Users Manual
MLX90308CCC & MLX90314AB
and
Demo Software
The MLX90308CCC and MLX90314AB sensor interfaces provide a versatile programmable
interface to a wide assortment of sensors and transducers. The sensor interface provides gain and
offset control, linearization and temperature correction for sensor. The device contains on chip
EEPROM for storage of the coefficients for making the signal corrections and adjustments. This
document describes the use of the Melexis software for programming and calibrating the sensor
interface. The MLX90308CCC and the MLX90314AB are identical circuits except that the
MLX90314AB has four times the gain as the MLX90308CCC. The package and pin out are the
same as well as the programming.
Within this document the terms ‘sensor interface’, ‘controller’, ‘chip’, and ‘ASIC’ are used
interchangeably and all refer to the MLX90308CCC or the MLX90314AB.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
1
11/28/01
1) Main Dialog Box
Upon starting the software the dialog box below, figure 1, will appear.
Status Display Window
title bar
Figure 1, Main dialog box.
Software Title
The applicable part number for which the software is intended to work with is located in the title
bar of this dialog box. The software and this document is intended to work with the
MLX90308CCC or MLX90314AB sensor interface.
Software revision
To determine the version of software which is currently running, a right mouse click on the title
bar will cause a pop up menu to be displayed. From this menu, select ‘About test…’. This will
cause the following pop up menu to be displayed, see figure 2.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
2
11/28/01
Software
Revision
Figure 2, Software revision dialog box.
Connect
The ‘Connect’ button will bring up the ‘Find Controller’ dialog box. This is needed to establish
communications between the PC and the sensor interface. A description of this dialog box is in
section 2.
Edit Parameters
The ‘Edit Parameters’ button is used to bring up the ‘Edit Parameters’ dialog box. This dialog
box is used for changing the operating parameters of the sensor interface. This dialog box is
described in section 3.
Reset Controller
The ‘Reset Controller’ button is used to send a reset command to the sensor interface. This is a
soft reset commanded by a read to EEPROM address 63. Address 63 of EEPROM is not a valid
address so no data is returned. Upon a successful reset the sensor interface responds with a BCh.
The software in turn notifies the user in the status display window of the successful completion
of the reset command.
When uploading new parameters to EEPROM, the sensor interface must be reset before the
changes take effect.
Protect Factory Clock Settings
Six bytes in the EEPROM are used to control the internal clock. These bytes are set by Melexis
during final test of the chip. The settings are to ensure that the clock rate is such the non-turbo
baud rate is as close to 2400 as possible. The software protects these bytes, 1, 32-36, and 39
from inadvertent modifications. This check box, default is checked, is used to give the user the
ability change the factory set locations if needed. The datasheet describes the function of each of
the factory set bytes.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
3
11/28/01
Download ASIC
This button downloads all of the data from the sensor interface to the PC. This data is from both
RAM and EEPROM.
The data must be downloaded from the sensor interface before it can be edited using the edit
parameters dialog box
If the data from the sensor interface does not appear to be correct, click on this button several
times and observe the data that is downloaded. If this data changes or is extremely slow to
download then the PC is not communicating at a close enough baud rate to the baud rate of the
chip.
Upload Ram
This button is used to upload only the RAM data to the sensor interface. Any changes to the data
from what is already in the RAM will take effect immediately. The ram data is overwritten with
data from EEPROM when the chip is reset.
Some of the locations used in RAM are for dynamic variables. This means that as soon as the
user loads values into these locations the chip’s firmware overwrites those locations with the
current operating data.
Upload EEPROM
This button is used to upload only the EEPROM data to the sensor interface. The entire block of
data, 48 bytes, is uploaded. The software automatically generates the correct checksum and
places it in byte 47 when the block is uploaded. This occurs whether or not the checksum test
has been selected. Changes to the EEPROM data do not take effect until the sensor interface has
been reset.
