target value for the laser bias current. the adc digitizes the src pin current and the average modulation current, producing 10-bit values is_adc and im_adc. the laser bias current is the difference between the src pin current and the average modulation current (equation 1). the system generates a digital representation of the laser bias current by calculating: ib_adc = is_rng ? is_adc C im_rng ? im_adc (26) where ib_adc is the result of a calculation. (the adc never digitizes the laser bias current directly.) the difference between the target value and the actual value is the servo loop error signal, ib_error. ib_error is multiplied by a constant, ib_gain, to set the loop gain. the result is integrated in a digital accumulator and applied to a 10-bit dac, increasing or decreasing the src pin current as required to drive the loop error to zero. the servo loop adjusts the src pin current every four milliseconds, producing 250 servo iterations per second. the simplified difference equation for the bias current servo loop is, assuming im_nom = 0: ib adc ib adc ib gain is rng ib error ib adc ib gain is rng ib set ib adc n n n n _() _ _ ( _ ) _ _ _ ( _ ) __ = ++ =+ + () - - - 27 32 1 32 1 1 1 1 ib_gain is a 5-bit digital value, so the scaling factor, ib_gain/32, takes on the discrete values 0, 1/32, 2/32, , 31/32. if ib_gain ? (is_rng + 1) = 16, then ib_gain ? (is_rng + 1)/32 = 0.5 and the error in the control loop is cut in half with each servo iteration. in this case the step response of the loop is given by, assuming im_nom = 0 : ib adc ib set ib gain is rng n n _ _ _(_ ) = + ? ? ? ? ? ? 11 1 32 (28) the step response has the familiar exponential settling characteristic of a first order system. the step response is shown in figure 26 for ib_gain ? (is_rng + 1) = 16. the remaining error is reduced by one-half with each servo iteration. in seven iterations, or about 28ms, the laser bias current settles to under 1% in this example. the mea- sured step response is shown in the typical performance characteristics. operatio u ib_adc 0 4 8 12 16 20 24 28 32 5100 f26 2 1 345678 time (ms) servo iterations ib_set figure 26. step response of the laser bias current for (ib_gain) ? (is_rngtl ) = 16 reducing ib_gain slows the settling time and increasing ib_gain speeds the settling time. for example, with ib_gain ? (is_rng + 1) = 1, the residual loop error is cut by 1/32 with each servo iteration. in this case it would take 145 servo iterations (about 580ms) to settle to 1%. with ib_gain ? (is_rng + 1) = 31, the residual servo loop error is cut by 31/ 32 with each servo iteration. in this case it would take only two servo iterations (about 8ms) to settle to 1%. setting im_nom 1 0 slows the settling time of the laser bias current somewhat. this effect can easily be compen- sated by increasing ib_gain. temperature compensation the set point value for the laser bias current, ib_set in figure 4, can change with temperature to compensate the temperature dependence of the laser diodes threshold current. temperature measurements are supplied either
ltc5100 27 5100f by an on-chip temperature sensor or by an external microprocessor, according to the setting of ext_temp_en. the temperature compensated expression for ib_set is given by: ib set ib nom ib tc t ib tc t __ _ _ = d +d+ ? ? ? ? 22 12 1 18 2 13 (29) ib_tc1 and ib_tc2 are the first and second order tempera- ture coefficients for the laser bias current. transmit enable, fault detection and eye safety the ltc5100 is compatible with the gigabit interface converter (gbic) specification, but includes additional features and safety interlocks. figure 27 shows the state diagram for enabling the transmitter and detecting faults. the en pin and soft_en control bit enable and disable the transmitter. the en pin may be programmed for active high or active low operation with the en_polarity bit. operatio u ready (disabled) 2 settling (enabled) 4 read eeprom 1 wait 64ms 3 settled (enabled) 5 ready (settled and disabled) 6 faulted (disabled) 8 power on reset or operating_mode = 0 succeeded or operating_mode = 1 100ms timeout enable and rapid_restart_en = 0 enable enable transmitter_enabled = 0 fault detection disabled transmitter_enabled = 1 fault pin asserted transmitter_enabled = 0 transmit_ready = 0 faulted = 1 settling (enabled) 10 fault detection disabled transmitter_enabled = 1 settling (enabled) 11 faulted twice (disabled) 12 fault detection enabled transmit_ready = 1 power on reset 5100 f27 faulted = 1 faulted_once = 1 faulted_twice = 1 faulted = 0 faulted_once = 1 25ms timeout 100ms timeout transmitter_enabled = 0 transmit_ready = 0 fault fault detection enabled transmit_ready = 1 fault detection disabled transmit_ready = 1 enable = en and soft_en fault = over_current or over_power or under_power failed timeout mem_load_error = 1 operating_mode = 1 ready (disabled) 9 enable enable fault settling (enabled) 7 enable faulted_once = 0 enable 40ms timeout enable and rapid_restart_en = 1 enable and rep_flt_inhibit = 0 enable and rep_flt_inhibit = 1 figure 27. state diagram for transmitter enable and fault detection
ltc5100 28 5100f the en pin and the soft_en bit must both be active to enable the transmitter, providing an extra degree of safety and allowing full software control of the transmitter enable function. as shown in figure 6, the en pin has a weak 10 m a current source that pulls it to the inactive state in case of an accidental open on the pin. the en and soft_en bits are inhibited until the ltc5100 has successfully loaded its registers from an eeprom or the operating_mode bit has been set, signaling that a microprocessor has assumed control of the chip. the first time the transmitter is enabled after initial power up, the servo loops find the correct dac settings for bias and modulation current through a feedback process. initial settling is typically within 300ms. if the transmitter is disabled and subsequently re-enabled, the previously determined dac settlings are restored. in this case set- tling occurs typically within 1ms. this feature is called rapid restart and can be overridden by setting the rapid C restart_en bit to zero. the ltc5100 has sophisticated eye safety and fault han- dling features. five types of faults are detected: low supply voltage, excessive laser bias current, overpower, underpower and eeprom memory load failure. table 1 summarizes these five faults and how they are handled in the ltc5100. faults are latched in compliance with gbic requirements. faults can be independently enabled (except for low sup- ply voltage and memory load failure) and are recorded in an internal register for readout over the serial bus. if two faults occur simultaneously, the fault with the highest priority (see table 1) is recorded in the flt_status register. this register indicates the cause of the fault and is cleared only when read (not when the fault itself is cleared.) low supply voltage and memory load failure are considered hard faults and cannot be masked or overrid- den. they prevent the transmitter from begin enabled until they are cleared. normally, a fault automatically disables the transmitter and shuts down the laser. in some systems it may be desirable to allow data transmission to continue after a operatio u table 1. fault detection and handling fault type laser laser laser eeprom memory power supply software overcurrent overpower underpower load fault undervoltage forced fault fault occurs when laser bias current monitor diode monitor diode eeprom load v dd drops the flt_pin_override exceeds the value current is 50% current is 50% starts but fails below 2.8v and force_flt bits in the ib_limit greater than the less than the to complete are set register set point set point priority 5 4 3 2 1 na cleared by yes yes yes yes no yes power-on reset latched in the yes yes yes yes yes no (not part of the flt_status flt_status register) register cleared from the yes yes yes yes yes no (not part of the flt_status flt_status register) register on read latched at the yes yes yes yes no (the fault pin yes (actually latched fault pin signals a fault as in the flt_config long as the supply register) voltage remains too low) enabled by over_current_en over_pwr_en under_pwr_en always enabled always enabled flt_pin_override and apc_en and apc_en glitch rejection 4 m s4 m s4 m s na 200mv typical na hysteresis
