Fluke 87V vs Fluke 117: The Magnitude That Matters on a Noisy Generator Feed

The myth: “Any True-RMS meter handles a generator feed the same — it’s just AC voltage.” A maintenance tech once chased a phantom 20 V discrepancy on a 480 V backup genny for three shifts. The meter read 497 V on one phase; a second meter read 480 V. Neither was “broken.” The magnitude of the error — and the proportion of harmonic content that drives it — is the only thing that separates a useful reading from a wild goose chase. Here’s where Fluke multimeter’s own lineup diverges in a way you can measure.

1. Harmonic proportion: why 2 % THD can inflate reading by 8 V on a 240 V genny

Take a portable generator with 6 % total harmonic distortion (typical for a brushed synchronous set under 75 % load). The True‑RMS of a 240 V fundamental plus 6 % third‑harmonic (14.4 V) is not 240 V; it’s √(240² + 14.4²) ≈ 240.4 V — negligible. But many generators produce both third and fifth, plus notch distortion from the AVR: the actual crest factor can shift to 1.55 instead of 1.414. A meter that measures RMS over a wide bandwidth (87V, typical bandwidth >100 kHz) will include that extra harmonic energy. The Fluke 117, with a more modest bandwidth (~20 kHz), naturally rolls off high‑order harmonics; its RMS reading may be lower by 0.5 % to 2 % depending on the generator’s waveform [based on typical true‑RMS chipset behaviour; illustrative].

Worked consequence: If you are setting the generator’s AVR to produce 240 V using a 117, you might actually be setting the fundamental to 244 V because the meter disregards some harmonic content. The load (a VFD or a switching power supply) sees the higher fundamental and may over‑voltage trip. The 87V’s wider passband gives you a reading that matches what a typical non‑linear load sees. Reversal: If you only care about the RMS voltage that a resistive load (heaters, incandescent lamps) experiences, the lower bandwidth of the 117 produces a reading closer to the heating value — for that specific load, the 117 is more correct.

2. Low‑pass filter: isolating the fundamental when you need it

On a generator that feeds a VFD (e.g., a pump or chiller), the meter’s reading can jump 5 % when the VFD is running because of carrier‑frequency ripple (4–16 kHz). The 87V includes a low‑pass filter with a corner at ~1 kHz that attenuates that ripple. Without the filter, the total RMS includes the carrier; the reading can be 8–12 V higher on a 480 V system (illustrative). The 117 has no LPF, so every reading on a VFD‑fed generator bus will include that carrier — you get a number that is technically correct (it’s all volts) but useless for setting the generator’s voltage regulator.

Magnitude proportion: The error is proportional to (V_carrier² / V_fundamental²). At 6 kHz carrier with 10 V ripple on 480 V, the reading error is about 0.02 % — trivial. But if the generator also has a notch‑type AVR that creates high‑frequency ringing, the error can exceed 1 %. The 87V lets you decide whether to include or exclude that energy; the 117 only gives you the raw sum. Reversal: If you are troubleshooting a noise problem, you want to see the carrier — the 117 reveals what a sensitive PLC might see. The 87V’s LPF can hide the problem.

3. Peak capture: catching the transient that resets the controller

A generator synchronising to a grid can produce a voltage spike of 1.8 per‑unit for a quarter‑cycle (~4 ms at 60 Hz). The 87V’s Peak Capture grabs events as short as 250 µs; the 117 has no dedicated peak‑capture spec. On a typical generator start‑up, a 300 µs spike of 900 V on a 480 V line is invisible to a meter that samples at 2 ms intervals. The 87V stores it; the 117 gives you a normalised RMS value that missed the transient.

Worked example: An automatic transfer switch (ATS) logic controller resets during generator test cycles. You suspect a transient. With the 117 you see a steady 480 V — no clue. With the 87V in Peak mode you capture a 790 V spike that coincides with the ATS reset. The magnitude proportion: the spike lasts 0.05 % of a cycle, so its energy contribution to RMS is negligible, but its amplitude disrupts electronics. The 87V’s peak spec is the right tool. Reversal: In a clean grid environment with stable voltage, peak capture adds no value; the 117’s simpler measurement is sufficient and faster.

4. Magnitude proportion summary: when each meter leads

When the 87V is overkill — and the 117 is the right call

If the generator feed is exclusively through a dedicated transfer switch to resistive loads (space heaters, water baths), the extra harmonic content is negligible. The 117’s True‑RMS with ~20 kHz bandwidth is accurate to within 0.2 % on a typical grid [illustrative], and you save ~$250. Also, the 117’s Auto‑V/LoZ mode prevents ghost voltages from coupled capacitance on long generator cables — a real failure mode that the 87V lacks unless you manually switch to LoZ.

Rule‑of‑thumb threshold: If the generator load includes any power electronics (VFD, battery charger, UPS, LED lighting), choose the 87V. If the load is 100 % resistive, the 117 is sufficient and more portable.

Topology/standards per the cited standards; all product ratings are manufacturer-stated values from the cited datasheets, current to 2026-06; derived/illustrative figures are labelled as such. This is not an independent head-to-head test. Fluke is a brand affiliated with this site; competitor names are used for identification only.