data sheet supplement differential two-wire hall effect sensor ic tle4942 -2 TLE4942-2c for all parameters not specified in this document the tle4942 data sheet is valid. p-sso-2-1 type marking ordering code package TLE4942-2 4202e4 q62705-k633 psso2-1 TLE4942-2c 42c2e4 q62705-k630 psso2-2
TLE4942-2, TLE4942-2c supplement TLE4942-2 - TLE4942-2c data sheet supplement february 2002 2 1 functional description offset correction peak detection offset = ( max + min) / 2 uncalibrated mode calibrated mode d1 initial settling time d1 d2 d2 d2 standstill pulse (cut off) pga switching 0 1 2 3 4 5 6 7 fig. 1: example for start-up behaviour uncalibrated mode: occasionally a short initial offset settling time t d,input might delay the detection of the input signal. (the sensor is "blind"). the magnetic input signal is tracked by the speed adc and monitored within the digital circuit. for detection the signal transient needs to exceed a threshold (digital noise constant d1). when the signal slope is identified as a rising edge (or falling edge), a comparator is triggered. the comparator is triggered again as soon as a falling edge (or rising edge respectively) is detected ( and vice versa). depending on the initial state of the comparator the ic output is first triggerd on the first or second detected edge. between the startup of the magnetic input signal and the time when its second extreme is reached, the pga (programmable gain amplifier) will switch to its appropriate position. this value is determined by the signal amplitude and initial offset value. the digital noise constant value is changing accordingly (d1 ? d2), leading to a change in phase shift between magnetic input signal and output signal. after that consecutive output pulses should have a nominal delay of about 180. during the uncalibrated mode the offset value is calculated by the peak detection algorithm as described in the tle4942 data sheet. transition to calibrated mode: in the calibrated mode the phase shift between input and output signal is no longer determined by the ratio between digital noise constant and signal amplitude. therefore a sudden change in the phase shift may occur during the transition from uncalibrated to calibrated mode.
TLE4942-2, TLE4942-2c supplement TLE4942-2 - TLE4942-2c data sheet supplement february 2002 3 calibrated mode: see tle4942 data sheet. additional notes: unlike the tle4942 the first output pulse might occur before the first zero-crossing of the magnetic input signal. therefore the maximum number of edges until the calibrated mode is active is increased by one for TLE4942-2. however, referring to the input signal the delay between startup of the signal and first calibrated output signal is identical with tle 4942. typically the phase error due to pga-transition (row 7 to 15) reduces the error caused by switching the mode from uncalibrated to calibrated. in very rare cases a further pga switching can occur during the calibration process. it can take place when the signal is extremely close to a pga switching threshold. this additional switching might delay the transition to calibrated mode by two to three more pulses. the probability of this case is mainly depending on variations of magnetic amplitude under real automotive conditions. (see appendix b) the direction detection feature is active in the calibrated mode only. therefore identical to tle4942 the correct direction information is available after the first three output pulses in calibrated mode. regarding the rare case mentioned before combined with other initial conditions this may lead to a worst case of 11 pulses before correct direction information is guaranteed. the typical value is 5 pulses.
