The dataTaker range of data loggers have 2 different types of analog input channel multiplexers, solid state (CMOS) multiplexer or relay multiplexer.

The models of the dataTaker data loggers which have each of these multiplexer types are as follows

Solid State (CMOS) Multiplexer
- dataTaker 50, 500 and600

Relay Multiplexer
- dataTaker 505, 605, 515 and615
- Channel Expansion Module (CEM-AD)

Throughout this section reference will be made to solid state or relay multiplexing, rather than to individual dataTaker models. Confirm from the above list which dataTaker model you are using.

The analog processing circuits of the dataTaker consist of four principal sections

a multiplexer for selecting the analog input channels for sampling

a multiplexer for selecting the measuring circuits and input configuration appropriate to each input signal type

a precision instrumentation amplifier with programmable gain

an analog to digital converter (ADC)

This chapter describes the operation of each of these sections, and the degree of influence that the user has over how each of these sections function.

Analog Input Multiplexer

The analog input channel multiplexer for the dataTaker is an array of either solid state (CMOS) analog gate devices or small electromechanical relays, that provide the analog input channel switching.

Each of the analog input channels is in effect a five pole switch, for which each pole is normally open.

These switches select the analog input signal to be measured, configure the channel to measure the input signal, and direct any required support to the sensor.

The switches for each of the analog input channels has the following functions

Select the +ve input

Select the -ve input

Direct the excitation source to the channel

Select the 'active' end of the 100.0 Ohm current shunt.

Select the 'grounded' end of the 100.0 Ohm current shunt.

A simplified schematic for the analog multiplexer switching of a single analog input channel of the dataTaker is illustrated in Figure 22.

There are additional multiplexer or Selector devices in the analog input circuits, which select the various support circuits during measurement of an input signal, and select the various internal calibration channels during an autocalibration.

 

 

Figure 22 - dataTaker Simplified Analog Input Channel

 

Each analog input channel also has input protection resistors, which are not shown in Figure 22 for simplicity.

The major difference between the two multiplexer types are the common mode voltage and the voltage ranges that the dataTakers with each multiplexer type can measure directly.

Solid State Multiplexer

The dataTaker 50, dataTaker 500 and dataTaker 600 data loggers have a solid state multiplexer, which uses CMOS 4052 quad analog gates.

The dataTakers which have the solid state multiplexer have three decades of voltage input ranges of  ±25.000 mV,  ±250.00mV and  ±2500.0 mV.

The 4052 devices have a maximum input voltage of approximately 20 Volts, and will withstand common mode voltages of up to ±3.5 Volts.

The multiplexer also provides protection for the dataTaker against excessive input voltages, because the 4052 devices will usually fail safe when damaged.

The multiplexer devices are socketed for easy replacement ñ see Appendix D for details of replacing the multiplexer devices in the field.

Relay Multiplexer

The dataTaker 505, dataTaker 515, dataTaker 605 and dataTaker 615 data loggers all have a relay multiplexer, which uses high precision electromechanical relays.

The dataTakers which have the relay multiplexer also have internal input attenuation which provides two categories of voltage input range as follows

when the internal attenuators are not selected, there are three decades of low voltage input ranges of  ±25.000 mV,  ±250.00mV and  ±2500.0 mV

when the internal attenuators are selected, there are three decades of high voltage input ranges of  ±7.000 V,  ±70.00 V and  ±100.0 V

The relay multiplexer has high withstanding input voltages when the contacts are open as follows

1.5 KVolts for 10 µS

500 Volts for 50 mS

100 Volts indefinitely

Instrumentation Amplifier

The precision instrumentation amplifier of the dataTaker is a high performance device, with three programmable gain stages that provide the three decade input ranges of the analog input channels. The instrumentation amplifier has gains of 1, 10, and 100 that correspond to voltage input ranges.

The dataTakers with the solid state multiplexer have three decades of voltage input ranges of  ±25.000 mV,  ±250.00mV and  ±2500.0 mV.