Load From File
This button is used to obtain previously saved data from a file. A standard dialog box is used for
locating the desired file to be loaded. Configuration data can be saved from one chip and stored
in another. The protection of the factory programmed clock settings will prevent the loaded file
from overwriting these values as they are unique for each chip. It is important to note that the
settings from one chip will not necessarily yield the same results on another chip. This is due to
the resistors implemented in silicon having about a 20% tolerance.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
4
11/28/01
Save to File
This button is used to save the RAM and EEPROM data to a file. The data can be used as a
starting point for calibrating another sensor-sensor interface. A standard dialog box is called to
allow the user to name the file and location. The data is saved in a simple readable ASCII text
file format with the addresses identified. Saved files can be edited and comments added.
Byte Editor
The Byte editor, see figure 3, allows the user to modify individual bytes in RAM or EEPROM
without specifically labeling the data contained in the byte. The byte editor can access all of the
bytes in RAM and EEPROM. The byte editor is accessed by double clicking on the desired byte
within the data windows on the main dialog box. Once a new data value is entered, clicking OK
sends the data to the sensor interface. The data is shown and set in a hex format.
New data value is entered
Figure 3, Byte editor.
2) Find Controller Dialog Box
The Find Controller dialog box, see figure 4, is used to establish communications between the
PC and the sensor interface. Each sensor interface is set during final test to have an internal
clock frequency that will result in a baud rate of 2400. Also, if the turbo bit is set, the baud rate
will be around 9600.
Find Controller
The Find Controller button initiates the search for the baud rate for the PC to communicate with
the sensor interface.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
5
11/28/01
Serial Port
The serial port pull down menu allows the user to select the serial port the sensor interface is
connected to. The options are Com1, 2, 3, and 4. The software assumes standard port addresses
and IRQs are defined in Windows. The default is Com 1.
Baud Rate
The baud rate pull down menu allows
the user to select a specific baud rate
or a range of baud rates for the
software to search. The default is the
2400 baud rate selection.
Search Status
The search status window informs the
user of the status of the search.
Figure 4, Find controller dialog box
3) Edit Parameters Dialog Box
The Edit Parameters dialog box is the most used dialog box of the software, see figure 5. This is
where the individual operating parameters are identified from the block of EEPROM data. All
user settable parameters are accessed using this dialog box. Also, important data from RAM is
displayed within this box. The items in this dialog box are arranged by functions in areas and
groups.
3.1) Mode
The Mode area of the dialog box contains all of the mode options contained in the mode byte,
byte 0 in EEPROM.
Voltage Mode
This box enables Voltage Mode output of the sensor interface. This is used for both absolute and
ratiometric configurations. Voltage mode can be used simultaneously with Current Mode.
Either Voltage Mode and/or Current Mode should always be enabled.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
6
11/28/01
Current Mode
This box enables Current Mode output of the sensor interface. Voltage mode can be used
simultaneously with Voltage Mode. Either Voltage Mode and/or Current Mode should always
be enabled.
Turbo
This box enables the Turbo mode of the sensor interface. This causes the internal clock rate to
be multiplied by four. The baud rate is also multiplied by four as a result. The faster clock rate
will provide wider bandwidth.
Figure 5, Edit parameters dialog box.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
7
11/28/01
Analog Mode
This box enables the chip to operate with a completely analog signal path from sensor input to
signal output. This provides the best accuracy however signal linearization is not available in
Analog Mode. Analog Mode is mutually exclusive with Digital Mode.
Digital Mode
Digital Mode breaks the signal path by converting the signal to digital data for manipulation by
the internal microcontroller. The signal is then converted back to analog before it is sent to the
output stage of the chip. Both the A to D and the D to A are 10 bits wide. Digital Mode allows
the processor to manipulate the data thus providing signal linearization. Digital Mode has a
smaller bandwidth than Analog Mode and some small quantization noise and error is typical.
Digital Mode is mutually exclusive with Analog Mode.
Level Steering
The Level Steering Mode uses the IO1 and IO2 pins as digital outputs to provide a 2 bit A to D
to an internal analog signal. The analog signal is chosen by the user for the Alarm/Level
Steering Source area of the dialog box. The trigger levels are also defined by the user and are
contained in the same part of the dialog box. This function is mutually exclusive with Alarm
Mode.