ltc5100 29 5100f fault has occurred. for example, the software in the host system may need to evaluate the cause of the fault before shutting down the laser. if auto_shutdn_en = 1, the ltc5100 automatically disables the transmitter after a fault. if auto_shutdn_en = 0, data transmission continues after a fault. the transmitter is not disabled until the host system drives the en pin inactive or clears the soft_en bit. low power supply voltage and memory load errors are considered hard faults and always disable the transmitter, regardless of the setting of auto_shutdn_en. the ltc5100 implements the gbic protocol for prevent- ing software from repeatedly re-enabling a faulted trans- mitter. when a first fault is detected, it can be cleared by disabling the transmitter. if the transmitter is re-enabled and a second fault occurs within 25ms after fault detection is enabled, the transmitter is permanently disabled. only cycling power to the ltc5100 can clear this condition. this feature is called repeated fault inhibit and can be overridden by setting the repeated_flt_inhibit bit to zero. the fault pin can be configured active high or active low with the flt_pin_polarity bit. the fault pin can be pro- grammed for open drain, 330 m a internal pull-up, 500 m a internal pull-up or complementary (push-pull) drive with the two flt_drv_mode bits. refer to figure 8 for an equivalent circuit of the fault pin. the fault pin can be overridden in software for testing purposes or to allow a microprocessor in the transceiver module to fully control the modules fault output. if the flt_pin_override bit is set, then the force_flt bit fully controls the state of the fault pin. the state of the ltc5100 can be monitored by reading the flt_status register. see table 21 for a description of the status bits. eye safety information communications lasers can emit levels of optical power that pose an eye safety risk. while the ltc5100 provides certain fault detection features, these features alone do not ensure that a laser transmitter using the ltc5100 is compliant with iec 825 or the regulations of any particu- lar agency. the user must analyze the safety require- ments of their transceiver module or system, activate the appropriate laser safety features of the ltc5100, and take any additional precautions needed to ensure compliance of the end product with the requirements of the relevant regulatory agencies. in particular, the ltc5100 produces laser currents in response to digitally programmed com- mands. the user must ensure software written to control the ltc5100 does not cause excessive levels of radiation to be emitted by the laser. power consumption and power management the power consumption of the ltc5100 is dependent on several variables, including the modulation current range (set by im_rng), the laser bias and modulation levels, and the state of the transmitter (whether enabled or disabled.) if power_down_en = 1, the ltc5100 turns off its high speed amplifiers when the transmitter is disabled, reduc- ing supply current to less than 5ma (typical). see the typical performance chacteristics for further information. high speed peaking control the ltc5100 has the ability to selectively peak the falling edge of the modulation waveform to accelerate the turn-off of the laser diode. the 5-bit peaking register controls this function. see the typical performance chacteristics for further information. lower values in the peaking register increase the falling edge peaking. analog-to-digital conversion overview the adc in the ltc5100 is a 10-bit, dual slope integrating converter with excellent linearity and noise rejection. a multiplexer allows digitizing six quantities: ? src pin current, i s ? average modulation current, i m ? laser diode voltage, v ld ? monitor diode current, i md ? termination resistor voltage, v term ? die temperature, t all of these measurements are available to the user via the i 2 c serial bus. operatio u
ltc5100 30 5100f conversion sequence the adc has a 1ms conversion time and operates in a four- cycle sequence. three of these cycles are dedicated to the needs of the servo controllers for laser bias and modula- tion current. one cycle is available to the user to convert any desired quantity. table 2 shows how the four conver- sion time slots are allocated. the temperature compensa- tion and servo loop calculations are done during the user cycle. the source and modulation dacs are also updated during this cycle. table 2. adc conversion sequence result stored in cycle apc mode ccc mode register 1 t t t_int_adc 2i m i m im_ adc 3i md i s imd_adc/is_adc 4 user user user_adc user access to the adc the results of each conversion cycle in table 2 are stored in user accessible registers. the last die temperature measurement can be read over the i 2 c bus at any time by reading the t_int_adc register. note that the quantity converted during the third cycle depends on whether the chip is in apc or ccc mode. the result of the third conversion cycle is stored in a register that is called imd_adc in apc mode and is_adc in ccc mode. there is only one register, but it is given two names to indicate the quantity it actually holds. the fourth cycle, called the user cycle, is available to digitize any of the six multiplexed signals. the result can be read out over the i 2 c serial bus. the signal to be digitized during the user cycle is selected by setting the three-bit field user_adc.adc_src_sel (see table 23). for example, setting adc_ src_sel = 2 programs the multiplexer to select the laser diode voltage, v ld . during the next user conversion cycle, v ld is converted and stored to the user_adc. data field. when the conversion is complete, user_adc.valid is set and user_adc.adc_src indicates the signal source whose converted value is stored in user_adc.data. reading or operatio u writing the user_adc register clears the valid bit. the valid bit remains cleared until the next user conversion is complete. user_adc.adc_src always corresponds to the signal source whose data is stored in user_adc.data, not the source that was most recently selected by writing user_adc.adc_src_sel. the valid bit and adc_src field are useful for monitoring when the adc has updated the user_adc.data field. table 3 gives an extended example of accessing the user_adc register. note that the content of the user_adc register is different for writing and for reading, even though the i 2 c command used to access this register is the same in both cases. see table 23 and table 24 for a detailed definition of the bit fields in the user_adc register. table 23 also shows how to convert adc digital codes to real-world quantities. direct microprocessor control of the laser bias and modulation current setting lpc_en to zero turns off the ltc5100s digital laser power controller (see figure 2). the source and modulation dacs (is_dac and im_dac) can then be written from the i 2 c serial bus, allowing an external microproces- sor or test computer to directly control the source and modulation currents. digital control and the i 2 c serial interface the ltc5100 has extensive digital control and monitoring features. these features can be used during final assembly of a transceiver module to set up the laser and verify performance. in normal operation, the ltc5100 can oper- ate standalone or under microprocessor supervision. op- erating standalone, the ltc5100 automatically loads its configuration and laser operating parameters (bias cur- rent, modulation current, monitor diode current) from a small external eeprom at power up. operating under microprocessor supervision, the microprocessor is in total control of setting up the ltc5100. i 2 c serial interface protocol the digital interface for the ltc5100 is i 2 c, a 2-wire serial bus standard that is fully documented in i 2 c-bus and how
