The properties required for signal transmission are determined by different factors. In addition to the requirements regarding accuracy and signal transmission speed, the input data of downstream devices, the properties of the signal to be transmitted, and if applicable the ambient conditions must also be observed.
3.1 Current or Voltage Transmission
The initial criteria for selecting an isolation amplifier or a transmitter are the input signal to be processed and the output signal required. The output signal is generally determined by the downstream devices such as controllers, indicators, PLC, PCS, etc., whereby many of these devices have either current or voltage inputs.
Current signals should be preferred if both options are available, particularly for longer transmission paths (see Fig. 14). Impressed current signals are considerably less sensitive to interference than voltage signals.
Figure 14: Transmission of a measurement signal over large distances
3.2 Input Resistance
The input resistances of modern isolation amplifiers are generally dimensioned in a way that they are sufficiently high for voltage inputs and sufficiently low for current inputs so that the signal being processed is practically not loaded. Only in a few cases (very low voltage signals with a high source resistance or low-load capability current signals) would the input resistance be a selection criterion for isolation amplifiers. The input resistance of the VariTrans P 41000 isolation amplifiers – which was developed specifically for shunt applications – is at approx. 100 kΩ relatively low in comparison to other isolation amplifiers. For shunt applications with resistances in the mΩ range, however, the resistance is still several orders of magnitude higher than required.
3.2.1 Input Voltage Drop
In various isolation amplifiers with a current input and in loop-powered isolators, the load on the input signal is specified as a voltage drop and not as an input resistance. This voltage drop is constant under normal operating conditions and – depending on the model – is max. 500 mV for active isolation amplifiers. In passive isolators, there is a voltage drop at the input resulting from the voltage requirement of the device plus the load voltage at the output. Therefore the load capability of the measurement signal and the load connected at the isolator output should both be known before using passive isolators. An exception are the passive isolators from Knick with load stop function: The current supplied at the primary side is maintained independent of the output load, without generating feedback.
3.3 Output Load Capability
The load capability of voltage outputs is generally indicated by the maximum current. Almost all manufacturers specify a resistance value for the load capability at current outputs. This specification does not indicate the load capability of the output currents of Knick isolation amplifiers absolutely correctly. Therefore the output load capability is “traditionally“ given as a voltage value.
The specification of a maximum permissible load voltage of 10 V therefore applies to every current value, whereas 500 Ω would apply exclusively to 20 mA.
3.4 Transmission Accuracy
Many Knick isolation amplifiers are distinguished by extraordinarily low transmission errors so that the accuracy requirements of virtually all industrial measurement tasks are exceeded. The long-term stability of Knick isolators ensures maximum transmission accuracy even past the five-year warranty granted for Knick isolation amplifiers and transmitters.
3.4.1 Quality of Measurement Signals
The highest possible accuracy of input signal transmission is not only required for measurement applications. Signal distortion due to polarity switching, overshooting during signal fluctuations and extreme ramp angles in rectangular wave transmission are the rule for many isolation amplifiers available on the market. These undesirable properties are initially not noticeable to the user. Unexplainable measurement errors are often only detected when the device is put into regular operation. In the cyclical, digital scanning of measured values, signal distortions, for example, due to overshooting, can cause serious measurement errors. Therefore signal transmission accuracy has traditionally played a large role in the development of Knick isolation amplifiers.
3.4.2 Residual Ripple
The output signal of isolation amplifiers and transmitters is principally superimposed by low interference voltages. These interference voltages are caused, for example, by the chopper frequency as well as by mains feedover. The amplitude of this interference voltage, referred to as residual ripple, should be as low as possible because otherwise measurement errors cannot be ruled out – especially with low modulation.
3.5 Temperature Coefficient (Gain Droop)
The temperature coefficient or gain droop is a specification for changes in gain caused by temperature changes. Droop rates are specified as a relative parameter in %/K or as an absolute value in nA/K or µA/K, for example. In absolute value specifications, you need to check whether the TC refers to the input or the output.
– The temperature coefficient at the output of an isolation amplifier is max. 10 nA/K.
