The handheld digital multimeter (DMM) is the most common of all electrical and electronic test instruments. Despite the pervasiveness of the the DMM, its specifications and features are often not fully understood. This may lead to overlooking its more advanced measurement solution capabilites or, worse yet, not understanding its limitations. This article will help DMM users to better understand the significance of the important performance specifications.
Traditionally, DMM displays have been specified as "3 1/2-digit" for example. The meaning of this is that there are three complete digits, each capable of displaying the numbers zero to nine, and one additional preceding digit which may display only a zero (may be blanked in this case) or a one for a full scale reading of 1999. Not exactly intuitive but it has been around long enough for most users to understand it. Newer DMMs have clouded the picture somewhat by increasing the full scale range to 3999 or 39999 or more. These have been dubbed 3 3/4- and 4 3/4-digits respectively. This description is even less intuitive than it was for 3 1/2-digits. A better approach which is now displacing the fractional digits is to specify the number of "counts" that may be displayed. For example, the 3 1/2-digit display is described as 2000 count (1999 plus the reading of 0). From that description, it becomes readily obvious what the display is capable of showing. 3 3/4-digits becomes 4000 counts and 4 3/4-digits becomes 40000 counts. Some confusion arises in cases where 3 3/4 digits has been used to denote 3000 or 5000 counts. Table 1 shows the relationship between digits and counts for the more common DMM displays.
Table 1: Digits vs Counts
The number of counts usually applies to the DC volts function. Fewer counts may be displayed on the same instrument for certain functions. For instance, a 40000 count DMM may be limited to 4000 counts when measuring capacitance.
Resolution is a measure of the smallest increment that may be discerned. At first glance, it would seem that 10.000 volts measured with a 40000 count DMM would be read to a resolution of 0.001 volt. This is usually the case where the DMM's A to D converter resolution exceeds that of the display but some have less resolution than the display. In this case, the last digit could read 0, 1, 2, 3, 5 ,6, 7, 9, etc. with a linearly increasing voltage. Notice that only eight out of the ten possible values were displayed. This is an artifact due to the digital nature of the conversion. In a more extreme case, only odd or even digits are be displayed, hence the need for a resolution spec separate from the display count.
The increased resolution of 20000 and 40000 count DMMs does not come without penalties. Longer settling times are required for the far right digits to reach their final value. To partially offset this, analog-like bargraphs are included. Because of their lower resolution, they provide near-instantaneous response to changing input signals. They permit peaking and nulling adjustment where rapid indication of the adjustment effects is needed.
The resolution specification for certain functions may be limited intentionally because of noise or accuracy limitations related to that particular function. The meaningless trailing display digits are usually blanked in such a situation. Capacitance is a good example of a function included in many high end DMMs that has accuracy limitations where it is desirable to limit the display to only those digits which are meaningful. Displaying more than that has two fundamental problems. The first is that the meaningless digits usually are rapidly changing and would only serve to distract the user from those digits that are significant. Secondly, the display of random numbers leads the user to think that the instrument is inaccurate. Even if the reading appears stable, it may not be accurate as with a 40000 count DMM having 0.5% accuracy in AC volts. Additional resolution without a corresponding accuracy improvement is not meaningful to anyone except the advertising department!
Accuracy, Uncertainty, and Repeatability
Accuracy is often the differentiating specification between similar appearing models. DC accuracy is usually used for the "banner" specification since it is usually better than the accuracy for other functions. DC accuracies of better than ±0.1 percent are just now becoming available. Added to the percentage figure will be a specified number of counts (sometimes referred to as "digits") due to rounding error and noise limitations. For DMMs, accuracy is specified as percent of reading as opposed to percent of full scale as specified for analog multimeters. If possible, avoid use of the bottom ten percent or so of any range since accuracy is badly degraded there. It isn't the percentage error that is the problem; it is the effect of the number of counts deviation becomes substantially larger in proportion to the measured value.
AC accuracy is usually less than that for DC. It is also optimized for 50-60 Hz. Other frequencies may have poorer accuracy. As with DC accuracy specifications, a number of counts (often greater than for DC) will be added to the accuracy percentage. Also, for waveforms other than a pure sine wave, additional inaccuracy will be encountered when measured with an average responding DMM. Even a true RMS responding DMM will have some accuracy limitations for waveforms with high peak amplitude components if measured near full scale. This will be discussed further under TRMS/AVG.
From a metrologist's standpoint, what appears in the manufacturer's specification sheet under accuracy is more properly deemed uncertainty with accuracy being reserved to indicate the probability of the reading being accurate. A specification of 1% uncertainty would have an accuracy of 99%. Common practice among instrument manufacturers is to use the term accuracy and for calibration labs to use uncertainty with both representing the same thing in real life.