ltc5100 31 5100f operatio u table 3. example of user adc cycle access read from adc write to adc_user cycle signal source adc_src_sel register adc_src valid data comment 1t v term 0v term (1) selected signal source is v term 2i m v term 0v term (1) 3i s v term 0v term (1) 4 user (v term )v term 0v term (1) 1t v term 1v term (2) adc updates data with new data, setting valid 2i m v ld v term 0v term (2) user selects new signal source, v ld , clearing valid 3i s v term 0v term (2) 4 user (v ld )v term 0v term (2) 1t v ld 1v ld (1) adc updates data with new data, setting vaild and changing adc_src to reflect the source of the new data 2i m v ld 1v ld (1) 3i s v ld v ld 0v ld (1) user reads the adc_user register, clearing valid 4 user (v ld )v ld 0v ld (1) 1t v ld 1v ld (2) adc updates data with new data, setting valid 2i m v ld 1v ld (2) 3i s v ld 1v ld (2) 4 user (v ld )v ld 1v ld (2) s write read w a ltc5100 address (7 bits) 0x14 command byte low byte a a high byte a p low byte a high byte n a p 5100 f28 sw a ltc5100 address (7 bits) 0x14 ltc5100 address 0x14 command byte s a a r to use it, v1.0 by philips semiconductor. the i 2 c bus address for the ltc5100 is 0x14 (hex). to communicate with the ltc5100, the bus master transmits the ltc5100 address followed by a command byte and data as defined by the i 2 c bus specification and shown in figure 28 and table 4. note that 16 bits of data are always transmitted, low byte first, high byte last. within each transmitted byte, the bit order is msb .. lsb. the register set and i 2 c command set for the ltc5100 are documented in table 7 through table 30. figure 28. i 2 c serial read/write sequences (ltc5100 responses are shown in bold italics) table 4. legend for the i 2 c protocol symbol meaning s start w write r read a acknowledge na no acknowledge p stop
ltc5100 32 5100f standalone operation on power-up the ltc5100 becomes an i 2 c bus master and attempts to load its configuration data from an exter- nal eeprom. if an eeprom responds, the ltc5100 reads 16-bytes of data and transfers this data to the internal register set. when a 16-byte transfer is completed without error, the ltc5100 becomes ready to enable the transmit- ter and begin driving the laser. if a bus error occurs during this transfer, the load sequence is aborted and a mem_load_error is generated, preventing the transmitter from being enabled until a successful memory load at- tempt is completed or until an external agent sets the operating_mode bit. every 64ms another attempt is made to load the eeprom until the memory is read or until operating_mode = 1. table 5 shows the memory map for the eeprom. the ltc5100 generates i 2 c address 0xae (1010_1110 binary) when accessing the eeprom, making it compat- ible with a wide range of eeprom sizes. table 6 details how the ltc5100 interacts with eeproms from 128 bits to 16k bits and from where it gets its data. the ltc5100 supports hot plugging in standalone mode. if the soft_en bit is set in the eeprom and the en pin is active, the ltc5100 loads its configuration data from the eeprom and immediately enables the transmitter. the transmitter is typically enabled and settled within the 300ms t_init period required by the gbic specification. operatio u table 5. eeprom memory map bit byte 7 6 5 4 3 2 1 0 15 reserved peaking (4:0) 14 ib_gain(4:0)/apc_gain(4:0) im_gain(2:0) 13 reserved imd_rng(1:0) t_nom(9:8) 12 t_nom(7:0) 11 im_tc2(7:0) 10 im_tc1(7:0) 9 reserved im_rng(1:0) im_nom(9:8) 8 im_nom(7:0) 7 ib_tc2(7:0)/imd_tc2(7:0) 6 ib_tc1(7:0)/imd_tc2(7:0) 5 reserved is_rng(1:0) ib_nom(9:8)/imd_nom (9:8) 4 ib_nom(7:0)/imd_nom(7:0) 3 reserved rep_flt_inhibit rapid_restart_en flt_drv_mode 2 lpc_en auto_shutdn_en flt_pin_polarity flt_pin_override force_flt over_pwr_en under_pwr_en over_current_en 1 reserved ib_limit 0 cml_en md_polarity ext_temp_en power_down_en apc_en en_polarity soft_en operating_mode
ltc5100 33 5100f operatio u table 6. effective base addresses for various sized eeproms generic part number 24lc00 24lc01b 24lc02b 24lc04b 24lc16b bits 128 1k 2k 4k 16k bytes 16 128 256 512 2048 device address (binary) 1010xxx. 1010xxx. 1010xxx. 1010.xxa 1010cba. word address space (binary) xxxx_nnnn xnnn_nnnn nnnn_nnnn nnnn_nnnn nnnn_nnnn ltc5100 generates 1010_111. 1010_111. 1010_111. 1010_111. 1010_111. device address = 0xae = 0xae = 0xae = 0xae = 0xae ltc5100 generates 0110_0000 0110_0000 0110_0000 0110_0000 0110_0000 word address = 0x60 = 0x60 = 0x60 = 0x60 = 0x60 effective base address 0000_0000 0110_0000 0110_0000 0001_0110_0000 0111_0110_0000 = 0x00 = 0x60 = 0x60 = 0x160 = 0x760 comments minimum size eeprom not big standard gbic ltc5100 loads ltc5100 loads eeprom. enough for gbic eeprom. from an area from an area loads every id. ltc5100 smallest eeprom outside the gbic outside the gbic byte in the loads from 0x60 that is big enough id data area id data area eeprom. to 0x6f to hold the ltc5100 data and the gbic id. ltc5100 loads from 0x60 to 0x6f, the first 16 bytes of the vendor area microprocessor controlled operation an external microprocessor or a test computer can take full control of the ltc5100 by setting the operating_mode bit. when this bit is set, the ltc5100 stops searching for an external eeprom and takes commands from the mi- croprocessor. it is even possible to combine standalone and microprocessor controlled modes. if an eeprom is present, the ltc5100 will load its configuration registers from the eeprom at power-up. a microprocessor or test computer can then read and write the ltc5100 registers at will. the primary purpose of the operating_mode bit is to stop the ltc5100s eeprom load attempts. once the ltc5100 has loaded itself from an eeprom (if present), it is not technically necessary to set the operating_mode bit to communicate with the ltc5100. the ltc5100 attempts to read the eeprom every 64ms until it successfully loads its registers or until the operating_mode bit is set. there is a finite chance that the microprocessor and the ltc5100 will generate an i 2 c bus collision if an eeprom load attempt coincides with the microprocessors attempt to access the ltc5100. in this case, the microprocessor will receive a nack (not-ac- knowledged) response to its transmissions. the micro- processor needs only to cease transmission in accordance with the i 2 c protocol and try again. if the microprocessor makes this second attempt within 64ms (typical), it is guaranteed not to collide with the ltc5100.