Alarm Mode
The Alarm Mode is used to force the output signal to a defined value when an input analog
signal exceeds a user defined range. The input analog signal is chosen from the Alarm/Level
Steering Source area of the dialog box. The defined range is between the upper trigger and the
lower trigger. If the chosen signal exceeds the upper trigger then the output is forced to the value
of the high output. If the chosen signal goes below the low trigger then the output is forced to
the level of the low output. This function is mutually exclusive with Level Steering Mode.
Internal Temp
This box selects the internal temperature sensor for temperature compensating the sensor. The
internal temperature sensor is mutually exclusive with the external temperature sensor.
External Temp
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
8
11/28/01
This box selects the external temperature sensor for temperature compensating the sensor. The
external temperature sensor is mutually exclusive with the internal temperature sensor.
Checksum Test
This box enables the Checksum Test to ensure the integrity of the contents of the EEPROM.
When enabled the checksum is calculated during a reset, either power-up or a soft reset. The
sum of all of the EEPROM locations should be 0FFh for the test to pass. Location 47 in
EEPROM is reserved as the checksum location. This location is set by the software to a value
that forces the sum to 0FFh. The software will calculate and use this value whether or not the
test is enabled. The software updates the checksum whenever the Upload EEPROM button is
used. If the test detects a failure, the output is forced to a value defined by the user, ‘Checksum
Error Output’. The chip will regain normal function when the correct checksum is calculated.
3.2) General Signal Control
The center of the dialog box contains data and control that is needed to control the sensor’s
signal for both analog and digital modes as well as for voltage and current outputs. The diagram
below, figure 6, illustrates the signal path within the sensor interface.
FLT
OFC
CMO
CURRENT
SENSOR
BRIDGE
SENSOR
VBP
OFFSET
IAO GAIN
VBN
10 Bit DAC
10 Bit DAC
Fine Offset
Fine Gain
Temp Gain
Input for
External
Temperature
Sensor
CMN
VOLTAGE
VMO
ADC DAC
Invert
2 Bit DAC
Analog/
GNO Digital
& Offset
2 Bit DAC
2 Bit DAC
Coarse Offset
Coarse Gain
8 Bit
RISC
Microcontroller
1 Bit DAC
Coarse Gain
IO1
IO2
TMP
TPO
Internal
Temperature
Sensor
10 Bit ADC
COMS
Regulator
VDD
TSTB
GND
VDD1
FET
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
9
11/28/01
Figure 6, MLX90308CCC/MLX90314AB Block diagram.
Invert Signal
The invert signal control swaps the VBN and VBP signals at the front end of the signal path.
This has the same effect as rewiring the connections from the sensor.
Temp Amp Gain
The temperature amplifier gain control (GNTP) is a 2 bit control on the gain of the amplifier for
the analog temperature signal. This gain adjustment applies to both the internal or external
temperature sensor.
Coarse Offset
The coarse offset (CSOF) is a 2 bit offset control on the front end amplifier. This provides a
very coarse control on the offset of the signal.
Coarse Gain
Coarse gain (CSGN) controls two parts of the signal path. The most significant bit controls the
output amplifier used for voltage output. The two least significant bits control the gain on the
input stage.
Fixed Gain
Fixed Gain (GN) is a 10 bit value controlling the fine gain of the gain amplifier stage. This
allows the user to precisely calibrate the output span.
Fixed Offset
Fixed Offset (OF) is a 10 bit value controlling the fine offset adjustment to the offset amplifier.
Filter
The filter value is used to filter the temperature value. This is a 4 bit number (0 to 6) where 0 is
not filtering and 6 heaviest filtering. The heavier filtering means that it will take longer for a
change in temperature to impact the filtered temperature value used in the control of the sensor
signal. The 4 MSBs of the byte 25 in EEPROM are kept cleared by the software.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
10
11/28/01
Checksum
The Checksum data box is for display only. The software calculates the checksum whenever an
upload to EEPROM is performed. This happens whether the Checksum test is enabled or not.
This field is read only.
Checksum Error
This data field defines what the output will do if the Checksum test fails. This value is sent to
the D to A directed at the output, either Voltage or Current, irregardless of whether the part is in
analog or digital mode. This is a 10 bit value, 0-1023.