Repeatability is often more important than absolute accuracy when making a series of measurements. You want the DMM to read the same value each time a measurement is made of exactly the same value. Repeatability is not usually specified directly but a limit is implied by the accuracy specification since it must always be within the specified accuracy. It may or may not be considerably better than the specified accuracy and can vary considerably from one DMM to another, even of the same manufacturer and model. Component aging, battery condition, temperature, and warm-up time may all effect repeatability to some degree. The only way to get a feel for a particular DMM's repeatability is to make a large number of measurements of a precision source under a variety of conditions. The source should have better than four times the rated accuracy of the DMM to be able to sort out repeatability and accuracy limitations.
In the past, calibration of DMMs has usually been accomplished by adjusting several internal "tweaks". Initial calibration by the manufacturer included those along with filing or applying solder to the high current range shunt resistor (actually a piece of buswire) to adjust its value. This was the stone age version of the laser trimming process used for film resistors. These manual adjustments are gradually being displaced by software calibrations stored in the microprocessor. These software tweaks are accessible through a particular keystroke sequence while a specified input signal is applied to the DMM input. This technique reduces time and cost for calibration and allows better accuracy than attainable with a potentiometer or resistor adjustment since both solder and Cermet potentiometers have undesirable temperature coefficients.
True RMS or TRMS reading DMMs populate the high end of available handheld instruments. They are assuming greater importance today in light of the heightened concern over power line harmonic distortion caused by widespread use of switching power supplies in office computers and todays proliferation of electronic devices. TRMS is an advantage when measurements of AC waveforms other than sinousoidal are required.
RMS measurements are a measure of the equivalent heating effect produced by a voltage and, to be accurate, must include any DC component present along with the AC component that most users associate with RMS readings. Many TRMS DMMs are not capable of measuring the TRMS value of the combined AC+DC and errors may result if there is any DC superimposed on the AC voltage.
Conventional non-TRMS DMMs measure the average of the absolute value of AC voltage and are calibrated so that the readings are corrected to that of the RMS value for a sine wave. This works well but errors occur if harmonics are present with the effect becoming progressively worse as the harmonic content increases. Table 2 shows the readings that would be obtained when average and TRMS responding DMMs are used to measure sine, triangular, square, and pulse waveforms and DC.
Table 2: Average VS. TRMS Comparison of Typical Waveforms
|Waveform||Factor1||Factor2||Pk-Pk||Peak||RMS||AC+DC||AC Only||Rdg||For Avg.|
*Positive going pulse between 0 and 1Volt
1Ratio of peak to RMS
2Ratio of RMS to average
First, it can be seen from the data that average responding DMMs can have a substantial error when measuring squarewaves with their rich harmonic content. TRMS is not always the best, especially if millivolt amplitude sine waves are to be measured. Average responding instruments are faster settling to the final value or for adjusting levels near zero volts than equivalent TRMS DMMs and may have less offset or zero errors. This is common to TRMS meters since most use the same type TRMS converter integrated circuit.
Second, a "TRMS" DMM that does not read AC+DC will exhibit a -42.2% error under the AC with DC offset condition shown in the table. In this case, if TRMS readings are required, it will be necessary to compute it from separate DC and TRMS AC measurements using the equation:
VTRMS = (VDC2 + VAC2)-2
In both of the previously described instances, understanding what the DMM is actually responding to will help to prevent misinterpretation of measurements. A few DMMs offer the ability to select average, AC TRMS, or AC+DC TRMS thus allowing the user to choose the optimum mode for the desired measurement.
Crest factor also should be considered when making AC voltage measurements of non-sinusoidal waveforms. Crest factor is defined as the ratio of the peak or "crest" voltage compared to the RMS voltage. The crest factor relates to ideal waveforms as shown in Table 2
True RMS reading DMMs will usually specify the maximum crest factor that they can handle accurately. This will usually be lowest near the full scale reading due to the saturation characteristics of the TRMS converter chip. Better accuracy may be obtained near mid-scale.
Form Factor is defined as the ratio of the RMS value to the average value and is the factor that when multiplied by the average value of a waveform will equal the RMS value.
The environmental concerns for DMMs are primarily temperature, humidity, altitude, shock, and vibration. The operating temperature range is usually limited by the liquid crystal display. At low temperatures they become sluggish and hard to read. At high temperatures contrast may degrade due to the change in optimum LCD operating voltage with temperature, especially when multiplexed. Exceeding the non-operating temperature range may result in permanent deformation of the LCD polarizer, particularly when combined with high humidity. After changing the temperature of a DMM significantly, 30 minutes or more should be allowed for it to stabilize to the new ambient temperature before making important readings. Also, examination of the accuracy specifications for DMMs will reveal an important fact -- the specified accuracy is valid only for a relatively narrow range. 23 ±5° C and less than 70 or 80% relative humidity is common. The instrument is certainly usable over a much wider range but its accuracy will be degraded to some degree. Most DMMs are relatively well protected against humidity; however, they are not hermetically sealed and excessive humidity may cause internal condensation and possible loss of accuracy. Altitude is not usually a problem, at least at normal measurement voltages since arcing is unlikely at these levels.