ltc5100 34 5100f register defi itio s u u table 7. register set overview register name constant current automatic power i 2 c command read/write reference register group control mode control mode code (hex) access information system operating sys_config 0x10 r/w table 8 configuration loop_gain 0x1e r/w table 9 peaking 0x1f r/w table 10 reserved 0x08 r/w table 11 laser setup coefficients ib imd 0x15 r/w table 12 ib_tc1 imd_tc1 0x16 r/w table 13 ib_tc2 imd_tc2 0x17 r/w table 14 im 0x19 r/w table 15 im_tc1 0x1a r/w table 16 im_tc2 0x1b r/w table 17 temperature t_ext 0x0d r/w table 18 t_nom 0x1d r/w table 19 fault monitoring flt_config 0x13 r/w table 20 and eye safety flt_status 0x12 r table 21 ib_limit 0x11 r/w table 22 adc user_adc 0x18 r/w tables 23, 24 t_int_adc 0x05 r/w table 25 im_adc 0x06 r/w table 26 is_adc imd_adc 0x07 r/w table 27 dac is_dac 0x01 r/w table 28 im_dac 0x02 r/w table 29 pwr_ limit_dac 0x03 r table 30
ltc5100 35 5100f register defi itio s u u table 8. register: sys_configsystem configuration (i 2 c command code 0x10) register reset value .bitfield bit (bin) function and values .reserved 15:8 .cml_en 7 0 current mode logic enable 0: floating differential input termination: 100 w across in + and in C 1: cml compatible input termination: 50 w from in + to v dd(hs) and from in C to v dd(hs) .md_polarity 6 0 monitor diode polarity 0: cathode connected to the md pin, sinking current from the pin 1: anode connected to the md pin, sourcing current into the pin .ext_temp_en 5 0 external temperature enable selects the source of temperature measurements for temperature compensation. 0: internal temperature sensor 1: externally supplied through the serial interface .power_down_en 4 1 power down enable allow power reduction when the transmitter is disabled. 0: no power reduction when the transmitter is disabled. 1: reduce power consumption when transmitter is disabled by turning off the high speed amplifiers. .apc_en 3 0 automatic power control enable select the means of controlling the laser bias current. 0: constant current control 1: automatic power control using feedback from the monitor diode .en_polarity 2 0 en pin polarity set the input polarity of the en pin. 0: active low: a logic low input level enables the transmitter. 1: active high: a logic high input level enables the transmitter. note: in order to enable the transmitter, both the en pin and soft_en bit must be asserted. .soft_en 1 0 soft transmitter enable enables transmitter through the serial interface. 0: disable the transmitter 1: enable the transmitter (if the en pin is active) note: in order to enable the transmitter, both the en pin and soft_en bit must be asserted. .operating_mode 0 0 digital operating control mode select whether the ltc5100 operates autonomously or under external control. 0: standalone operation: configuration parameters are loaded from an external eeprom at power up. 1: externally controlled operation: configuration parameters are set by an external microprocessor or test computer.
ltc5100 36 5100f register defi itio s u u table 9. register: loop_gaincontrol loop gain (i 2 c command code 0x1e) register reset value .bitfield bit (bin) function and values .reserved 15:8 .ib_gain 7 0 bias current or apc loop gain (.apc_gain in apc 6 0 this bit field modifies the open-loop gain of the bias current servo control loop. the effect of this mode) 51 bit field differs in constant current control (ccc) mode and in automatic power control (apc) mode. 40 in ccc mode, this bit field is called lb_gain. in apc mode, this bit field is called apc_gain. 30 constant current control (ccc) mode (apc_en = 0): the loop gain and settling time are independent of is_rng. the default value of ib_gain yields stable but slow settling of the laser bias current for any value of is_rng. automatic power control (apc) mode (apc_en = 1): the open-loop gain of the bias current servo loop depends on the value of is_rng. the default value of apc_gain yields stable but potentially slow settling of the laser bias current for any value of is_rng. im_gain 2 0 modulation current loop gain 1 0 this bit field modifies the open-loop gain of the modulation current servo loop. the open-loop. 01 gain is approximately im_gain/32. the loop gain and settling time are independent of im_rng. the default value of im_gain yields stable but slow settling of the laser modulation current. table 10. register: peakinghigh speed modulation peaking (i 2 c command code 0x1f) register reset value .bitfield bit (bin) function and values .reserved 15:5 .peaking 4 1 peaking control for the modulation output 3 0 this bit field controls the high speed peaking of the modulation output. decreasing the value of 20 peaking increases the undershoot on the falling edge of the modulation signal. the peaking control 10 can be used to compensate for slow laser turn-off characteristics. 00 table 11. register: reservedreserved for internal use. this register is for test puposes only. do not write to this register (i 2 c command code 0x08) register reset value .bitfield bit (bin) function and values .reserved 15:7 .reserved 6 1 reserved for internal use, do not write. 50 40 30 .reserved 2 1 reserved for internal use, do not write. 10 00
ltc5100 37 5100f table 12. register: ib (imd)laser bias current register (monitor diode current in apc mode) (i 2 c command code 0x15) register reset value .bitfield bit (bin) function and values .reserved 15:12 .is_rng 11 0 source current range 10 0 is_rng sets the full-scale range of the src pin current. the table below shows the available ranges. values for is_rng nominal full- binary decimal scale src pin value value current (ma) 00 0 9 01 1 18 10 2 27 11 3 36 see the electrical specifications for guaranteed limits in each range. .ib_nom 9 0 bias current or monitor diode current setting at the nominal temperature (.imd_nom in 8 0 this bit field has different functions depending on apc_en. apc mode) 7 0 this bit field is an unsigned 10-bit integer. 60 constant current control (ccc) mode (apc_en = 0): ib_nom sets the laser bias current at 50 temperature t = t_nom. the physical bias current at t_nom is given by: 4 0 (typical) 30 20 automatic power control (apc) mode (apc_en = 1): imd_nom sets the monitor diode current at the 10 temperature t = t_nom. the physical monitor diode current at t_nom is given by: 00 register defi itio s u u i= ib_nom 1024 b ( _ ) is rng ma + 19 ia d nom md drng =m ? ? 425 4 8 1024 . exp ln( ) im _ im _ table 13. register: ib_tc1 (imd_tc1)laser bias/monitor diode current first order temperature coefficient (i 2 c command code 0x16) register reset value .bitfield bit (bin) function and values .reserved 15:8 .ib_tc1 (.imd_tc1 7 0 first order temperature coefficient for bias current or monitor diode current in apc mode) 6 0 this bit field is a signed 8-bit, twos complement integer. thus its value ranges from C128 to 127. 5 0 this bit field has different functions depending on apc_en. 40 constant current control (ccc) mode (apc_en = 0): ib_tc1 sets the first order temperature 30 coefficient for the laser bias current. the nominal scaling is 2 C13 / c or 122ppm/ c per lsb. 20 automatic power control (apc) mode (apc_en = 1): imd_tc1 sets the first order temperature 10 coefficient for the monitor diode current. see laser bias current control in apc mode 00 in the operation section for details.