A change in temperature of 20 K causes a change in the output current of 20 · 10 nA = 200 nA.
– The TC of an isolation amplifier is 0.0025 %/K.
A change in temperature of 20 K causes a change in gain
of 20 · 0.0025% = 0.05%.
3.6 Offset Voltage, Offset Current
In real amplifiers, the output variable is not exactly zero even when the input signal is zero. The input offset voltage of an amplifier is defined as the voltage that must be applied to the input to bring the output to zero. It therefore acts as an input voltage or an additional voltage acting in series with the input signal (see Fig. 15).
Figure 15: Offset voltage
The input offset current of an amplifier also acts as an additional input signal (see Fig. 16). In amplifiers with a voltage input, the offset current causes a voltage drop at the internal resistor of a signal voltage source, which is added to the input signal. The offset voltage and offset current are so low in Knick isolation amplifiers that they are negligible for normal applications. Offset influences should only be considered for very special applications such as the 1:1 transmission of very small measurement signals or the transmission or amplification of very high-resistance signals. The polarity of offset parameters depends on the device and therefore is given as an absolute value without plus or minus sign.
Figure 16: Offset current
3.7 Cutoff Frequency
Isolators and transmitters are principally designed for the transmission or amplification of direct voltage signals. In order to be able to transmit fast changes in the measured value almost without delay, Knick devices can also transmit alternating quantities to a certain extent. The upper cutoff frequency is up to 12 kHz for sinusoidal signals – depending on the model. As is common in electronics and telecommunications, the upper cutoff frequency is defined as the frequency at which the gain is attenuated by 3 dB compared to the DC gain, i.e. it corresponds to approx. 71 % DC gain (see Fig. 17).
Figure 17: Cutoff frequency
3.8 Common-Mode Behavior
If the same voltage Vcm is applied to ground at both inputs of a symmetrical amplifier, the input voltage remains at Vin = 0. This operating mode is called common-mode modulation. In an ideally symmetrical amplifier, the output voltage Vout would also remain at 0. This is not the case in real amplifiers, however, i.e. a voltage deviating from 0 will appear at the output (see Fig. 18). Common-mode modulation always exists when the signal voltage is not at ground potential, i.e. when there is a potential difference between the (two) input lines and ground, for example, when measuring voltages across a shunt lying at a high potential against ground.
Figure 18: Common-mode modulation
Common-mode voltages can also occur as common-mode interference, for example, during switching processes, due to stray pickup in the signal lines or due to compensating currents. The ratio between an applied common-mode voltage and the resulting output voltage is called the common-mode gain. However, in practice, the deviation from the ideal common-mode behavior of an amplifier, specified as common-mode rejection, is of greater interest. The common-mode rejection ratio CMRR is defined as the ratio of differential-mode to common-mode gain or as the logarithmic ratio of an applied common-mode voltage Vcm to a signal voltage Vd that would produce the same output signal: CMRR = 20 · log (Vcm/Vd) [dB].
Transient common-mode voltages can be caused both by single or periodic switching processes such as in thyristor-controlled converters. For the VariTrans P 40000 series isolation amplifiers, the TransShield technology was implemented to suppress such common-mode pulses. Compared to conventional designs, it enables very compact high-voltage transformers with low leakage. Thanks to the resulting space advantage, the VariTrans P 41000 shunt isolators can be installed in an only 22.5 mm wide modular housing. Common-mode interferences such as high transients are reliably isolated and cause hardly any measurement errors at the output. The term T-CMRR (Transient Common Mode Rejection Ratio) has been chosen for the corresponding data specification.
It describes the ratio of differential DC gain to common-mode gain of a transient interference signal with a slew rate of 1000 V/µs (see Fig. 19).
Figure 19: Test circuit for measuring the T-CMRR
The VariTrans P 40000 series devices are therefore particularly well suited for measurements on shunts where common-mode pulse voltages or rapidly changing common-mode voltages are to be expected. The isolation amplifiers achieve a T-CMRR of 115 dB. The common-mode rejection for 50 Hz interferences is 150 dB.