HighSpark vs OEM ignition coil test, 2022 Toyota Tacoma TRD Off-Road 3.5L V6
Raw GTC505 spark analyzer captures, published by Mod in Japan (modinjapan.com)
Captured November 12, 2026

WHAT IS IN THIS ZIP
Ten unmodified CSV exports from a GTC505 engine ignition analyzer (header identifier GTC505 213I), used in coil-pack mode. The files are byte-for-byte copies of the originals, with the original filenames. SHA-256 checksums are listed in SHA256SUMS.txt.

SETS (all are normalized captures, taken after a drive cycle that let the engine ECU adapt and normalize to the installed coils)
  HighSpark Premium, normalized:   GTC505_trends_0053.csv, GTC505_trends_0054.csv, GTC505_trends_0055.csv
  HighSpark Standard, normalized:  GTC505_trends_0058.csv, GTC505_trends_0059.csv, GTC505_trends_0060.csv, GTC505_trends_0061.csv, GTC505_trends_0062.csv
  OEM (factory) coils, normalized: GTC505_trends_0070.csv, GTC505_trends_0071.csv
  Note: HighSpark Standard is the street coil and corresponds to the files labeled "Regular" in our data folders. HighSpark Premium is the race coil. These are coil versions, not fuel grades.

TEST CONDITIONS
  Vehicle: one 2022 Toyota Tacoma TRD Off-Road with the 3.5L V6.
  After each coil set was installed, the truck was driven (about 15 minutes) so the ECU's adaptive learning could normalize. The captures were then taken with the truck parked, with no road load, while the engine was revved through its RPM range.
  Parked, no-load testing was chosen because it was safer and gave a cleaner signal than a road test with a 2.5 meter extension cable. A loaded test is planned for a later session.
  Not recorded: ambient conditions, spark plug type and gap, probe position.

FILE LAYOUT
  Row 1: instrument header, for example "GTC505 213I,system:,coil pack,cycle:,four-stroke".
  Row 2: column names as exported: Spark Index, Engine Speed (RPM), Dwell Time (ms), Peak Voltage (V), Burn Time (ms).
  Rows 3 onward: 300 spark rows with five numeric fields.
  Note: despite the Row 2 text, the data fields are, in order:
    1. Engine speed (RPM)
    2. Dwell time (ms, whole milliseconds)
    3. Peak voltage (V, usually negative because of secondary polarity; we use the absolute value divided by 1,000 as kV)
    4. Burn time (ms, whole milliseconds)
    5. Unused (always 0.00 in these files)
  Rows that are all zeros are blank sparks and carry no data.

HOW WE ANALYZED THEM (summary; see the full paper for detail)
  Cleaning, applied in analysis only (the files here are not changed):
    - drop RPM at or below 300 or at or above 7,000
    - drop dwell of 0
    - drop absolute peak voltage of exactly 2,400 V (treated as an instrument error code)
    - drop RPM more than 40% away from a centered 9-point rolling median
    - drop readings of 16 kV or more (presumed sensor noise)
    - drop dwell of 6 ms or longer (a capture artifact cluster at about 7 ms)
  Readings at or below 3 kV are counted as a detection floor (the analyzer bottoms out near 2.5 kV).
  The median is the main statistic reported per RPM band.

FULL METHOD
  The full write-up, charts, and limitations are at:
  https://www.modinjapan.com/mij-tech-research/highspark-vs-oem-ignition-coils-tacoma