ltc5100 38 5100f table 14. register: ib_tc2 (imd_tc2)laser bias/monitor diode current second order temperature coefficient (i 2 c command code 0x17) register reset value .bitfield bit (bin) function and values .reserved 15:8 .ib_tc2 (.imd_tc2 7 0 second order temperature coefficient for bias current or monitor diode current in apc mode) 6 0 this bit field is a signed 8-bit, twos complement integer. thus its value ranges from C128 to 127. 5 0 this bit field has different functions depending on apc_en. 40 constant current control (ccc) mode (apc_en = 0): ib_tc2 sets the second order temperature 30 coefficient for the laser bias current. the nominal scaling is 2 C18 / c 2 or 3.81ppm/ c 2 per lsb. 20 automatic power control (apc) mode (apc_en = 1): imd_tc2 sets the second order temperature 10 coefficient for the monitor diode current. see laser bias current control in apc mode 00 in the operation section for details. table 15. register: imlaser modulation current (i 2 c command code 0x19) register reset value .bitfield bit (bin) function and values .reserved 15:12 .im_rng 11 0 modulation current range 10 0 im_rng sets the full-scale range of the modulation current binary decimal nominal full-scale moda and modb pin current value value peak-to-peak (ma) average (ma) 00 0 9 4.5 01 1 18 9 10 2 27 13.5 11 3 36 18 see the electrical specifications for guarant eed limits in each range. these currents represent the peak-to-peak current at the moda and modb pins. (the moda and modb pins are tied together on chip). .im_nom 9 0 modulation current setting at the nominal temperature 8 0 this bit field is an unsigned 10-bit integer. im_nom sets the average modulation current delivered at 70 the moda and modb pins. (the moda and modb pins are tied together on chip). 60 the peak-to-peak current is twice the average current for a data stream with 50% duty cycle. 50 the modulation current reaching the laser depends on its dynamic resistance relative to the 40 termination resistor. 30 20 10 00 register defi itio s u u
ltc5100 39 5100f table 16. register: im_tc1laser modulation current first order temperature coefficient (i 2 c command code 0x1a) register reset value .bitfield bit (bin) function and values .reserved 15:8 .im_tc1 7 0 first order temperature coefficient for modulation current 6 0 this bit field is a signed 8-bit, twos complement integer. thus its value ranges from C128 to 127. 50 im_tc1 sets the first order temperature coefficient for the modulation current. the nominal scaling 40 is 2 C13 / c or 122ppm/ c per lsb. 30 20 10 00 register defi itio s u u table 17. register: im_tc2laser modulation current second order temperature coefficient (i 2 c command code 0x1b) register reset value .bitfield bit (bin) function and values .reserved 15:8 .im_tc2 7 0 second order temperature coefficient for modulation current 6 0 this bit field is a signed 8-bit, twos complement integer. thus its value ranges from C128 to 127. 50 im_tc2 sets the second order temperature coefficient for the laser bias current. the nominal 40 scaling is 2 C18 / c 2 or 3.81ppm/ c 2 per lsb. 30 20 10 00 table 18. register: t_extexternal temperature (i 2 c command code 0x0d) register reset value .bitfield bit (bin) function and values .reserved 15:10 .t_ext 9 0 externally supplied temperature for temperature compensation calculations (unsigned 10-bit 80 integer) 7 0 by convention the scaling of t_ext is 512k or 239 c full scale, corresponding to 0.5 c/lsb. 60 however, any scaling is permissible as long as the temperature compensation coefficients are also 50 appropriately scaled. 4 0 t_ext = (t + 273 c)/0.5 c, where t is the external temperature in degrees celsius. 30 20 10 00
ltc5100 40 5100f table 19. register: t_nomnominal temperature (includes imd_rng) (i 2 c command code 0x1d) register reset value .bitfield bit (bin) function and values .reserved 15:12 .imd_rng 11 0 monitor diode current range 10 0 imd_rng sets the full-scale range of the monitor diode current. binary decimal md pin current range ( m a) value value nom min nom max 00 0 4.25 34 01 1 17 136 10 2 68 544 11 3 272 2176 .t_nom 9 0 nominal temperature 8 0 t_nom is the temperature with respect to which all temperature compensation calculations are 70 made. t_nom is usually the temperature at which the ltc5100 and laser diode were set up in 60 production. 5 0 the scaling is 512k or 239 c full scale, corresponding to 0.5 c/lsb 4 0 t_nom = (t + 273 c)/0.5 c, where t is the nominal temperature in degrees celsius. 30 20 10 00 register defi itio s u u
ltc5100 41 5100f register defi itio s u u table 20. register: flt_configfault configuration (refer also to table 1) (i 2 c command code 0x13) register reset value .bitfield bit (bin) function and values .reserved 15:12 rep_flt_inhibit 11 0 repeated fault inhibit 0: allow repeated attempts to clear a fault and re-enable the transmitter. 1: inhibit repeated attempts to clear a fault. only one attempt to clear a fault is allowed. if the fault recurs within 25ms of re-enabling the transmitter, the transmitter is disabled until power is cycled. rapid_restart_en 10 1 rapid_restart_en 0: rapid restart disabled: the servo controller settings for the laser bias and modulation currents are reset to zero when the transmitter is disabled. when re-enabled, the laser currents start from zero and settle typically within the 300ms standard initialization time, t_int, from the gbic specification. 1: rapid restart enabled: the servo controller settings for the laser bias and modulation currents are retained when the transmitter is disabled. when re-enabled, the retained servo values are loaded into the src_dac and mod_dac, allowing settling typically within the 1ms standard turn-on time, t_on, from the gbic specification. flt_drv_mode 9 0 fault pin drive mode 8 0 00: open drain (3.3ma sink capability) 01: open drain, 280 m a internal pull up 10: open drain, 425 m a internal pull up 11: push-pull (3.3ma source and sink capability) lpc_en 7 1 laser power controller (lpc) enable 0: lpc disabled: allows external control of the src_dac and mod_dac registers from the serial interface. this setting gives an external microprocessor or test computer full control of the src_dac and mod_dac registers. 1: lpc enabled: the lpc continuously updates the src_dac and mod_dac registers to servo control the laser. (any values written to these registers over the serial interface will be overwritten by the lpc.) auto_shutdn_en 6 1 automatic transmitter shutdown enable 0: disabled: when a fault occurs the ltc5100 continues to drive the laser. this mode allows a microprocessor or test computer to mediate the decision to shut down the transmitter. the microprocessor can turn off the transmitter by driving the en pin inactive or by clearing the soft_en bit in the sys_config register. 1: enabled: when a fault occurs, the transmitter is automatically disabled. flt_pin_polarity 5 1 fault pin polarity 0: active low: the fault pin is driven low to signal a fault. 1: active high: the fault pin is driven high to signal a fault. flt_pin_override 4 0 fault pin override 0: the fault pin is driven active when a fault occurs. 1: internal control of the fault pin is overridden. when a fault occurs, the fault is detected and latched internally, but the fault pin remains inactive. this mode allows a microprocessor or test computer to mediate fault handling. the microprocessor can drive the fault pin active by setting the force_flt bit. force_flt 3 0 force the fault pin output. force_flt gives a microprocessor or test computer full control of the fault pin, allowing external mediation of fault handling. 0: force the fault pin inactive. 1: force the fault pin active. this bit has no effect unless flt_pin_override = 1. over_pwr_en 2 1 enables detection of a laser overpower fault. 0: disabled, 1: enabled under_pwr_en 1 1 enables detection of a laser underpower fault. 0: disabled, 1: enabled over_current_en 0 1 enables detection of a laser overcurrent fault. 0: disabled, 1 enabled
ltc5100 42 5100f register defi itio s u u table 21. register: flt_statusfault status (i 2 c command code 0x12) register reset value .bitfield bit (bin) function and values .reserved 15:11 .transmit_ready 10 0 transmit ready indicates that the laser bias and modulation currents have settled to within specification and the ltc5100 is ready to transmit data. a fault clears this bit. 0: not ready, 1: ready .transmitter_ 9 0 transmitter enabled enabled indicates that the transmitter is enabled and the laser bias and modulation currents are on (though not necessarily settled.) the transmitter is enabled when the en pin and soft_en bits are active and no faults have occurred. a fault clears this bit. 0: transmitter is disabled, 1: transmitter is enabled. .en_pin_state 8 varies en pin state indicates the logic level on the en pin. the en_polarity bit has no effect on en_pin_state. the power-on reset value reflects the state of the en pin. 0: en pin is low. 1: en pin is high. .faulted_twice 7 0 faulted twice (only active when rep_flt_inhibit is set) 0: either no faults or only one fault has been detected. 1: a second fault has been detected within 25ms of attempting to clear a first fault. the transmitter is disabled and can only be re-enabled by cycling the power. .faulted_once 6 1 faulted once (only active when rep_flt_inhibit is set) indicates that a first fault has been detected. after a fault occurs, faulted_once will be set at the moment the transmitter is disabled (by setting the en pin of soft_en bit inactive). if the transmitter is subsequently re-enabled and a second fault occurs within 25ms, the faulted_twice bit is set. if no fault occurs within 25ms, the faulted_once bit is cleared. 0: a first fault has not been detected or has been cleared. 1: a first fault has been detected. .faulted 5 1 faulted 0: the ltc5100 is not in the faulted state. 1: a fault has occurred and the ltc5100 has entered the faulted state (the transmitter is not disabled unless auto_shutdn_en is set). .under_votlage 4 1 undervoltage fault indicator (always enabled) cleared-on-read indicates that a power supply undervoltage event occurred. 0: no fault, 1: undervoltage fault detected. the undervoltage bit is always set at power up. read the flt_status register immediately after power-up to clear this bit. .mem_load_error 3 0 memory (eeprom) load error indicator (always enabled) cleared-on-read indicates that an attempt to load the registers from eeprom was started but did not complete successfully. 0: no fault, 1: eeprom load failed. .over_power 2 0 laser overpower fault indicator (enabled by over_pwr_en) cleared-on-read indicates that a laser overpower fault occurred. overpower occurs when the monitor diode current exceeds its set point. an overpower fault can occur only in apc mode. 0: no fault, 1: overpower fault detected. .under_power 1 0 laser underpower fault indicator (enabled by under_pwr_en) cleared-on-read indicates that a laser underpower fault occurred. underpower occurs when the monitor diode current falls below its set point. an underpower fault can occur only in apc mode. 0: no fault, 1: underpower fault detected. .over_current 0 0 laser overcurrent fault indicator (enabled by over_current_en) cleared-on-read indicates that the laser bias current exceeded the value set in the ib_limit register. 0: no fault, 1: overcurrent fault detected.
