As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
A low pass filter on a clamp meter, usually labelled as a drive mode, attenuates everything above a corner frequency so that the slow fundamental the motor actually turns on is what reaches the converter. Use it whenever you are reading voltage, current or frequency on the output side of a variable speed drive and you want a number that corresponds to what the drive thinks it is producing. Do not use it when the high frequency content is the thing you are investigating, which covers bearing current complaints, screen and gland problems, harmonic disputes with a supplier, and any check on filter or reactor hardware. The filter does not know the difference between noise and evidence.
What a drive mode filter is actually throwing away
Inside the instrument the filter is a simple arrangement of resistance and capacitance between the sensing stage and the converter. Its behaviour is described by a single number, the corner frequency, at which the output has fallen to about 0.707 of the input. Below that corner almost everything passes. Above it the response falls away steadily, roughly by a factor of ten for every tenfold increase in frequency when the filter has one pole.
That gentle slope is the part people misread. A drive mode is not a gate that admits the fundamental and blocks the carrier. It is a slope, and a substantial fraction of the carrier can still get through if the carrier is not far enough above the corner. The table below gives the surviving fraction for a single pole filter at various multiples of its corner, which is the arithmetic that decides whether a particular instrument will clean up a particular drive.
| Frequency as a multiple of the corner | Fraction of amplitude that survives | What that means in practice |
|---|---|---|
| One tenth | 0.995 | A motor fundamental passes essentially untouched |
| One half | 0.894 | Low order harmonics largely survive |
| Equal to the corner | 0.707 | The defined corner, about three decibels down |
| Double | 0.447 | More than half removed, but far from gone |
| Five times | 0.196 | A carrier well clear of the corner is strongly suppressed |
| Ten times | 0.0995 | Attenuated by a factor of ten |
Carrier frequency against fundamental, and which one you need
Two quite different frequencies live on the same pair of output conductors. The fundamental is the slow one the drive is synthesising, anywhere from a fraction of a hertz during a soft start up to several hundred hertz on a high speed spindle. The carrier is the rate at which the output switches are actually opening and closing, usually some thousands of times a second, chosen by the designer as a compromise between audible noise, switching losses and how smooth the current turns out.
Almost every question an installer asks concerns the fundamental. Is the drive putting out the voltage it claims. Is the motor drawing balanced current on all three terminals. Is the output frequency consistent with the speed the process needs. None of those are answered by a reading that includes the carrier, because the carrier contributes real energy to the total without contributing anything the motor turns into torque. The guide to choosing a clamp meter covers which measurement functions go together, and the comparison between a multimeter and a clamp meter covers where each instrument belongs.
How little any of these listings say about the corner frequency
| Instrument | Filtering stated | Corner frequency stated | Other published behaviour relevant to drives |
|---|---|---|---|
| FLIR CM276 clamp meter | Drive mode described as a low pass filter for voltage, current and frequency | Not stated in the listing bullets | A 160 by 120 infrared imager, worklights, app connectivity with alarm thresholds |
| POROMETISTO CM06 clamp meter | Drive mode and low impedance mode both stated | Not stated in the listing bullets | Inrush capture over roughly a 100 ms window, stated as applying to alternating current only |
| UNI-T UT219PV clamp meter | Not stated in the listing bullets | Not stated in the listing bullets | True RMS alternating and direct current, maximum and minimum hold, data logging, relative measurement |
| Megger DCM1500S clamp meter | Not stated in the listing bullets | Not stated in the listing bullets | 2000 V direct, true RMS 1500 V alternating, 1500 A alternating and direct, a hold function |
The FLIR CM276 is unusually explicit about the mechanism, stating that its drive mode uses a low pass filter to give accurate voltage, current and frequency readings where switching electronics would otherwise interfere. It still does not say where the corner sits. The POROMETISTO CM06 names both the drive mode and the low impedance mode and adds a detail many listings omit, that its inrush capture runs for about a tenth of a second and applies only to alternating current.