The dataTakers with the relay multiplexer also have internal input attenuation which provides two categories of voltage input range as follows

when the internal attenuators are not selected, there are three decades of low voltage input ranges of  ±25.000 mV,  ±250.00mV and  ±2500.0 mV

when the internal attenuators are selected, there are three decades of high voltage input ranges of  ±7.000 V,  ±70.00 V and  ±100.0 V

Autoranging by Instrumentation Amplifier

The programmable gains of the instrumentation amplifier are automatically selected by the dataTaker, according to the level of each input signal to be measured. It is not necessary for the user to define ranges for measuring analog input signals.

During operation the dataTaker manages the instrumentation amplifier gain as follows

when the dataTaker is initially programmed to scan a group of analog channels, each channel is set to an initial gain which is optimal for the particular input signal type being measured.

as each input channel is sampled, the analog to digital converter reads the channel at the previously defined gain. If the first sample is not within this range, then the gain is changed accordingly and the channel is sampled again. This process continues until an in-range or over-range sample is obtained.

when an in-range sample is obtained, the gain setting is saved for that channel, and becomes the initial gain setting when the channel is next sampled.

when an over-range sample is obtained, the gain is set to one and becomes the initial gain setting when the channel is next sampled. An over-range error is generated.

Autoranging occurs at thresholds of approximately 15% and 120% of the full scale for the selected gain, for a gain increase and gain decrease respectively.

The tendency for 'gain hunting' is reduced by a hysteresis of approximately 10% around the threshold.

When the input signal is greater than the full scale threshold of the highest gain range, then an ëE11ñinput(s) out of rangeí error message is generated. A value of -99999.9 units or +99999.9 units is returned, depending on the polarity of the input signal.

Locking the Gain

The instrumentation amplifier gain can be locked for individual analog input channels by the Gain Lock channel option, which is declared as a part of the individual channel specifications in the schedule lists of Schedules.

When the gain is locked, the instrumentation amplifier is preset to the defined gain setting and does not autorange. All measurements for that channel are made at the set gain. Locking the gain for a channel does not affect autoranging for other channels.

The gain is locked by a channel option command

channel(GLn)

where GL specifies Gain Lock, and n is the gain required.

The gain can be locked by direct command from DeTransfer as part of the channel specification in a schedule list. For example

will lock the instrumentation amplifier gain to 10 (±250.00 mV).

In DeLogger, the gain lock setting for each analog channel is specified in the Program Builder via Channel Options : GainÖ

 

 

The Input Bias Current

The instrumentation amplifier of the dataTaker requires only a small input bias current, typically less than 10 nanoamps, which must be supplied by the sensor providing the input signal.

By default, the dataTaker terminates most voltage type inputs to ground via internal
1 MOhm resistors. These input termination resistors provide a path for the bias current. However the input termination resistors may be switched out of circuit if required (See Section II ñ The Analog Input Channels).

If a current path is not provided, then the common mode input range will be exceeded and erroneous readings will result.

Some sensors such as pH and specific ion glass electrodes have a high output impedance, and are unable to supply the required input bias current. Such sensors should therefore be connected to the dataTaker via a high impedance buffer amplifier.

However if the input signal source is either connected directly to the ground of the dataTaker, or is related to the ground of the dataTaker, and the common mode range is not exceeded, then there is generally no difficulty in providing the bias current.

The Analog to Digital Converter

The analog to digital converter of the dataTaker data loggers is a voltage controlled oscillator (VCO) type. This conversion technique has several advantages over other techniques as follows

the input is integrated over a programmable sampling period, providing high rejection of input noise

trade-offs are possible between speed and resolution

the frequency counter also allows measurement of frequency and period signals

The main limitation of VCO type analog to digital converters is the conversion speed.

However the VCO technique removes the need for cumbersome input signal filtering for mains or line hum. Mains or line induced hum is a prominent source of noise in most measurement systems, often producing serious measurement errors. The analog to digital converter of the dataTaker is designed to minimize errors produced by mains or line induced noise, by sampling the input channels over a period of one mains or line cycle.