ltc5100 43 5100f table 22. register: ib_limitlaser bias current limit (i 2 c command code 0x11) register reset value .bitfield bit (bin) function and values .reserved 15:7 .ib_limit 6 0 laser bias current limit 5 0 this bit field is an unsigned 7-bit integer 4 0 sets the detection level for an over_current fault. when the laser bias current exceeds this level an 30 over_current fault is generated (provided over_current_en is set). 20 the physical bias current level is given by: 10 00 i= ib_limit 128 b(limit) ( _ ) ( ) is rng ma typical + 19 register defi itio s u u table 23. register: user_adcwriting (i 2 c command code 0x18) register reset value .bitfield bit (bin) function and values .reserved 15:3 .adc_src_sel 2 0 adc source select 1 0 selects the signal to be converted by the adc during the user adc cycle 00 user adc signal sources signal select signal (binary) name description scaling 000 i s source current (src pin i s = adc_code/1024 ? (is_rng + 1) ? 9ma current) 001 i m average modulation i m = adc_code/1024 ? (im_rng + 1) ? 9ma current (moda +modb pin current) 010 v ld laser diode voltage v ld = adc_code/1024 ? 3.5v 011 i md monitor diode current i md = 4.25 m a ? 4 lmd_rng ? exp[in(8) ? adc_code/ 1024] 100 t temperature t( c) = adc_code ? 0.5 c C 273 c 101 v term termination resistor v term = adc_code/1024 ? (is_rng + 1) ? 400mv voltage 110 reserved reserved 111 reserved reserved
ltc5100 44 5100f register defi itio s u u table 24. register: user_adcreading (i 2 c command code 0x18) register reset value .bitfield bit (bin) function and values .reserved 15 .adc_src 14 0 adc signal source 13 0 specifies the signal source of the last user adc conversion. see table 23 for the definition of these 12 0 signal sources. adc_src reflects the last signal source converted. it does not necessarily hold the last value written to the adc_src_sel bit field. .reserved 11 .valid 10 0 adc data valid indicates that the result in the data bit field (defined below) contains newly converted data since the last time adc_src_sel was written or this register was read. immediately after power up valid is false. valid becomes true as soon as the first user adc conversion is completed. 0: the adc result is not a valid conversion of the most recently selected adc source. 1: the adc has finished conversion and the result is valid. .data 9 0 adc data (10-bit unsigned integer) 8 0 contains the result of the last user adc conversion. see table 23 for the definition of the available 70 signal sources. 60 50 40 30 20 10 00 table 25. register: t_int_adcinternal temperature adc (i 2 c command code 0x05) register reset value .bitfield bit (bin) function and values .reserved 15:10 .t_int_adc 9 0 adc reading of the internal (die) temperature (10-bit unsigned integer) 8 0 this bit field contains the result of the last conversion of the ltc5100s internal die temperature. 70 6 0 the scaling is 512 k or 239 c full scale, corresponding to 0.5 c/lsb. 50 4 0 t = t_int_adc ? 0.5 c C 273 c, where t is the internal temperature in degrees celsius. 30 20 10 00
ltc5100 45 5100f table 26. register: im_adcmodulation current adc (i 2 c command code 0x06) register reset value .bitfield bit (bin) function and values .reserved 15:10 .im_adc 9 0 adc reading of the modulation current (10-bit unsigned integer) 8 0 im_adc contains the last adc conversion of the average modulation current delivered at the moda 70 and modb pins. (the moda and modb pins are tied together on-chip.) the peak-to-peak current 60 s twice the average current for a data stream with 50% duty cycle. the modulation current 50 reaching the laser depends on its resistance relative to the termination resistor. 4 0 the average physical current at the moda and modb pins is given by: 30 20 10 00 register defi itio s u u table 27. register: is_adc (imd_adc)source current/monitor diode current adc (i 2 c command code 0x07) register reset value .bitfield bit (bin) function and values .reserved 15:10 .is_adc 9 0 adc reading of the src pin current or monitor diode current (.imd_adc in 8 0 this bit field has different functions depending on apc_en. apc mode) 70 constant current control (ccc) mode (apc_en = 0): is_adc contains the last adc conversion of 60 the src pin current. the physical src pin current is given by: 50 40 30 automatic power control (apc) mode (apc_en = 1): imd_adc contains the last adc conversion of 20 the monitor diode current. the physical monitor diode current is given by: 10 00 i= adc_code 1024 s ( _ ) ( ) is rng ma typical + 19 i= a4 md imd_rng 425 8 1024 . exp ( ) _ m ? ? in adc code table 28. register: is_dacsouce current dac (i 2 c command code 0x01) register reset value .bitfield bit (bin) function and values .reserved 15:10 .is_dac 9 0 dac setting for the source current (the src pin current) 8 0 read access to this dac is always available. write access is only valid if lpc_en = 0. 70 60 50 40 30 20 10 00 i= is_dac 1024 s ( _ ) ( ) is rng ma typical + 19 i= adc_code 1024 m (im_ ) ( ) rng ma typical + 19
ltc5100 46 5100f table 29. register: im_dacmodulation current dac (i 2 c command code 0x02) register reset value .bitfield bit (bin) function and values .reserved 15:10 .im_dac 9 0 dac setting for the peak-to-peak modulation current (the combined moda and modb pin currents) 8 0 read access to this dac is always available. write access is only valid if lpc_en = 0. 70 60 50 40 30 20 10 00 register defi itio s u u table 30. register: pwr_limit_dacoptical power limit dacread only (i 2 c command code 0x03) register reset value .bitfield bit (bin) function and values .reserved 15:7 .pwr_limit_dac 6 0 dac setting for the over and underpower fault detection comparator (read only) read only 5 0 this bit field has different functions depending on apc_en. 40 constant current control (ccc) mode (apc_en = 0): pwr_limit_dac has no function in this mode. 30 its contents are undefined. 20 automatic power control (apc) mode (apc_en = 1): pwr_limit_dac tracks the value of the monitor 10 diode current. the laser power controller continuously updates the pwr_limit_dac with the 00 most recent adc reading of imd. reading the dac will return the value of imd_adc shifted right by three bits. i= im_dac 1024 m (im_ ) ( ) rng ma typical + 19
