Why this roast exists

This was a ~52g sample roast done as part of validating whether the Gene Café CBR-301 behaves consistently at very small batch sizes.

Coffee

• Coffee: Costa Rica Fermented and Washed Geisha (Monte Brisas)

Setup (what I actually did)

• Roaster: Gene Café CBR-301
• Batch size: 52.5g green
• Set temperature: 250°C
• Start condition: cold start (start temp logged at 21°C)
• Cooling: initiated via the roast stop/cool action (this is the “end time” used below)

Measured results (from the roast log)

• Roast date/time: 2026-02-09 20:07:26
• Weight in → out: 52.5g → 45.1g
• Weight loss: 14.1%
• Total roast time (cooling initiated): 11:10 (670s)
• End temperature: 250°C
• Max temperature: 251°C
• Fan behaviour: Fan 2 → 3 at 4:22, then 3 → 2 at 7:50
• Crack marker in app: 10:20
  • Development (flag → cooling): 0:50 (~7.5% DTR)

What I observed during the roast (notes, not claims)

• This roast felt “textbook” in terms of how smoothly it ran.
• The roaster still showed the familiar “overheat / airflow” behaviour even at a tiny batch size (fan stepping to 3, then later returning to 2).
• First crack was audible to me, and I aimed for roughly ~1 minute after crack before stopping.

Interpretation (kept deliberately modest)

• The main thing I’m taking from this is that the machine still expresses its “normal” airflow/protection behaviour at ~52g.
• Weight loss (14.1%) gives me a solid anchor for roast level that doesn’t depend on how clearly I hear crack.

What I’ll do next

• Brew after a short rest window and keep notes simple: body, balance, finish (no flavour-note chasing).
• If I repeat small samples again, I’ll aim to keep start conditions consistent (cold start) so the timing comparisons mean something.

Source: Gene Café roast log export (JSON).