When you move a DA system from the controlled environment of the laboratory to validation testing or the manufacturing floor, problems may arise. Many factors can combine to destroy measurement accuracy or possibly damage the equipment. Understanding these factors is crucial to measurement success.
This two-part article addresses five recurring roadblocks to successful measurements, from most common to least common, together with some real-life examples.
Crosstalk, a phenomenon where the contents of one DA channel are superimposed on another, can cause measurement errors that range from subtle to major and these errors may go undetected. In its most exaggerated form, a nearly exact duplicate of one channel appears on an adjacent channel to which nothing is connected.
The revolution in PC-based instrumentation includes the active use of multiplexers and is driven by the promise of low cost per channel, with a target per-channel cost of $30–$40. Along the way, however, the hallmark of traditional instrumentation has been dropped—having an isolation amplifier for each channel. The system under test is connected directly to the multiplexer's inputs but the multiplexer's inputs have a capacitance that stores a charge that is directly proportional in magnitude to the sample rate and the impedance of the signal source. This inherent characteristic causes crosstalk.
Consider an application where the multiplexer's input is connected directly to the output of an isolation amplifier. In this situation, the impedance the multiplexer sees is stable and low; 10 Ω is a typical value. Crosstalk is greatly minimized or eliminated altogether because the impedance of the source is low enough to bleed off the charge on the multiplexer's input capacitance before the ADC reports a value.
Even under this nearly ideal impedance situation, a high sample rate can boost crosstalk by minimizing the capacitive discharge time on the multiplexer's channels. In effect, the capacitance has less time to bleed off its charge before the A/D conversion takes place, resulting in crosstalk where none existed before.
As source impedance and sample rate increase, the probability of crosstalk increases as well. To prevent this from happening, keep these points in mind:
- Minimize the source impedance of the signal source. Use isolation amplifiers to keep it below 100 Ω; at very high sample rates, even this value may be too high.
- When source impedance cannot be controlled, an isolation amplifier is needed between the signal source and the DA card multiplexer. An instrument with a built-in isolation amplifier on each channel is needed to provide protection from stray signal paths.
- These strategies become more important as the sample rate increases. The best and most predictable results are obtained when using an instrument with a fixed scan interval; this helps control any high-sample-rate bursts.
Although crosstalk tops the list of most frequent causes of measurement problems, common-mode voltage (CMV) tops the list in its ability to distort data.
When using familiar measurement methods, such as a battery-powered, handheld digital voltmeter (DVM), input readings are almost impervious to CMV problems. Many engineers assume that they can extend the success of DVMs to PC-based measurement approaches; in most cases the results range from poor to disastrous.
The problems encountered are tied to two specifications on the manufacturer's data sheet: full-scale input range and maximum input voltage (without damage). Full-scale input range indicates the magnitude of voltage connected across the instrument's inputs (the normal-mode voltage or NMV), which can be successfully measured. As its name implies, maximum input voltage indicates how much NMV the instrument will tolerate before incurring damage.
The CMV, which appears simultaneously and in phase on each of the instrument's inputs with respect to power ground, combines with the NMV to test the limits of the measurement system. Most DA products for the PC permit measurements when the sum of CMV and NMV is equal to or less than the instrument's full-scale input range. Measurements can be made under these conditions only if the DA product's input is configured for differential operation as shown in Figure 1. Using this basic rule for differential measurements, here are the possible measurement results, listed from best to worst:
- (CMV + NMV) ≤ Full Scale Range: This is a good measurement, subject to the common-mode rejection (CMR) specification of the instrument.
- Full Scale Range ≤ (CMV + NMV) ≤ Maximum Input Voltage (without damage): The measurement could be latched at plus or minus full scale. Although there are no usable measurement results, there is also no damage.
- (CMV + NMV) > Maximum Input Voltage (without damage): There is potential for damage to the DA product and to the attached computer.
Figure 1. DA differential operation
For most DA products, it doesn't take long to reach the destructive stage. Most will tolerate a maximum input voltage (without damage) of ±30 VDC or peak AC. In the realm of production measurements, with 120 V to 440 VAC motor supplies or 24 VDC process current supplies and the high probability of ground loops, this limit can be very quickly and irrevocably exceeded. How can the comparatively expensive DA instrument be used to collect the same measurements that are made so effortlessly and safely with the handheld DVM? The answer is to choose a product that provides isolation.
Isolation is what its name implies. As with a battery-powered DVM, there is no electrical connection between the common connection associated with the instrument's front-end input terminals and the power common connection associated with the back-end of the instrument and the computer.
As such, the instrument's front end is free to float at a level defined by the magnitude of the CMV, without damage and with complete measurement accuracy. Here, the maximum CMV that can be tolerated is not dictated by its maximum input voltage specification, or even by its full-scale range, but rather by the voltage at which the isolation barrier breaks down. For example, with most Dataforth isolated signal conditioning products, isolation barrier breakdown occurs above 1500 VAC or 2200 VDC, which is much higher than the expected CMV of most production applications.
Figure 2 describes a typical application where isolation allows a measurement in the presence of a high CMV. Isolation can be provided in more than one form: input-to-output, channel-to-channel, and a combination of both. For the vast majority of multichannel production applications, both input-to-output and channel-to-channel isolation are needed. Such an arrangement allows each channel's input to float with respect to all other channels.
Figure 2. Measurements with high CMV present
A CMV on channel one, for example, will not disrupt measurements on the other channels, even if they are referenced to power supply ground, the same CMV, or an entirely different CMV. In contrast, systems designed with just input-to-output isolation essentially tie together the CMVs of all the channels. A CMV on one channel floats all channels at the same voltage with potentially disastrous results if another channel is connected to a ground-referenced signal or a different CMV.
There is only one reason to buy a product that offers only input-to-output isolation, and that reason is cost. For example, it is less expensive to build a single isolation barrier into an 8-channel DA instrument than to build one for each channel. The cost savings usually are reflected in a lower system price. However, in the actual application of such an instrument, expensive repairs may offset the initial cost savings later.
Before leaving the topic of isolation, one point needs to be established firmly and clearly: Do not confuse a product that offers differential measurement capability with one that offers isolation. These are two entirely different and unrelated features.
Some engineers are still under the impression that a product with differential measurement capability allows them to apply the instrument in high CMV conditions. As discussed earlier, differential but nonisolated inputs tolerate only moderate CMVs without damage. For more trustworthy measurement results lower CMVs are required.
In part two of this article we will discuss additional sources of measurement error—DC common-mode rejection, AC common-mode rejection, and measurement range and input protection—that can be avoided by using DA instruments with appropriate isolation.
Dataforth acknowledges and credits Roger Lockhart, DATAQ Instruments, Inc., for the technical content of this Application Note.