ltc5100 47 5100f applicatio s i for atio wu uu high speed design and layout figure 29 and figure 30 show the schematic and layout of a minimum component count circuit for standalone op- eration. the exposed pad of the package is soldered to a copper pad on top of the board, and nine vias couple this pad to the ground plane. the four v ss pins (pins 1, 4, 9, and 12) have webs of copper connecting them to the cen- tral pad to reduce ground inductance. the laser modula- tion current returns to the ground plane primarily through the exposed pad. any measures that reduce the induc- tance from the pad to the ground plane improve the modu- lation waveforms and reduce rfi. fault sda scl v dd(hs) v dd v ss in + in v ss v ss moda modb v ss en src ltc5100 md 13 12 r1 50 c1 10nf z o = 50 11 10 9 1 l1 ferrite bead enable +tx_data ?x_data fault v dd + 3.3v v ss 2 3 4 14 15 16 8 fiber c3 10nf 24lc00 eeprom sot23 package programming pads 5100 f29 7 6 5 z o = 50 v cc nc scl v ss sda in + v ss in eeprom c1 r1 l1 c3 2 3 v ss v ss 5100 f30 moda modb 1 16 15 14 13 v ss fault sda scl v dd(hs) v dd en src md 4 11 10 12 9 5678 scl v cc v ss sda nc figure 29. schematic of a minimum component count circuit figure 30. layout of the minimum component count circuit using 0402 passive components
ltc5100 48 5100f applicatio s i for atio wu uu the termination resistor, r1, and its decoupling capacitor, c1, are placed as close as possible to the ltc5100 to reduce inductance. inductance in these two components causes high frequency peaking and overshoot in the current delivered to the laser. r1 and c1 are folded against each other so that their mutual inductance and counter- flowing current partially cancel their self-inductance. c1 has two vias to the ground plane and a trace directly to pin 12. the layout shows the eeprom placed next to the ltc5100. however, placement of the eeprom is not critical. it can be placed several centimeters from the ltc5100 or on the back of the pc board if desired. the transmission line connecting the modb pin to the laser has a short length of minimum width trace. the net inductance of this section of trace helps compensate on- chip capacitance to further reduce reflections from the chip. fault sda scl v dd(hs) v dd v ss in + in v ss v ss moda modb v ss en src ltc5100 md 13 12 r1 50 c1 10nf z o = 50 11 10 9 1 l1 ferrite bead enable +tx_data ?x_data fault v dd + 3.3v v ss 2 3 4 14 15 16 8 fiber c3 10nf 24lc00 eeprom sot23 package programming pads 5100 f29 7 6 5 z o = 50 c2 10nf v cc nc scl v ss sda c3 in + v ss in eeprom c1 r1 l1 2 3 v ss v ss 5100 f32 moda modb 1 16 15 14 13 v ss fault sda scl v dd(hs) v dd en src md 4 11 10 12 9 5678 c2 scl v cc v ss sda nc figure 31. schematic of a minimum output reflection coefficient circuit figure 32. layout of the minimum reflection coeffieicnt circuit using 0402 passive components
ltc5100 49 5100f applicatio s i for atio wu uu figure 31 and figure 32 show the schematic and layout of a minimum reflection coefficient, minimum peaking solu- tion. two capacitors, c1 and c2 are used to further reduce the inductance in the termination network. c2 has two vias to the ground plane. temperature compensation the ltc5100 has first and second order digital tempera- ture compensation for the laser bias current, laser modu- lation current, and monitor diode current. recall that in constant current control mode, the ltc5100 provides direct temperature compensation of the laser bias current and the laser modulation current. in automatic power control mode, the laser bias current is under closed-loop control and the ltc5100 provides temperature compen- sation for the monitor diode current and the laser modu- lation current. the simplest procedure for determining the temperature coefficients (tc1 and tc2 in equation 12, equation 18, equation 23, and equation 29) is as follows: ? select a nominal or representative laser diode and assemble it into a transceiver module with the ltc5100. ? set all temperature coefficients to zero. ? place the transceiver module in a temperature chamber and find the values of ib_nom, im_nom, and imd_nom that give constant average optical power and extinction ratio at several temperature points. ? record the ltc5100s temperature reading, t_int, at each temperature point. ? select a convenient value for t_nom, the nominal tem- perature. (it is customary, but not mandatory, to use 25 c for the nominal temperature.) ? find the best values of tc1 and tc2 by fitting the quadratic temperature compensation formula (equa- tion 12) to the experimental values of ib_nom, im_nom, imd_nom, and t_int. to configure the ltc5100 for normal operation, set the nominal current to the value found at the nominal tem- perature. set tc1 and tc2 to the values determined by the best fit of the data. for standalone operation, store these values in the eeprom. for microprocessor operation, store the values in the microprocessors internal non- volatile memory or in another source of nonvolatile memory and load them into the ltc5100 after power-up. the above procedure not only corrects for the laser temperature drift, but also corrects the small temperature drift found in the ltc5100s internal references. demonstration board figure 33 shows the schematic of the dc499 demonstra- tion board. details of the use of this demo board and accompanying software can be found in the dc499 demo board manual. figure 34 shows the layout of the demo board and table 31 gives the bill of materials for the demo board. the core applications circuit for the ltc5100 vcsel driver appears inside the box in figure 33. this is the complete circuit for an optical transceiver module, includ- ing power supply filtering. it consists of the ltc5100 with eeprom for storing setup parameters, l1 and c3 for power supply filtering, and r1, c1, and c2 for terminating the 50 w modulation output. the circuitry outside the box in figure 33 is for support of the demonstration. 5v power enters through 2-pin connector p2 and is regulated by u3 to 3.3v to power the ltc5100. connector p1 sends 5v power and serial control signals to another board, allow- ing a personal computer to control the ltc5100. u4 produces 1.8vdc to bias the modulation output for elec- trical eye measurements. high speed data enters the ltc5100 through sma con- nectors j1 and j2. the ltc5100 high speed inputs are internally ac coupled with rail-to-rail common mode input voltage range. the input signal swing can go as much as 300mv above v dd or below v ss without degrading perfor- mance or causing excessive current flow. the high speed inputs may be ac coupled, in which case the common mode voltage floats to mid-supply or 1.65v nominally. a common cathode vcsel can be attached to the demo board via sma connector j3. r1 establishes a precision, low reflection coefficient 50 w modulation drive. by main- taining a wide band microwave quality 50 w path, the length of the connection to the laser can be arbitrarily long. the ltc5100 generates 20% to 80% transition times of 60ps (80ps 10% to 90%), corresponding to an instanta- neous transition filtered by a 4.4ghz gaussian lowpass filter. at these speeds the primary limitation on line length is high frequency loss. for high grade, low loss laboratory cabling with silver coated center conductor and foamed ptfe dielectric, a practical limit is about 30cm.