TLE4942-2, TLE4942-2c supplement TLE4942-2 - TLE4942-2c data sheet supplement february 2002 4 1 0 1 2 3 4 5 6 7 8 9 10 2 45s...737 ms 180 ( u n cal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 180 (cal) 180 (cal) 180 (cal) 3 45s...737 ms 45s ...180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 180 (cal) 180 (cal) 4 45s...737 ms 45s ...180 ( uncal) 180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 180 (cal) 5 45s...737 ms 45s ...180 ( uncal) 180 ( uncal) 180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 6 45s...737 ms 45s ...180 ( uncal) 180 ( uncal) 180 ( uncal) 180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 7 45s...737 ms 180...300 (pga, uncal) 180...220 180 ( uncal) 180 ( uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 180 (cal) 180 (cal) 8 45s...737 ms 180...300 (pga, uncal) 180 ( uncal) 180...220 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 180 (cal) 9 45s...737 ms 180...300 (pga, uncal) 180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 10 45s...737 ms 180...300 (pga, uncal) 180 ( uncal) 180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 11 45s...737 ms 180...300 (pga, uncal) 180 ( uncal) 180 ( uncal) 180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 12 45s...737 ms 45s ...180 ( uncal) 180...260 (pga, uncal) 180...220 180 ( uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 180 (cal) 180 (cal) 13 45s...737 ms 45s ...180 ( uncal) 180...260 (pga, uncal) 180 ( uncal) 180...220 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 180 (cal) 14 45s...737 ms 45s ...180 ( uncal) 180...260 (pga, uncal) 180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) 15 45s...737 ms 45s ...180 ( uncal) 180...260 (pga, uncal) 180 ( uncal) 180 ( uncal) 180 ( uncal) 90...270 150..200 (cal/ uncal) 150..200 (cal/ uncal) 180 (cal) table1: overview of the startup-behaviour. in the first row the pulse number is given. pulse number 0 is the last pulse before signal startup, e.g. the standstill (stopped) pulse. the following rows show different possibilities for the nominal delays between the pulses. numbers are calculated for sinusoidal input signals. additionally the specified tolerances have to be taken into account (e.g. jitter) rows 2..6: behaviour at small input amplitudes ( d b< approx. 3.5mt) rows 7..11: behaviour at initial phases of ?90 .. 0 rows 12..15: behaviour at initial phases of 0.. 90 remark: the additional pga switching can only occur once per row. therefore also the additional phase shift marked "150..200 (cal/ uncal)" will only occur once per row. (see example)
TLE4942-2, TLE4942-2c supplement TLE4942-2 - TLE4942-2c data sheet supplement february 2002 5 example: the 14th row describes the behaviour shown in fig. 1: the standstill pulse length can be cut by the first detected speed pulse, therefore the minimum distance between the rising edges will be 45s. the distance between the first and second detected speed pulse is determined by the initial signal phase and amplitude and a possible first pga switching. as the first pulse length also can theoretically be cut off by the following pulse, the minimum distance could be 45s. the rising edge between the first signal minimum and the first signal maximum can cause the pga switching into a lower gain range. as a result the digital noise constant value can increase in relation to the signal amplitude. that typically leads to an increased delay between the second and the third pulse, its maximum value is 260. the following minimum and maximum are necessary for peak detection. after offset correction, the delay between the 5 th and the 6 th pulse can have a maximum value 270. as this marks the transition from uncalibrated to calibrated mode, the following consecutive pulses (4, 5, 6 ...) will be spaced 180 nominally. same example with numbers : d b = 10mt sin ( w t + j ). j = 30 typical startup-behaviour at a sinusoidal input signal of 10mt amplitude, initial phase= 30. j 1 43,6 2 133,9 3 333,2 4 513,2 5 693,2 6 900 7 1080 8 1260 9 1440 10 ... d j 90,3 (pga, uncal) 199,3 (pga, uncal) 180 ( uncal) 180 ( uncal) 206,8 ( offset- correction) 180 (cal) 180 (cal) 180 (cal) 180 (cal) ? this corresponds to row 14 in the table, behaviour similar to fig. 1 as a special (and rare) case instead of an offset correction after edge number 5, a further (extra) pga switching could occur before edge number 5. pga switching inhibits an immediate offset update. it can happen if one of the signal peaks is exactly at a pga switching threshold (speed-adc overflow). in this case the offset update (switching from uncalibrated mode to calibrated mode) would be delayed by two to three further edges. the referring phase shifts of the example would then be as follows: j 1 43,6 2 133,9 3 333,2 4 513,2 5 727,5 6 907,5 7 1087,5 8 1260 9 1440 10 ... d j 90,3 (pga, uncal) 199,3 (pga, uncal) 180 ( uncal) 214,3 (extra pga) 180 ( uncal) 180 ( uncal) 172,5 ( offset- correction) 180 (cal) 180 (cal) ? this corresponds to row 13 of the table.