3.3) Alarm/Level Steering
The Alarm and Level Steering part of the dialog box is used to enter the operating parameters for
either function. The data fields are shared because the memory locations where the data is stored
are shared. Alarm Mode and Level Steering Mode are mutually exclusive.
Alarm Mode
Alarm Mode provides the capability range limit the output, see figure 7. The trigger levels and
the alarm levels are programmed by the user. These levels are programmed as 8 bit values which
correlate to 8 MSBs of the digitized source signal or the 8MSBs of the output D/A. The source
signal is also selected by the user. Alarm Mode is mutually exclusive with Level Steering.
Low Trigger
This value is the source signal threshold below which the low output value will be sent to output.
This is programmed as an 8 bit number which is compared to the 8 MSBs of the of the digitized
source signal.
Low Output
This value determines the output when the digitized source signal goes below the low trigger
value. This value is programmed as an 8 bit number correlating to the 8 MSBs of the output (10
bit value).
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
11
11/28/01
Alarm Function
OUTPUT
MLX90308
Output
High
Output
Low
Output
Low
Trigger
High
Trigger
Input Signal
Figure 7, Alarm Function
High Trigger
This value is the source signal threshold above which the high output value will be sent to
output. This is programmed as an 8 bit number which is compared to the 8 MSBs of the of the
digitized source signal.
High Output
This value determines the output when the digitized source signal goes above the high trigger
value. This value is programmed as an 8 bit number correlating to the 8 MSBs of the output (10
bit value).
Source
The source signal for Alarm Mode can be chosen from 6 points which the microcontroller has
access to. These are four internal points; input amplifier out (IAO), gain amplifier out (GNO),
temperature amplifier out (TPO), and voltage mode output (VMO) or two external points IO1
and IO2.
Level Steering
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
12
11/28/01
The Level Steering function provides two digital outputs acting as a two bit analog to digital
converter, see figure 8. The converted analog signal is selected by the user as well as the points
at which the two bits toggle. Level Steering is mutually exclusive with Alarm Mode. The
transition points, levels 1,2,& 3, are defined as 8 bit values. These correspond to the 8 MSBs of
the signal level displayed in the lower left corner of the edit parameters dialog box (also stored in
bytes 60 and 61 of RAM).
Level Steering Function
IO1, IO2
MLX90308
Parameter
1 -1
1-0
0-1
0-0
(I02,I01)
Level 1
Level 2
Level 3
Input
Figure 8, Level steering.
Level 1
Level 1 is the user defined point where IO1, IO2 transition from 0, 0 to 0, 1 with an increasing
source signal. The opposite is true for a decreasing signal, there is no hysteresis.
Level 2
Level 2 is the user defined point where IO1, IO2 transition from 0, 1 to 1, 0 with an increasing
source signal.
Level 3
Level 3 is the user defined point where IO1, IO2 transition from 1, 0 to 1, 1 with an increasing
source signal.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
13
11/28/01
Source
The source signal for Alarm Mode can be chosen from 4 points which the microcontroller has
access to. These are four internal points input amplifier out (IAO), gain amplifier out (GNO),
temperature amplifier out (TPO), and voltage mode output (VMO).
3.4) Temperature & Signal Data
The temperature and signal data is displayed for use by the user in determining the temperature
and linearization coefficients.
Update
The Update button is used to request the Temperature (Temp) and Signal data be updated. The
software reads the appropriate RAM locations and displays the data when this button is used.
Whenever the temp or signal data is needed, the displayed data has to be updated.
Temp Value
The temperature value is displayed from RAM locations 58 and 59. The data is 10 bits and is
right justified in the 16 bit field. The software shifts the data left before being displayed by the
software (divided by 64). The data is only updated when the update button is pressed.
Signal In Value
The digitized value of the sensor’s signal is displayed from RAM locations 46 and 47. While in
Analog Mode, this value is the digitized GNO (gain amplifier output) and is proportional to the
output. In Digital Mode, this data is the 'signal in value' (RAM locations 46 and 47) or the
'signal out value' (RAM locations 60 and 61). These are, respectively, the signal values before
and after the linearization calculation. Setting the PC values to +1.0 (1024 decimal) means that
the 'signal in value' = 'signal out value'. (Both ADC and DAC are 10 bits). The data is only
updated when the update button is pressed.
