Video summary

I Did Caffeine Analysis: Some Unexpected Results!

Main summary

Key takeaways

Product Review

Product reviewed

Lighttells caffeine analyzer — an inexpensive (relative) handheld/bench-style caffeine testing machine used for measuring caffeine concentration in brewed coffee.

How it works (key user-experience points)

  • Uses single-use test chips inserted into the machine.
  • Each caffeine test also requires a single-use reagent kit (includes chip + pipettes + reagent).
  • Testing steps:
    1. Add coffee sample
    2. Mix with reagent
    3. Shake for ~10 seconds
    4. Cover the chip reading area
    5. Start analysis via phone connection
  • Output is given as:
    • Total caffeine in the drink, and
    • Caffeine per 100 mL (deciliter / dL)

Cost / practicality (major con)

Although the machine cost is described as cheaper than other analyzers (~$2,500), the consumables dominate ongoing cost:

  • Test kit cost: £7.50 per test
  • Chips/reagents can’t be reused, so every measurement requires fresh consumables.

Main findings about coffee (what the reviewer tested with the analyzer)

1) Filter coffee vs espresso (big surprise)

Using the same starting coffee and aiming for comparable extraction:

  • Double espresso (full shot): ~110 mg caffeine
  • Filter coffee (full beverage): ~170 mg caffeine

Result: Filter coffee had ~50% more caffeine despite similar extraction assumptions.

2) Caffeine extraction over time: filter vs espresso

Method: Brew normally, but collect liquid into multiple bowls/cups to analyze caffeine vs extraction “by mass” across the brew.

  • Filter coffee: caffeine extraction rises quickly early, then slows/levels as brew progresses.
  • Espresso: caffeine shows a lag early:
    • Total extraction happens quickly, but caffeine extraction ramps up later.

Comparison note (grinders):

  • With a Niche grinder, the caffeine lag is present.
  • With a bigger grinder (OPTION-O P100), the caffeine lag is more obvious, even though extraction starts fast—suggesting caffeine extraction may be linked to contact time / early brew dynamics.

Extra add-on (interruption by another speaker):

  • For longer espresso drinks (e.g., 3:1 “longer espresso/lungo”), expect more caffeine because caffeine is accelerating toward the end of extraction.

3) Aeropress contact time: diminishing caffeine returns

  • Dose tested: 12 g coffee to 200 g water
  • Contact time tested: from 30 seconds up to 20 minutes
  • Results:
    • More steep time increases extraction at first (expected).
    • But caffeine extraction flat-lines after ~2 minutes during the tested phase.

Conclusion: Very long steep times may taste better, but don’t keep increasing caffeine much past that threshold.

4) Instant coffee caffeine myth vs fresh coffee

The reviewer previously believed instant coffee might have more caffeine (often robusta-based), but testing showed the opposite.

Instant coffee vs pour-over / fresh:

  • Using manufacturer’s typical weak preparation (example strength: 0.9% beverage strength):
    • Instant: just under 40 mg caffeine per deciliter
  • If prepared to match filter strength (made stronger):
    • Instant: ~60 mg/dL
  • Measured pour overs:
    • Often ~80 mg/dL

Takeaway: Instant coffee can have about half the caffeine of comparable specialty pour-overs per serving size.

Additional context / comparison:

  • Instant manufacturers appear to extract a higher fraction of caffeine than typical assumptions (reviewer mentions a “ceiling” around ~30% for other extractions; instant seems to go higher).
  • Even the “strongest/harshest” cheap instant tested still reached only about ~55 mg/dL under industry-standard preparation.
  • Comparator example:
    • V60 brewed with Lavazza (contains some robusta): over 100 mg/dL
    • This was higher than specialty pour-over and much higher than instant.

5) Myth: darker roasts have less caffeine

Roasts tested:

  • Light (under-roasted)
  • Medium (normal)
  • Dark (to second crack)

Brewed to matching extraction levels (pour-over comparisons). Results (caffeine concentration):

  • Light: ~62 mg/dL
  • Medium: ~67.5 mg/dL
  • Dark: ~72 mg/dL

Conclusion: Darker roast had the highest caffeine concentration, contradicting “dark = less caffeine.”

Reviewer’s hypothesized explanation:

  • Darker roasts are less dense and more porous, requiring less mass to reach the same dose (and potentially enabling more caffeine extractability).
  • Even needing slightly more dark-roasted mass didn’t fully explain the increase; reviewer suspects extractability/porosity/contact effects also matter.
  • Note: a “sublimation” claim was mentioned as disputed by the data.

Pros / cons of the Lighttells analyzer (as stated/implied)

Pros

  • Enables direct, repeatable caffeine testing rather than relying on literature assumptions about brewing quality.
  • Produces actionable caffeine metrics (total and per 100 mL/dL).
  • Facilitates experiments isolating variables (grind/contact time/brew method/roast level).

Cons

  • High ongoing consumable cost: £7.50 per single test.
  • Single-use chips and reagents reduce flexibility and increase waste.
  • Test workflow described as somewhat “annoying”, requiring reagent prep each time.
  • Requires careful planning because testing is expensive.

Overall verdict / recommendation

Recommended for serious coffee researchers, enthusiasts, or labs who need real caffeine measurement across brew variables.

Not cost-effective for casual users due to single-use consumables (£7.50/test) and the need to run many test iterations for meaningful comparisons.

Unique points mentioned (consolidated)

  1. Lighttells caffeine analyzer; ~$2,500 device cost.
  2. Phone connection for operation.
  3. Measurement types: total caffeine and caffeine per 100 mL (dL).
  4. Requires single-use chip + reagent kit per test.
  5. Consumables cost: £7.50 per test.
  6. Brew-method surprise: filter ~170 mg vs espresso ~110 mg for comparable starting coffee assumptions (~50% more for filter).
  7. Time-course experiments show:
    • Filter: fast early caffeine extraction then slows.
    • Espresso: caffeine lags early, catching up later.
  8. Grinder comparison (Niche vs OPTION-O P100) changes clarity/extent of caffeine lag.
  9. Long espresso (3:1 lungo): more caffeine because caffeine accelerates toward the end.
  10. Aeropress contact time: caffeine rises then flat-lines after ~2 minutes (even up to 20 minutes).
  11. Instant coffee myth corrected:
    • Instant typically has much less caffeine than specialty pour-over.
    • Instant ~40 mg/dL (as instructed), up to ~60 mg/dL when matched strength.
    • Pour-over often ~80 mg/dL.
  12. Strongest cheap instant tested still ~55 mg/dL at standard preparation.
  13. Lavazza V60 example (robusta-containing): >100 mg/dL.
  14. Roast myth corrected:
    • Dark roast: ~72 mg/dL
    • Medium: ~67.5 mg/dL
    • Light: ~62 mg/dL
  15. Reviewer hypotheses: density/porosity and extractability differences; sublimation claim disputed by data.
  16. Variability across numbers due to differences in coffee beans and testing many coffees/grinders/brewers; Ethiopian coffees often showed slightly less caffeine.
  17. General rule-of-thumb: higher-grown coffee can have less caffeine, but with exceptions.

Speakers / contributors

  • Main speaker/reviewer: ran all tests, reported results and explanations; discussed machine operation and experiments.
  • Interrupted speaker (brief comment): noted that lungos/longer espresso (3:1) should yield more caffeine for the same overall extraction due to late-stage caffeine acceleration.

Original video