TLE4942-2, TLE4942-2c supplement TLE4942-2 - TLE4942-2c data sheet supplement february 2002 6 circuit description see tle4942 data sheet
TLE4942-2, TLE4942-2c supplement TLE4942-2 - TLE4942-2c data sheet supplement february 2002 7 2 additions/changes for tle4942?2 versus tle4942 (all values are valid for constant amplitude and offset of input signal, f<2500hz) parameter symbol min. typ. max. unit conditions signal behaviour after undervoltage or standstill>t stop n dz -start 2 3 pulses pulses magnetic edge ampl i tude according to startup early limit, b ? d edges that occur b e fore n dz -start can be suppressed 1hz f 2000hz f > 2000hz t d,input has to be taken into account 38 45 52 s shortest time delay between pulse 0 (stop pulse) and pulse 1 t d,input has to be taken into account 38 45 52 s shortest time delay between wheel speed pulse 1 and 2 systematic phase error of output pulses during startup- and uncalibrated mode -88 +88 systematical phase error of ? uncal? pulse; nth vs. n+1th pulse (does not include jitter) phase shift change during pga switching 0 80 phase shift change du r ing transition from uncalibrated to calibrated mode d f switch -90 +90 number of pulses in uncalibrated mode n dz-startup 6 pulses in rare cases (see appendix b) n dz-startup 8 pulses number of pulses with invalid su p plementary info r mation n dr-startup 9 pulses after n dr-startup pulses the supplementary information is correct (starting with the n dr-startup +1-th pulse the pulse length is correct) in rare cases (see application notes) n dr-startup 11 pulses
TLE4942-2, TLE4942-2c supplement TLE4942-2 - TLE4942-2c data sheet supplement february 2002 8 parameter symbol min. typ. max. unit conditions jitter during startup and u n calibrated mode s jitclose (1 s -value) 3 4 % % -40c t amb 150c: 150c t amb 170c: s jitfar (1 s -value) 5 7 % % -40c t amb 150c 150c t amb 170c s jitac (1 s -value) 3 % see tle 4942 spec ma g netic field amplitude change necessary for early startup of the ?2 ve r sions startup early limit, b ? d % x b ? ? * 2 limit + > (x = 10) startup early limit, b ? d 0.7 0.7 1.6 +10% 1.76 3.0 +10% 3.3 mt mt these magnetic field changes are necessary for startup with the second magnetic edge permitted time for edges to exceed startup early limit, b ? d startup early slow limit, t d 590 ms nece s sary for startup with the second edge f < 1s behaviour at magnetic input signals slower than t stop (self-calibration time period): unlike the tle4941 magnetic changes exceeding startup early limit, b ? d can cause output switching of the tle4941-2, even at f significantly lower than 1hz. depending on their amplitude edges slower than startup early slow limit, t d might be detected. if the digital noise constant ( startup early limit, b ? d ) is not exceeded before a new initial self-calibration is started, the output of the corresponding edge will be inhibited. this depends on signal amplitude and initial phase. 3 additional remarks all additional parameters for TLE4942-2 are guaranteed by design, based on lab chara c terisations. for series production additional to the parameters of tle4942 (standard type) only n dz -start is tested.
TLE4942-2, TLE4942-2c supplement TLE4942-2 - TLE4942-2c data sheet supplement february 2002 9 appendix b: TLE4942-2 application notes release 1.0 1. occurrence of initial calibration delay time t d,input identical to tle 4942, tle 4942 c application notes. 2. magnetic input signal extremely close to a pga switching threshold during signal startup: after signal startup normally all pga switching into the appropriate gain state occurs within less than one signal period. this is included in the calculation for n dz-startup . for the very rare case that the signal amplitude is extremely close to a switching threshold of the pga and the the full range of following speed adc respectively, a slight change of the signal amplitude can cause one further pga switching. it can be caused by non-perfect magnetic signal (amplitude modulation due to tolerances of pole-wheel, tooth wheel or air gap variation). this additional pga switching can result in a further delay of the output signal ( n dz-startup ) up to three magnetic edges leading to a worst case of n dz-start =9 and n dr-startup =11. however, the speed signal startup, comprised of n dr-startup and t d,input is not affected by this behaviour for tle 4942-2. 3. - 6. identical to tle4942, tle4942c application notes.
|