3.5) Temperature Compensation
The MLX90308’s temperature compensation algorithm is piece-wise with up to 4 temperature
segments, see figure 9. Within each temperature segment (‘gap’), the correction is a first order
calculation. There are separate temperature coefficients for gain and offset.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
14
11/28/01
Temperature
Compensation
OF & GN
Temperature Gaps
There are four temperature gaps possible
with the MLX90308. The gaps are defined
with respect to the filtered digitized
temperature value, displayed in the ‘Temp
Value’ box (saved in RAM locations 58 and
59). The first temperature gap is always
bound at the low end by a digitized
temperature value of zero. The fourth
temperature gap is bound at the high end by
a digitized temperature value of 102310
(3FFhex).
MLX90308
OFTC4
OFTC3
OFTC2
GNTC4
GNTC3
Parameter
OFTC1
1 Gap
GNTC2
i4
GNTC1
Whenever fewer than four gaps are used, the
last gap should have an upper bound of
102310. This is to ensure that if the
temperature exceeds the last point it won’t
enter into an undefined temperature gap.
i3
i2
i1
0
Application's temperature
range
T1
T2
T3
3FFh
Temperature
Figure 9, Temperature Compensation
The first temperature gap is slightly different than the other three. The compensation is based on
the temperature difference between the current temp and the T1 point (the upper end of the gap).
The other gaps use the lower end of the gap to measure the temperature difference. For instance
in the second gap, the temperature difference is current temperature minus T1. This difference
has to be accounted for in the compensation procedure.
Temperature Points
The temperature points T1, T2 and T3 are defined by the user to differentiate between the four
possible temperature compensation ranges. The low endpoint T0 is defined as the minimum
digitized filtered temperature value, zero. This means that the coefficients for the first gap,
OFTC1 and GNTC1, will apply to the signal until the digitized filtered temperature reaches zero.
The high endpoint T4 is defined as the maximum digitized filtered temperature value, 102310.
This means that the coefficients for the last gap, OFTC4 and GNTC4 will apply to the signal
from the T3 point and up until the digitized filtered temperature reaches its maximum, 102310.
When defining these points, the number that is used is simply copied from the ‘Temp Value’ box
at the desired temperature (contents of ram locations 58 and 59).
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
15
11/28/01
GNTC
The GNTC values are the gain temperature compensation values for each of the four temperature
gaps, GNCT1, GNTC2, GNTC3, and GNTC4. These values are used to adjust the signal up or
down based on the relative temperature within a particular gap. Mathematically this is
represented by the following equation:
DAC_GAIN = CSGN * {GN[9:0] + [GNTCi *∆Ti]}
Where:
∆Ti = (T1 – temp_f) for gap 1
∆Ti = (temp_f – Ti-1) for gaps 2,3,and 4
Temp_f = Filtered temperature
DAC_GAIN = The digital value that adjusts the gain on the gain amplifier.
Ti = Temperature segment point i = 2, 3, or 4
GNTCi = Gain TC for a given temperature segment i,
CSGN = Course Gain.
GN[9:0] = Fixed Gain (doesn’t change with temperature)
Figure 10, 12 Bit fixed point numbers
The GNTC and OFTC values are 12 bit
fixed point signed numbers, see figure
10. That means that the most significant
bit is a sign bit. The next bit is one or
zero (whole number). The remaining 10
bits are to the right of the decimal point.
This gives the variable the range of –
1.9990234 to +1.9990234. The table
below shows how the fixed point
numbers are represented as the binary
number progresses from 0 to 4095.
Decimal
Value
Hexadecimal
Equivalent
Real Number
Equivalent
0
0000h
+0.00
1023
3FFh
+0.9990234
1024
400h
+1.000
2047
7FFh
+1.9990234
2048
800h
-0.000
3071
0BFFh
-0.9990234
3072
0C00h
-1.000
4095
0FFFh
-1.9990234
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
16
11/28/01
OFTC
The OFTC values are the offset temperature compensation values for each of the four
temperature gaps, OFCT1, OFTC2, OFTC3, and OFTC4. These values are used to adjust the
signal’s offset up or down based on the relative temperature within a particular gap. The offset
adjustment is additive from one gap to the next. That is the offset adjustment for the fourth gap
is added to the total offset adjustment of the third gap which is also added to the total offset
adjustment of the second gap. Mathematically this is represented by the following equation:
For the first temperature gap:
DAC_OFFSET = OF[9:0] + OFTCi * (T1 – Temp_f).