Shock and vibration are important with the abuse that DMMs are subjected to and not just in the field. Their small size and long test leads result in a number of drops from benchtop height. Consequently they should be designed to survive a 3 to 4 foot drop onto a hard surface. This specification is much easier to understand and appreciate than the usual range of amplitudes and frequencies prescribed in MIL specs and related environmental test procedures.
Most if not all of today's DMMs are remarkably durable and hold up admirably to everyday abuse that you wouldn't consider subjecting a laboratory meter to. And the accuracy and functionality of handheld DMMs today often exceed that of laboratory instrument of just a few years ago.
Historically, each country has developed its own safety requirements for electrical devices sold within their borders. They were not necessarily legal requirements but were increasingly dictated in recent years. The proliferation of differing national standard writing bodies (UL, CSA, EN, DIN, etc.) and international distribution of products led to the need for "harmonization" of the standards. This need is now being led by the IEC development process.
For DMMs, the standards for electrical measurement instruments are the ones of interest here. Many other safety standards for medical, information technology, and home entertainment equipment exist. Much changing of standards has occurred during the 1990s. In the United States, UL 1244 is being replaced by ANSI/ISA S82.01-1994 and UL 3111-1, and in Canada, CSA C22.2, No. 231 is being replaced by CSA C22.2 No. 1010.1-92. All of these recent standards are harmonized with IEC 1010-1 which replaced IEC 348. IEC 1010-1 in turn is being adopted in the European Union as EN 61010-1. Note the similarity of the numbers for these standards. All contain 1010-1 in their designation with the exception of UL 3111-1. In this case, the 1010 designation had been used previously by UL for another standard and was unavailable. In any event, new DMMs must conform to an IEC 1010-1 harmonized standard. New instruments are certified to these while those that have been on the market for some time may or may not be certified to IEC 348, UL 1244, CSA 231, etc. which were in effect at the time of introduction but which were not mandatory.
In the United States, the marking "NRTL" indicates a Nationally Recognized Testing Laboratory. These laboratories perform safety testing equivalent to Underwriters Laboratories.
Additionally, DMM inputs must now be marked with an overvoltage Category I, II, or III for expected transients. These refer to the following usage:
CAT I Signal level, special equipment or parts of
equipment, telecommunication, electronics
CAT II Local level mains, appliances, portable equipment
CAT III Distribution level mains, fixed installation
CAT IV Service drop to building (outside)
Similar overvoltage transients are expected for a 600 volt CAT III or 1000 volt CAT II mains line but a meter input designed for 600 volt CAT III may be internally limited and not suitable for 1000 volt CAT I or II use. It is therefore important to observe any voltage and current limitations indicated on the DMM or in its manual. Damage to the DMM or operator injury may occur if they are exceeded. The current ranges are usually protected against overcurrent by fuses. Fuses of the identical type must be used for replacement. They are rated for the voltage indicated on the DMM or in its manual. Exceeding this voltage or replacement with a lower voltage fuse can result in a hazardous sustained arc if the fuse opens.
Certification of products is not necessarily performed by the standards formulating organizations themselves. Other independent testing laboratories such as ETL are internationally recognized and perform their testing to UL, CSA, or IEC standards.
Instruments sold internationally now must also meet strict Electromagnetic Compatability (EMC) requirements, both for generation of emissions and for immunity to nearby electromagnetic and electrostatic fields. The EMC standards used for DMMs have evolved from a hodge-podge of local standards to international CISPR and IEC standards and, finally, to the harmonized EN standards. In order to sell into the European Union or EU, (formerly known as the European Community or EC) under directive 89/336/EEC, DMMs must now meet EN 55011 for radiated emissions, EN 50082-1 for susceptability, and display the CE conformity mark (Figure X). EN 55011 was adopted from the earlier CISPR 11 and Class A is for instruments used in a commercial environment as opposed to Class B which is for home use. EN 50082-1 is the generic electromagnetic immunity standard. It refers to a series of IEC standards for the individual test procedures. Those pertaining to DMMs are IEC 801-2-1984 for ESD immunity testing and IEC 801-3-1984 for RF immunity. Of these, the ESD immunity is probably of the most interest to the average DMM user since it assures that static zaps of this frequently handled instrument won't kill it. The normal testing is for protection to 8 kV. Testing to 15 kV is better yet since this is a realistic value that may be encountered on a dry day at an indoor location. It is expected that a new EN standard will replace EN 50082-1 in early 1997. It will refer to IEC 1000-4-2 and IEC 1000-4-3 that will replace IEC 801-2-1984 and IEC 801-3-1984 respectively.
The versatility of the DMM and the design compromises required to allow additional features or optimization of certain specifications for advertising purposes require a thorough understanding of your DMM to achieve maximum accuracy and to apply it to problem solving in new applications. The competitive trend among DMM manufacturers toward greater accuracy, wider measurement range, and special features such as waveform viewing combined with compactness and reasonable cost are certain to continue. You, the user, will be the ultimate winner with new problem solving capability. But, take a little time to understand the limitations as well as the features in order to realize the maximum return for your DMM investment, as small as this investment may be in today's competitive environment.