The analog to digital converter of the dataTaker is optimized to the extent that while an input signal for a channel is being converted, the raw data for the previous channel is being calculated through to appropriate engineering units. Calculation time for each analog channel is approximately 5-10 mS depending on whether any linearization is required for thermocouples, RTDs, etc. or if polynomials have to be applied to the data.

Timing diagrams for the sampling of the analog input channels, the calculation of the primary data, and the linearization of supported sensor types, are presented in detail in the Appendix B at the back of this manual.

ADC Sample Period and Line Frequency

The analog to digital converter of the dataTaker data loggers is designed to sample an input channel for one mains or line cycle period. One mains or line cycle is a period of 16.67 mS for the 60 Hz line frequency, and 20 mS for the 50 Hz line frequency.

Sampling inputs over one line or mains frequency period maximizes rejection of line or mains hum that may have been induced into the sensor cables, etc. As a consequence of this technique, the conversion period or sampling rate of the analog to digital converter is a function of the local line frequency.

The local line frequency must be defined to the dataTaker, and is set via the Line Frequency switch of the DIP switch. The line frequency should be set to 50 Hz for Australia, Europe, Asia, New Zealand, etc, and to 60 Hz for Canada and USA.

For 50Hz operation the Line Frequency switch must be set OFF, while for 60Hz operation the Line Frequency switch must be set ON.

The DIP switch of the dataTaker is accessed by removing the top cover of the logger. The settings for the Line Frequency switch is illustrated below in Figure 23.

 

 

Figure 23 - The Line Frequency Switch of the DIP Switch

 

The dataTaker is factory set to a line frequency of 60 Hz (Line Frequency switch ON) for North America, and 50 Hz (Line Frequency switch OFF) for all other countries.

Increasing the ADC Sampling  Rate

Increasing the setting of the line frequency for the dataTaker has the effect of decreasing the ADC sampling period, and as a consequence increases the ADC sampling rate.

It is not mandatory to set the mains or line frequency to the local line frequency if noise induced from line hum is not a problem. Therefore higher line frequencies may be selected in a quest for increased sampling speed.

However as the setting for the line frequency is increased and so the sampling period decreased, the resolution and accuracy of the ADC decreases.

The ADC resolution and the defined line frequency are related by the relationship

R = 3100000/F_line

where the resolution R is the number of quantizing levels in the measuring range, and F_line is the line frequency.

This ADC resolution can also be expressed in terms of the input voltage resolution by the formula

Input Resolution        = (V_fs x F_line)/3,100,000

where V_fs is the full scale voltage and F_line is the line frequency

For the 25 mV range and 50 Hz line frequency

Input Resolution        = (0.025 x 50)/3100000

                            = 0.4 mV

 

For the 2.5 Volt range and 60 Hz mains/line frequency

Input Resolution        = (2.50 x 60)/3100000

                            = 48.3 mV

 

The mains or line frequency can be set to frequencies other than 50 Hz or 60 Hz by the Parameter11 command (See Section III ñ Parameter Commands).

Using DeTransfer, the mains or line frequency can be set by the command for example

which sets the line frequency to 100 Hz, and therefore sets the period for analog to digital conversion to 10 mS.

The current setting for the mains or line frequency can be read in DeTransfer by the command

Using DeLogger, the mains or line frequency can be set and read using these same commands either

directly, in the Entry Screen of the Text View

as part of a program, in the Pre Schedule Initialization Commands under the Special Commands button of the Settings tab of the Program Builder

 

 

The line frequency can be set within the range of 48 Hz and 1000 Hz.

Increasing the line frequency beyond 200 Hz will not increase the conversion speed appreciably, because the dataTaker spends at least 5 mS per channel to scale the data into engineering units.

Timing diagrams for input channel sampling, the calculation of primary data and the linearization of supported sensor types are presented in the Appendix B at the back of this manual.

The ADC Settling Period

Whenever an input channel is selected for sampling, a settling period or delay is introduced to allow the input to stabilize before the sample is taken.

There are two reasons for varying the ADC settling period

sensors with an output impedance of greater than 50 KOhm will require a longer settling time, because of the small internal capacitances of the dataTaker input channels and ADC

the ADC settling period provides a means for modifying the sampling period or sampling rate. This is mainly useful when sampling one or two channels faster than once per second.