For the second temperature gap:
DAC_OFFSET =OF[9:0] + OFTC2 * (Temp_f – T2)
For the third temperature gap:
DAC_OFFSET = OF[9:0] + [OFTC2 * (T2-T1)] + OFTC3 * (Temp_f – T3)
For the fourth temperature gap:
DAC_OFFSET =OF[9:0] + [OFTC2 * (T2-T1)] + [OFTC3 * (T3 – T2)] +
OFTC4 * (Temp_f – T4)
OF[9:0] = Fixed offset
3.6) Signal Linearization
The MLX90308’s signal linearization algorithm is piece-wise with up to 5 signal segments, see
figure 11. Within each signal segment (‘gap’), the correction is first order. There are separate
correction coefficients for each gap.
There are up to five signal gaps possible with the sensor interface. The gaps are defined with
respect to the signal’s digitized value, displayed in the ‘Signal Value in’ box and 'Signal value
out'-box.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
17
11/28/01
Signal Linearity
Correction
Output
MLX90308
PC5
PC4
Output (units)
PC3
PC2
PC1
P1
0
P2
P3
P4
P5
Input Signal (units)
Figure 11, Signal linearization
The signal linearization is an adjustment to the output level based on the input signal level.
Adjustment coefficients can be set for five different signal ranges. The adjustment is a first order
correction based on the relative signal level within the particular gap plus the total correction of
the previous gaps. The output is obtained by the following formula:
Output = (Signal – Pi) * Pci + Pi where
Signal = input signal measurement;
Pi = programmed signal point (i=2,3,4,5).
Pci = programmed coefficient.
The PCi coefficients are 12 bit fixed point signed numbers, the same format as the OFTC and
GNTC variables. The Pi values are coded on 10 bits (0-3FFh). These values are copied from
the Signal Value box.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
18
11/28/01
4) Programming
4.1) Baseline Calibration
The baseline calibration involves setting the operating modes and temperature gain along with
the coarse and fixed gain and offset.
Setting the operating modes Voltage and/or Current mode, turbo mode, Analog or Digital mode,
Level steering or alarm mode (digital mode only), internal or external temperature sensor, and
Checksum test. The checksum test, if used, should be enabled after all other settings are
complete.
The invert signal bit swaps the differential inputs to the signal path. This has the same effect as
swapping the connections to the VBP and VBN pins of the chip.
Temperature amplifier gain, GNTP, should be set such that the analog converter doesn’t underrun or over-run at the temperature extremes. When using the internal temperature sensor over
the full temperature span of the device (-40C to +140C) typically a GNTP setting of 2 will work.
When using an external temperature sensor the voltage on the TMP pin must stay within the
ranges as described in the datasheet. The more of the voltage range used the greater the
temperature compensation adjustment range will be.
The coarse gain and fixed gain should be set before the coarse offset and fixed offset. Gain and
offset are inter-related, the offset is multiplied by the gain. It is much easier to program the gain
first then offset. It may be necessary to make some minor adjustment to the coarse offset settings
before adjusting the gain. This is only needed if the output clips at either high end or low end. It
is difficult to precisely calculate the offset and gain values. The amplifier circuitry within the
chip uses resistors implemented in silicon. These resistors have around a 20% tolerance, thus the
gain and offset will vary from chip to chip. Each chip is tested to provide a gain and offset
adjustment capability within a specified range. For calculations a typical value can be taken.
4.2) Temperature calibration
The temperature compensation capability of the 90308/90314 is piece-wise with first order
compensation in each segment (gap). The compensation is based on the difference between the
current digitized filtered temperature and the appropriate temperature point. The equations have
been previously described. The first temperature gap is slightly different than the other three.