Jobs where switching the filter in will hide the fault
Bearing damage complaints come first. Circulating shaft currents are driven by the common mode content of the switched output, so the quantity you need to see is exactly the quantity the filter removes. The same applies to screen termination problems, where a poorly landed cable screen shows up as high frequency current flowing somewhere it should not, and to earth leakage that only appears while a drive is running.
Harmonic questions are the second category. If you are comparing a site against a supply agreement or trying to work out why a transformer is running warm, you need the content above the fundamental counted rather than discarded. Commissioning checks on output reactors, sine filters and common mode chokes are the third, because the whole purpose of that hardware is to change what sits above the corner, and an instrument with its own filter engaged cannot tell you whether the hardware is doing anything. The Megger DCM1500S is a reminder that a very capable instrument can publish high ranges and true RMS conversion without stating any filtering at all, which makes it an honest unfiltered reference and a poor choice for a quick fundamental reading.
A short sequence for a drive you have never met
Start unfiltered and start on current, because the jaws disturb nothing. Clamp each output conductor in turn and note the three readings. Switch the drive mode in and repeat. A small change on all three tells you the carrier contribution is modest and the installation is behaving. A large change on one conductor only is worth stopping for, because the asymmetry is the finding rather than the magnitude. Then move to voltage, taking the same pair of readings and recording both.
Take the frequency reading last and take it filtered, because an unfiltered frequency count on a switched output is frequently meaningless. Compare the filtered output frequency with what the drive keypad reports, and compare the filtered current with the motor nameplate. If anything disagrees, go back and look at the unfiltered numbers before blaming either instrument. Two reviews on this site walk through the physical handling side of the same work, the Klein clamp meter write up and the Ziboo clamp meter write up, both of which deal with jaw access in a crowded enclosure.
Frequently Asked Questions
Is a drive mode the same thing as true RMS?
No, and the two solve opposite halves of one problem. True RMS describes how the instrument converts whatever reaches the converter into a heating equivalent value. A drive mode describes what is allowed to reach the converter in the first place. An instrument can be true RMS with no filter, in which case it includes the switching energy, or it can have a filter without true RMS, in which case it is applying a sine assumption to a filtered waveform. Serious drive work wants both.
What corner frequency do these filters usually sit at?
Commonly somewhere in the region of one kilohertz, which is comfortably above any motor fundamental and comfortably below most switching carriers. That is a general industry pattern rather than a figure you can assume for a particular instrument. Where the corner sits determines how much carrier survives, so the only reliable source is the manual for the specific model in front of you.
Will the filter change the frequency reading as well as the amplitude?
That is often the main reason to engage it. A frequency counter looking at an unfiltered drive output has to decide which of the many edges crossing its threshold constitute a cycle, and it frequently latches onto the carrier or onto something in between. Remove the high frequency content and the counter sees a recognisable fundamental, which is the number that tells you what speed the motor is being asked to run at.
Does the filter affect a current measurement taken with the jaws?
Yes, in the same way it affects a voltage measurement, because the filtering happens after the sensing element and before the converter. Motor current is already smoothed considerably by the winding inductance, so the difference between filtered and unfiltered current is usually far smaller than the difference between filtered and unfiltered voltage. A large gap on the current side is itself worth noting.
Should I record the filtered or the unfiltered number?
Record both, labelled. The filtered value is what you compare against the drive display and the motor nameplate. The unfiltered value, read alongside it, tells you how much energy is sitting above the corner, and a change in that gap between one visit and the next is a genuine signal about cable condition, earthing or filter components.
Can a filter rescue a measurement that is otherwise out of specification?
No. Filtering changes which part of the spectrum is counted, it does not widen a crest factor limit, extend a safety category or improve the basic accuracy of the converter. If the instrument was never rated for the place you are standing or the peaks you are feeding it, a drive mode does nothing about either problem.