The ADC settling period is 10 mS by default, and may be set to any period in the range of 0 mS to 30000 mS in increments of 1 mS by the Parameter10 command (See Section III ñ Parameter Commands).

Using DeTransfer, the ADC settling period is set by the command for example

sets the ADC settling period to 20 mS.

The current setting for the ADC settling period is read in DeTransfer by the command

Using DeLogger, the ADC settling period is set and read using these same commands either

casually in the Entry Screen of the Text View

as part of a program, in the Optional Initialization Commands under the Special Commands button of the Settings tab of the Program Builder

Altering the ADC settling period alters the ADC sampling rate. This effect is separate from the effect on ADC sampling rate of changing the line frequency.

Autocalibration of the ADC

The analog to digital converter of the dataTaker is automatically recalibrated to compensate for any temperature or ageing drifts in the ADC circuits. Such drifts will result in measurement errors, particularly when low level signals are being measured.

The input zero voltage, the gain, and the current source for resistance measurement, are corrected by autocalibration. An autocalibration is carried out in response to changes in the measured internal zero voltage.

An autocalibration can also be forced at any time by enabling the Autocalibration Switch command /K (see below).

The Number of Calibration Samples

During each autocalibration cycle, each internal calibration channel is sampled a number of times, and the samples are averaged.

Averaging of the samples from calibration channels provides for an accurate calibration, and minimizes the effects of local noise during the calibration cycle.

The number of calibration samples taken during a calibration cycle is 3 samples by default, but may be set within the range of 1 to 10 samples by the Parameter23 command (See Section III ñ Parameter Commands).

Using DeTransfer, the number of calibration samples is set by the command for example

which sets the number of calibration samples to 2.

The current setting for the number of calibration samples is read in DeTransfer by the command

Using DeLogger, number of calibration samples is set and read using these same commands either

casually in the Entry Screen of the Text View

as part of a program, in the Optional Initialization Commands under the Special Commands button of the Settings tab of the Program Builder

There is little benefit in increasing the number of calibration samples beyond the default. However decreasing the number of calibration samples will reduce the calibration time, with only a minor reduction in accuracy.

The time taken for a calibration cycle when Parameter23=3 (default) is 600 mS, which can cause unpredictable discontinuity's in data collection during rapid scanning.

Autocalibration of the ADC can be enabled and disabled by the /K or /k Calibration Switch command.

Using DeTransfer, autocalibration is enabled by the command

and is disabled by the command

Using DeLogger, autocalibration can be enabled or disabled using the same commands either

directly, in the Entry Screen of the Text View

as part of a program, in the Optional Initialization Commands under the Special Commands button of the Settings tab of the Program Builder

When autocalibration is enabled by the /K command, a new calibration is immediately performed. Therefore ADC recalibration can be forced at any time by the /K command.

Autocalibration by Drift in the Zero Voltage Reference

The basis for determining the need for automatic recalibration of the ADC circuits is the drift in the internal zero voltage reference. The zero voltage drift permitted before a recalibration is performed can be defined by the user.

Drift in the internal zero voltage may be in either direction, and is due to warming up of the logger when first powered up, ambient temperature changes, and aging of circuit components.

The internal zero voltage reference is read before each scan of the analog input channels. If the zero voltage has drifted from zero by more than a defined value, then a calibration cycle is initiated.

The default zero voltage drift for autocalibration is ±4 µV, and can be set within the range of 0 to ±10,000 µV by the Parameter0 command (See Section III ñ Parameter Commands).

Using DeTransfer, the command for example

sets the allowed zero reference voltage drift before autocalibration to ±100 µV. The zero voltage drift is set in increments of 1 µV.

Setting P0=0 will ensure an autocalibration before every scan, while setting P0=10000 (±10 mV) will in practice disable autocalibration.