ltc5100 50 5100f the lasers monitor diode (if needed) can be attached to either pin of 2-pin header h2 (labeled md) or to the test turret labeled md. h2 is a 2mm, 2-pin header with 0.5mm square pins. the demo board includes an eeprom that provides non- volatile storage for the ltc5100s configuration settings and parameters. for example, the eeprom stores param- eters for the laser bias and modulation levels as well as temperature coefficients and fault detection modes. the ltc5100 transfers the data in the eeprom to its internal registers at power up. the ltc5100 is designed for hot plugging and can be configured to load the eeprom and enable the transmitter as soon as power is applied. be applicatio s i for atio wu uu careful with this mode of operation! it is possible to leave the eeprom in a state that automatically turns the laser on at power up. the ltc5100s fault output is available at the test turret labeled fault. the fault pin can be software config- ured with several output pull-up options, including open drain. the demo board has three jumpers for enabling the transmitter, observing the electrical eye diagram, and measuring the ltc5100s power supply current. details of the use of these jumpers are given in the dc499 demo manual. 9 10 11 12 4 50 fault j1 sma j2 sma in in + 3 2 1 13 14 15 16 8 7 6 5 u1 ltc5100 u2 24lc00 eeprom 5-lead sot23 package 128 bits v cc scl v ss sda nc 50 r1 49.9 c2 10nf (optional) 1 1 jp2 elec eye remove jumper before attaching a laser diode! termination resistor 50 v dd c3 10nf c1 10nf j3 sma scl gnd sda scl gnd sda h3 gnd md src en 5v p1 sda scl gnd1 gnd2 5v gnd p2 jp1 enable trans 12 h2 md 12 + r5 22.1k c7 0.1 f 2 4 3 5 u4 lt1812 1 1.8v v + v out v r4 10 r3 26.7k c5 10 f d3 r2 22.1k jp3 3a schottky i dd 12 c4 10 f nc u3 lt1762ems8-3.3 v cc 5v 5% 150ma max out sense byp gnd 1 2 3 4 8 7 6 5 in nc nc shdn v dd2 v dd1 d2 3a schottky c6 10 f ltc5100 core applications circuit 5100 f33 l1 blm15ag121pn1d 3.3v 5v v ss in + in v ss v ss moda modb v ss v dd en src md fault sda scl v dd(hs) gnd + + + figure 33. schematic diagram of the dc499 demo board
ltc5100 51 5100f applicatio s i for atio wu uu figure 34 layout of the dc499 demo board (silkscreen and top layer copper) table 31. bill of materials for the dc499 demo board reference quantity part number description vendor telephone 5v, v dd1 , v dd2 , sda, 12 2501-2 1-pin terminal turret test point mill-max (516) 922-6000 scl, fault, en, src, md, gnd(3) c1, c2, c3 3 grp155r71e103ja01 0.01 m f 25v 5% x7r 0402 capacitor murata (770) 436-1300 c4, c5, c6 3 12066d106mat 10 m f 6.3v 20% x5r 1206 capacitor avx (843) 946-0362 c7 1 0603yc104kat 0.1 m f 16v 10% x7r 0603 capacitor avx (843) 946-0362 d2,d3 2 b320a 3a schottky rectifier diode diodes, inc. (805) 446-4800 d4 0 option (no load) n/a (no load) n/a h2, jp1, jp2, jp3 4 2802s-02g2 2mm 2-pin header comm con (626) 301-4200 h3 1 2802s-03g2 2mm 3-pin header comm con (626) 301-4200 j1, j2, j3 3 142-0701-851 50 w sma edge-lanch connector johnson components (800) 247-8256 l1 1 blm15ag121pn1d 0402 ferrite bead murata (770) 436-1300 p1 1 70553-0004 5-pin right angle header molex (630) 969-4550 p2 1 70553-0001 2-pin right angle header molex (630) 969-4550 r1 1 cr05-49r9fm 49.9 w 1% 1/16w 0402 resistor aac (800) 508-1521 r2, r5 2 cr16-2212fm 22.1k 1% 1/16w 0603 resistor aac (800) 508-1521 r3 1 cr16-2672fm 26.7k 1% 1/16w 0603 resistor aac (800) 508-1521 r4 1 cr16-10r0fm 10 w 1% 1/16w 0603 resistor aac (800) 508-1521 u1 1 ltc5100 qfn 4mm 4mm ic ltc (408) 432-1900 u2 1 24lc00 128-bit ic bus serial eeprom 5-pin sot-23 microchip u3 1 lt1762ems8-3.3 low noise ldo micropower regulator ic ltc (408) 432-1900 u4 1 lt1812cs5 op amp with shutdown ic ltc (408) 432-1900 h3 1 ccij2mm-138g 2-pin 2mm shunt comm con (626) 301-4200 information furnished by linear technology corporation is believed to be accurate and reliable. however, no responsibility is assumed for its use. linear technology corporation makes no represen- tation that the interconnection of its circuits as described herein will not infringe on existing patent rights.
ltc5100 52 5100f linear technology corporation 1630 mccarthy blvd., milpitas, ca 95035-7417 (408) 432-1900 l fax: (408) 434-0507 l www.linear.com related parts part number description comments ltc1773 current mode synchronous buck regulator design note 295 high efficiency adaptable power supply for xenpak 10gbps ethernet transceivers ltc1923 high efficiency thermoelectric cooler controller lt ? 1930a 2.2mhz step-up dc/dc converter in 5-lead sot-23 design note 273 fiber optic communication systems benefit from tiny, low noise avalanche photodiode bias supply ? linear technology corporation 2003 lt/tp 0903 1k ? printed in usa u package descriptio uf package 16-lead plastic qfn (4mm 4mm) (reference ltc dwg # 05-08-1692) 4.00 0.10 (4 sides) note: 1. drawing conforms to jedec package outline mo-220 variation (wggc) 2. all dimensions are in millimeters 3. dimensions of exposed pad on bottom of package do not include mold flash. mold flash, if present, shall not exceed 0.15mm on any side 4. exposed pad shall be solder plated pin 1 top mark 0.55 0.20 16 15 1 2 bottom view?xposed pad 2.15 0.10 (4-sides) 0.75 0.05 r = 0.115 typ 0.30 0.05 0.65 bsc 0.200 ref 0.00 ?0.05 (uf) qfn 0802 recommended solder pad pitch and dimensions 0.72 0.05 0.30 0.05 0.65 bcs 2.15 0.05 (4 sides) 2.90 0.05 4.35 0.05 package outline

▲Up To Search▲

Price & Availability of LINEARTECHNOLOGYCORP-LTC5100EUFPBF

	To Download LINEARTECHNOLOGYCORP-LTC5100EUFPBF Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .