Compare Fast and Time-Lag Fuses at the Same Starting Current

Two cartridge fuses can both be rated 250 mA and take different amounts of time to open. Apply a measured current near 1 A to one fast-acting fuse and one time-lag fuse. Watch the source voltage as well as the current. That second trace tells you whether the current stopped while the bench supply was still on.

Build a covered low-voltage fixture

Before connecting any input or output cables, connect the RIGOL DHO804's rear Ground Terminal to earth ground using the supplied ground cable, as required by the DHO800 User Guide. Its Type-C power input does not ground the instrument through the power cord. Use the supplied power adapter. Keep the SPD3303X's protective earth connected. The probe ground clips connect to G0 as described below; they do not replace this required chassis earth lead.

A current-limited SIGLENT bench supply drives a covered fuse, a 2.2 ohm power resistor and a 0.100 ohm low-side shunt in series. Both RIGOL probe grounds connect to the supply return. The current falls both when a fuse opens and when the supply is disabled, while the positive supply voltage distinguishes the two cases.
Channel 1 measures the shunt and channel 2 measures the supply terminal. The lower traces show calculated shapes around the cutoff. The drawing omits the DHO804 power adapter and separate chassis earth lead; connect them as described before adding any input or output cables. Full-size PNG · Editable SVG

Use new Littelfuse 0217.250MXP fast-acting and 0218.250MXP time-lag glass cartridge fuses, both 5 × 20 mm and 250 mA. Use a new 0217002.MXP 2 A cartridge from the 217 series for the control. Check carton markings and the fuse family's suffix before installation. Use a covered 5 × 20 mm holder with insulated terminals. Littelfuse lists the complete 345 holder 03455LF2H, or the matching body 03453LF2ZX010 and 5 × 20 mm knob 03455LF1HX020.

Install one fuse at a time with the power supply output off. Connect the positive output of an independent SIGLENT SPD3303X channel through the holder, a 2.2 Ω, 10 W power resistor and a 0.100 Ω, 1%, 1 W shunt in series. Return the lower shunt terminal to the channel negative terminal. This return is G0. The power resistor limits available current even when the supply has voltage across it before the current regulator responds. Keep it on a heat-resistant surface. Use the covered low-energy fixture on the bench only, with no mains wiring or battery connected.

Set the SPD3303X to a 4.50 V maximum and 1.00 A current. A current-regulated bench source raises its output voltage as needed until it reaches that voltage ceiling. This available voltage is called compliance. The 217 datasheet lists about 1.23 Ω nominal cold resistance for its 250 mA fuse and the 218 lists about 1.05 Ω. The cold voltage demand at 1 A is therefore about 3.53 V and 3.35 V respectively, including the other resistors. Resistance can rise while the fuse heats. Watch the actual current and reject a record where it cannot hold the near-1 A plateau. Keep the 4.50 V ceiling in place. At 1 A the 2.2 Ω resistor dissipates 2.2 W and the shunt dissipates about 0.10 W. With only those two nominal resistors across 4.50 V, their combined resistance would cap the ideal current at about 1.96 A, even before the current regulator acts. Check the actual capture for overshoot.

Connect two RIGOL PVP2350 probes to a DHO804, set both probes and channel readouts to 1×, and use direct-current (DC) coupling. Put channel 1's tip directly at the upper shunt terminal and channel 2's tip at the supply positive terminal. Put both ground clips directly at the lower shunt terminal, G0. Preserve the scope's protective earth. The shunt converts current to voltage by I = Vshunt / Rshunt. At 1.00 A and 0.100 Ω it gives 100 mV. Use short sense connections directly at the resistor terminals.

Use the control to qualify the source and capture

This test consumes the two 250 mA cartridges if they open. Wear eye protection, keep the holder closed while energized and turn the output off before handling a cartridge. Start each run with a new sample at room temperature. For this low-voltage fixture, record pre-test continuity with a Fluke 115 or equivalent on the loose fuse, and recheck after the test with the supply off and the fuse cool. The AC breaking capacity on the fuse label and sheet does not qualify any specific DC protection application.

  1. Install the 2 A control. Arm a single rising-edge capture on channel 1 near 50 mV with at least 0.10 s of pre-trigger trace and at least 4.2 s after trigger. Use a simultaneous memory capture with both channels enabled. Select a depth around 100 kpts and verify the actual sample interval is 0.50 ms or finer. Set the voltage scales to keep both traces visible.
  2. Enable the supply. Confirm the current reaches about 1 A quickly. Turn the output off after roughly half a second. Check that both current and the supply voltage fall, and confirm the control fuse still has continuity after power is off. If the current overshoots 1.10 A, rises from 0.10 A to 0.90 A in more than 20 ms, or fails to stay near 1 A, resolve the supply or measurement first. A slow front-panel turn-on can obscure a short fuse event.
  3. With the output off, repeat once with each fresh 250 mA fuse. Allow at least 0.1 s after a current collapse to see whether the source stayed energized, within an absolute four-second energized limit. Switch the source off at that limit even if no cutoff was seen. A late cutoff that leaves too little evidence is unresolved. After the fuse cools, remove it and verify permanent loss of continuity. Keep both raw captures and the blank observation sheet with the physical checks.

Compare timing within this fixture

From each scope capture use Storage → Save → Choose → Save Wave, Memory and the CSV (comma-separated values) format. Keep the native exports. In a copy remove instrument header lines and retain the simultaneous time, shunt voltage and supply voltage in seconds and volts. Give each row the corresponding name, control, fast or delay. Combine the rows under one header, mode,time_s,shunt_v,supply_v. Preserve original sample times within each capture.

Run python3 fuse_opening.py capture.csv --shunt-ohm 0.100. Use a verified low-resistance measurement if available, or the precision shunt's nominal value. A two-wire ohmmeter can include too much lead resistance to improve the nominal value at 0.100 Ω.

The program estimates the elapsed time from the first rising 0.90 A crossing to the first falling 0.10 A crossing. The measurable fixture window is 8 ms to 4 s. The program screens the intervening plateau, rejects a return of current after the first cutoff and requires comparable pre-cutoff current across the runs. It identifies a possible opened current path when the source stays energized through the cutoff and above 4.0 V in the following observation window. The separate unpowered continuity check establishes whether the cartridge stayed open. Computed tenths of a millisecond are formatted estimates. Interpret them in light of the sample interval, current rise and voltage measurement uncertainty.

Littelfuse 400% opening-time rows for these nominal ratings
TypeMinimumMaximum
217 fast10 ms300 ms
218 time-lag150 ms3000 ms

These published test ranges overlap. They provide context for why the timing classes differ, while the small, current-limited fixture measures its own timing marker. Your observed order depends on the individual samples and the qualified current. The test does not measure interrupting capacity, event energy in a fault circuit or suitability for any application. It does not judge production compliance against the datasheet table.

Exit 0 means the three waveforms meet the timing comparison's gates, with current collapse while the source remained on in both 250 mA traces. Exit 1 reports incomplete or ambiguous waveforms. Exit 2 reports a file or parsing error. The example and automated tests use synthesized waveforms. Physical bench results are left to the reader.

Challenge

One sample's current collapses after 200 ms and the source stays near 4.5 V. Can the time identify the fuse family? What does a later power-off continuity check add?

Answer

200 ms lies inside both published time bands, so elapsed time alone cannot identify the family. The live source trace shows the current path stopped with voltage still available. Loss of continuity measured with the cartridge removed and the circuit off confirms that this fuse remained open.

Complete Python analyzer

Download the standalone analyzer or use the experiment archive for the editable diagram, observation sheet, synthetic data and tests.

Complete source for fuse_opening.py
#!/usr/bin/env python3
"""Compare low-energy fuse opening traces with a manual output-off control."""

import argparse
from bisect import bisect_left, bisect_right
import csv
import math
from pathlib import Path
from statistics import median
import sys
from typing import NamedTuple


HEADER = ("mode", "time_s", "shunt_v", "supply_v")
MODES = ("control", "fast", "delay")
ROUNDING = 1e-9


class EvidenceError(ValueError):
    """The records do not support a controlled timing comparison."""


class Outcome(NamedTuple):
    elapsed: float
    plateau_a: float
    post_v: float
    classed_as: str


def require(ok, label, reason):
    if not ok:
        raise EvidenceError(f"{label}: {reason}")


def parse(rows):
    rows = iter(rows)
    require(tuple(next(rows, [])) == HEADER, "CSV", "use mode,time_s,shunt_v,supply_v")
    records = {name: [] for name in MODES}
    for number, row in enumerate(rows, 2):
        label = f"line {number}"
        require(number <= 330_001 and len(row) == 4, label, "use four fields and at most 330000 rows")
        require(row[0] in records, label, "unrecognized mode")
        try:
            sample = tuple(map(float, row[1:]))
        except ValueError as error:
            raise EvidenceError(f"{label}: enter numeric seconds and volts") from error
        require(all(map(math.isfinite, sample)), label, "samples must be finite")
        records[row[0]].append(sample)
    return records


def threshold_crossing(time, current, index, level):
    before, after = current[index-1], current[index]
    frac = (level-before)/(after-before)
    return time[index-1] + frac*(time[index]-time[index-1])


def analyze(rows, mode, ohm):
    require(bool(rows), mode, "no capture")
    time, shunt, supply = zip(*rows)
    current = [value / ohm for value in shunt]
    require(-10 <= time[0] <= time[-1] <= 10
            and all(0 < b-a <= .0005 + ROUNDING for a,b in zip(time,time[1:])),
            mode, "non-monotonic time or sample gap above 0.50 ms")
    require(all(-.08 <= v <= 4.60 for v in supply), mode, "measured supply outside the low-voltage fixture")
    require(min(current) >= -.04 and max(current) <= 1.10 + ROUNDING,
            mode, "overshoot or reversed shunt voltage")
    low = next((i for i,v in enumerate(current) if v >= .10), None)
    first = next((i for i,v in enumerate(current) if v >= .90), None)
    require(low is not None and first is not None, mode, "no near-1 A current pulse")
    quiet_start = bisect_left(time,time[low]-.030)
    quiet_end = bisect_right(time,time[low]-.005)
    require(low > 0 and time[low] - time[0] >= .030 - ROUNDING
            and quiet_end - quiet_start >= 5
            and all(i <= .04 and v <= .20 for i,v in
                    zip(current[quiet_start:quiet_end],supply[quiet_start:quiet_end])),
            mode, "need a quiet pre-trigger with the supply output off")
    first_s = threshold_crossing(time,current,first,.90)
    low_s = threshold_crossing(time,current,low,.10)
    require(first_s-low_s <= .020 + ROUNDING, mode, "current rise took longer than 20 ms")
    end = next((j for j in range(first+1,len(time)) if current[j] <= .10), None)
    require(end is not None, mode, "no low-current endpoint captured")
    end_s = threshold_crossing(time,current,end,.10)
    elapsed = end_s-first_s
    require(.008 - ROUNDING <= elapsed <= 4.0 + ROUNDING,
            mode, "duration outside the measurable 8 ms to 4 s fixture window")
    inner_start = bisect_left(time,first_s + .002)
    inner_end = bisect_right(time,end_s - .002)
    middle = current[inner_start:inner_end]
    require(len(middle) >= 5 and min(middle) >= .93 - ROUNDING and max(middle) <= 1.07 + ROUNDING
            and .95 - ROUNDING <= median(middle) <= 1.05 + ROUNDING,
            mode, "no qualified near-1 A plateau")
    post_start = bisect_left(time,end_s+.005)
    post_end = bisect_right(time,end_s+.055)
    require(time[-1] >= end_s+.055 - ROUNDING, mode, "no post-event observation window")
    require(max(current[end:post_start], default=0.0) <= .10 + ROUNDING
            and max(current[post_start:]) <= .05 + ROUNDING,
            mode, "current returned after the first cutoff")
    after = supply[post_start:post_end]
    if mode == "control":
        require(max(after) <= .20 + ROUNDING, mode, "control never showed a disabled source")
        kind = "output disabled"
    else:
        edge_start = bisect_left(time,end_s)
        require(min(supply[edge_start:post_end]) >= 2.0 - ROUNDING
                and min(after) >= 4.0 - ROUNDING, mode, "source fell or did not remain energized after cutoff")
        kind = "current path opened"
    return Outcome(elapsed,median(middle),median(after),kind)


def compare(records, ohm):
    require(.095 <= ohm <= .105, "fixture", "shunt must be near 0.100 ohm")
    results = {mode: analyze(records[mode],mode,ohm) for mode in MODES}
    plateaus = [r.plateau_a for r in results.values()]
    require(max(plateaus)-min(plateaus) <= .05 + ROUNDING,
            "fixture", "current plateaus do not match within 0.05 A")
    return results


def main():
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("observations")
    parser.add_argument("--shunt-ohm", type=float, default=.100)
    args = parser.parse_args()
    try:
        with Path(args.observations).open(newline="",encoding="utf-8-sig") as stream:
            results = compare(parse(csv.reader(stream,strict=True)),args.shunt_ohm)
    except EvidenceError as error:
        print(f"UNRESOLVED: {error}",file=sys.stderr)
        return 1
    except (OSError,UnicodeError,csv.Error) as error:
        print(f"ERROR: {error}",file=sys.stderr)
        return 2
    for mode, outcome in results.items():
        print(f"{mode:8} {outcome.elapsed*1000:8.1f} ms  {outcome.plateau_a:.3f} A  "
              f"{outcome.post_v:.2f} V after cutoff  {outcome.classed_as}")
    print("Confirm permanent fuse opening with a separate power-off continuity check.")
    return 0


if __name__ == "__main__":
    sys.exit(main())

Related experiments and primary sources