The first gap uses the temperature difference between the current temperature and the T1
temperature point (the upper end of the temperature gap). The second temperature gap also uses
the T1 point for determining the temperature differences (the low end of the second gap). The
third and fourth gaps also use the temperature point at the low end of their gap. This means that
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
19
11/28/01
programming the temperature compensation for the first gap is slightly different than the other
gaps. The compensation coefficients for the first gap (OFTC1 and GNTC1) apply to digitized
filtered temperature values from T1 down to zero. The fourth gap coefficients apply to digitized
filtered temperature range from T3 to 1023 (decimal).
The temperature points T1 thru T3 should always be in increasing order from T1 to T2 to T3. If
the temperature sensor has increasing signal with increasing temperature then the compensation
procedure is intuitively easy. This is the case with the internal temperature sensor. If the
temperature sensor has decreasing signal with increasing temperature then the compensation will
start at a hotter temperature and go towards cold. The procedure below is written with regard to
the filtered digitized temperature not the real physical temperature.
4.3) Procedure
1) Set the desired operating mode and number of gaps. It is also helpful to set T2 and T3 to
1023.
2) At the T1 temperature do a baseline calibration, the span then the offset by using CSGN[2:0]
and GN[9:0] then CSOF[1:0] and OF[9:0]. The T1 temperature is somewhat arbitrary, typically
its room temperature. If the temperature sensor goes down with increasing temperature then the
T1 value will be towards the high end of the application’s temperature range. Update the
Temp/Signal values. Copy the digitized temperature reading to the T1 box. If one temperature
gap is desired then T1 will be the highest digitized temperature value.
3) Lower the temperature to the lowest operating temperature and recalibrate the output using
GNTC1 and OFTC1 only. If the temperature sensor goes down with increasing temperature then
raise the temperature to its maximum value and re-calibrate the output using GNTC1 and
OFTC1. Do not change the T1, OF, or GN values.
4) Raise the temperature until the second temperature point, T2. If the temperature sensor goes
down with increasing temperature then lower the temperature to the T2 value. This point may be
arbitrary and can be based on how far the output has deviated from it's desired value. Recalibrate the sensor by adjusting GNTC2 and OFTC2 only.
5) Update the Temp/Signal values displayed on the screen. Copy the digitized temperature value
to the T2 box. This must happen AFTER step 4.
6) Repeat steps 4 and 5 for T3 and setting GNTC3, OFTC3. T1, T2, and T3 must be in ascending
order.
7) For the last temperature gap, raise the temperature to the highest operating temperature
point and re-calibrate the output using GNTC4 and OFTC4. There is no T4, it is assumed to be
1023 (the maximum value of the digitized filtered temperature).
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
20
11/28/01
For less than four temperature gaps the last temperature point should be set to 1023. This means
that the last gap extends out to the end of the temperature range.
4.4) Signal linearization
The signal linearization is only available when the device is in digital mode. This means that the
entire calibration procedure should be done in digital mode. A calibration in analog mode may
not yield a valid calibration in digital mode.
It is the best to first do the baseline calibration then linearize the sensor signal and then then
temperature compensation. The baseline offset calibration must be perform in that way that the
digitized signal has the value 0 or 1 LSB ( be careful with clamping at 0 value).
The total number of linearization gaps plus the temperature gaps cannot be more than five. This
is because the temperature coefficients and the linearization coefficients share the same memory
locations.
1) Set all of the PC values to 1024. This corresponds to a real value of +1.000. Also, set P2 thru
P5 to a value of 1023. (P1 = 0)
2) With no signal (zero pressure) find the correct value of P1 to have the correct voltage on the
output.
3) Raise the signal to the signal point (pressure point) P2. This is an arbitrary value (can be
based on where the output varies from the desired output by a set amount). Adjust the output to
the desired level by adjusting PC1.
4) Set P2 by copying the pressure value by copying the digitized pressure value to the P2
location. This must happen after step 3.
5) Repeat steps 3 and 4 for P3, P4, and P5 setting PC2, PC3, and PC4.
6) Raise the pressure to its maximum and set PC5
If fewer than 5 pressure gaps are used the last P value (highest) should be set to 1023.
MLX90308CCC & MLX90314AB Software Users Manual
Rev 2.0
21
11/28/01