The current setting for the autocalibration zero voltage drift can be returned into DeTransfer by the command

Internal Calibration and Checks

There are five calibration standards within the dataTaker which determine accuracy of the logger. These standards are used during autocalibration, and are as follows

a 2MHz quartz crystal for frequency measurement

two precision current sources of 250.0 µA and 2.500 mA for resistance measurement

a 2.500 Volt 0.05% precision reference voltage for all voltage and current measurements

a precision 100 Ohm 0.1% reference resistance

The internal reference voltage can be trimmed should it drift with age (see below).

Internal Temperature 

The dataTaker and the Channel Expansion Modules (CEM-AD) have an internal temperature channel, which is accessible to the user. This temperature channel is used as the default cold junction temperature reference for thermocouple support.

Using DeTransfer, the internal temperature channel of the dataTaker can be read at any time by the command

Using DeLogger, the internal temperature channel of the dataTaker can be read

directly, in the Entry Screen of the Text View

by including the Internal : DeTransfer Temperature channel in a program

 

 

The channel type LM35 specifies that the internal temperature sensor is an LM35 which is powered by the analog circuits during reading (See Section II ñ Measuring Temperature with IC Temperature Sensors). The channel number 1% is the internal temperature channel for the dataTaker.

The internal temperature channel of the Channel Expansion Module (CEM-AD) can be read in the same way by the command

Channel number 1% is the internal temperature channel for the Channel Expansion Module (CEM-AD), and n: specifies the CEM-AD module address number of 1 or 2.

Trimming Errors in the Reference Voltage

All measuring systems ultimately compare an input voltage with an internal reference voltage. If the internal reference voltage is not accurate, then the reading of the input voltage will be in error.

The dataTaker assumes that the internal 2.500 Volt reference voltage is accurate.

However should the internal reference voltage drift with age, then the voltage reference can be trimmed by either of two techniques as follows

the 2.500 Volt voltage reference can be trimmed in hardware by adjusting a trim potentiometer. The trim potentiometer is located on the lower circuit board, and is accessible through a screwdriver hole in the upper circuit board beside the memory card connector. The adjustment range is ±0.1%.

the 2.500 Volt voltage reference can also be trimmed by entering a software trim factor

The internal voltage reference is 2.500 Volt ± 0.05%. Therefore the voltage reference can vary by ±1.25 mV from the nominated value.

The 2.500 Volt reference can be trimmed in software as follows

connect an accurately known precision voltage source to any of the analog input channels of the dataTaker, and read the voltage. This precision voltage source should preferably be greater than 1.5 Volt.

using the data for the known voltage and the measured voltage for the precision source, calculate the trim factor by

            DV = ( V_source - V_measured )

            Trim = (DV x ( V_ref / V_source ))

In the following example, the internal reference voltage trim factor is calculated from measurements using a 1 Volt precision source

            V_ref = 2500.00 mV

            V_source = 1000.00 mV

            V_measured = 1000.02 mV

            DV         = (V_source - V_measured )

                        = ( 1000.00 - 1000.02 )

                        = -0.02 mV

            Trim       = ( DV x (V_ref / V_source ))

                        = ( -0.02 x ( 2500.00 / 1000.00 ))

                        = -0.05 mV

                        = -50 µV

The magnitude of the reference voltage trim factor is declared to the dataTaker in terms of 10's of microvolts. The sign of the calculated trim factor must also be entered.

The reference voltage trim factor is entered into the dataTaker using the Parameter1 command (See Section III ñ Parameter Commands).

Using DeTransfer, the command for example

loads a trim factor of -50 µV into the dataTaker.

The current setting for the reference voltage trim factor can be read by DeTransfer by the command

Using DeLogger, the trim factor can be set and read using the same commands either

directly, in the Entry Screen of the Text View

as part of a program, in the Optional Initialization Commands under the Special Commands button of the Settings tab of the Program Builder

The default setting for P1 is 0 µV.

If a precision voltage source is not available, then the actual internal 2.500 Volt reference of the dataTaker can be measured with a precision multimeter, and a software trim factor determined as follows

if the measured voltage is greater than 2.500 Volt, then the trim factor is +1 for every 10 µV of error

if the measured voltage is less than 2.500 Volt, then the trim factor is -1 for every 10 µV of error

Calculate the trim factor required, and define this to dataTaker using the Parameter1 command.