Last roasts of the day
This was meant to be a tidy little roast session.
Two roasts of the same Honduras Catuai Natural. One cold start. One warm start. Then a third Honduras Pacas Washed to finish the day.
In practice, it became something more useful than a clean comparison.
One roast landed as expected.
One roast accidentally ran at a lower target temperature.
And the final roast, which also happened to be my 100th roast event if I include sample roasts, landed lighter than I expected.
That feels about right for this project.
Not a dramatic milestone.
Not a perfect experiment.
Just three more roasts, each with something to teach.
The coffees
The first two roasts used:
Honduras Catuai Natural
El Puente / Marysabel Caballero
88graines
HO2501
The third roast used:
Honduras Pacas Washed
Los Primos / Manuel Vallecillo
HO2520
All three were roasted on the Gene Café CBR-301.
These were personal roasts, not formal calibration roasts.
That distinction matters. The numbers are useful, but I do not want to pretend the session was cleaner than it was.
Session summary
| Roast | Coffee | Start | Batch in | Target | Time | Output | Weight loss | Note |
|---|---|---|---|---|---|---|---|---|
| Roast 95 | Honduras Catuai Natural HO2501 | Cold | 250 g | 249°C | 13:25 | 213.9 g | 14.4% | Clean personal cold roast |
| Roast 96 | Honduras Catuai Natural HO2501 | Warm | 245.6 g | 244°C | 12:55 | 210.7 g | 14.2% | Accidental lower-temp warm roast |
| Roast 97 | Honduras Pacas Washed HO2520 | Warm | 250 g | 250°C | 13:01 | 215.7 g | 13.7% | 100th roast event, lighter than expected |
Roast 95 — Honduras Catuai Natural, cold start
Roast 95 was the cleanest of the three.
It was a cold-start roast with a full 250 g batch. The logged target temperature was 249°C, and the roast was stopped at 13:25, exactly where I intended.
The result was:
250 g in → 213.9 g out
14.4% weight loss
That puts it in a lighter-balanced range for this machine. Not underdeveloped by the number alone, but certainly not a pushed roast.
The useful point is that the plan was followed, and the roast landed calmly.
Key observations
| Marker | Value |
|---|---|
| Start temperature | 22°C |
| 200°C | 4:42 |
| 230°C | 7:19 |
| 245°C | 9:12 |
| Cooling | 13:25 |
| Weight loss | 14.4% |
The fan behaviour was interesting. The fan stayed at level 2 for longer than I might have expected, then moved to level 3 quite late in the roast.
That is worth noting, but not over-reading.
For the purposes of this session, Roast 95 became the personal cold-start anchor.
Roast 96 — Honduras Catuai Natural, warm start, lower target
Roast 96 was supposed to be the warm-start partner to Roast 95.
But there was a catch.
The batch was slightly smaller at 245.6 g, and the machine was warm, starting at 65°C. But the target temperature was accidentally set or logged at 244°C, not 249°C or 250°C.
That means this roast should not be treated as a clean cold-versus-warm comparison.
The result was:
245.6 g in → 210.7 g out
14.2% weight loss
At first glance, that is surprising. A warm start and a slightly smaller batch might normally suggest a higher final weight loss. But the lower target temperature changes the whole interpretation.
Key observations
| Marker | Value |
|---|---|
| Start temperature | 65°C |
| 200°C | 3:24 |
| 230°C | 5:44 |
| 244°C | 7:30 |
| Cooling | 12:55 |
| Weight loss | 14.2% |
The warm start clearly accelerated the early roast. It reached 200°C and 230°C much faster than Roast 95.
But the lower target temperature appears to have limited the later roast energy.
So this roast became useful in a different way.
Not:
Warm start versus cold start.
But:
Warm start with lower top-end energy.
That is still a useful field note.
It reinforces something I keep coming back to with the Gene Café CBR-301:
A warmer start does not automatically mean a more developed roast if another part of the setup changes.
Roast 97 — Honduras Pacas Washed, warm start
Roast 97 was the final roast of the day.
It was also, if I include the three earlier sample roasts as roast events, my 100th roast event.
Formally, it is Roast 97.
As a project milestone, it is the 100th roast event.
That distinction feels messy, but honest.
This roast was a warm-start 250 g roast of Honduras Pacas Washed, with the target correctly set to 250°C. I planned to stop at around 13:00, and the roast ended at 13:01.
The result was:
250 g in → 215.7 g out
13.7% weight loss
That was lighter than I expected.
Key observations
| Marker | Value |
|---|---|
| Start temperature | 66°C |
| 200°C | 3:46 |
| 230°C | 6:21 |
| 245°C | 8:01 |
| 250°C | 8:43 |
| Cooling | 13:01 |
| Weight loss | 13.7% |
This is the interesting part.
The roast had:
- a warm start
- a full 250 g batch
- the correct 250°C target
- a planned 13-minute stop
And still, it landed at only 13.7% weight loss.
That does not make it a failed roast. It does mean the cup has to decide what the number means.
It may be a clean, lighter filter roast.
It may be a little underdeveloped.
It may simply be a coffee that responds differently from the Catuai Natural.
I do not know yet.
And that is the point.
What this session taught me
The useful lesson from this session is not that one roast was right and another was wrong.
The useful lesson is that three roasts in one day can look tidy on paper and still resist a simple conclusion.
Roast 95 showed that a cold 250 g Catuai Natural at 13:25 can land around 14.4%.
Roast 96 showed that a warm start does not guarantee a higher weight loss if the target temperature is lower.
Roast 97 showed that even with a warm start, full batch, 250°C target, and 13-minute stop, a different coffee can still land lighter than expected.
That is a good reminder.
The machine matters.
The setup matters.
The coffee matters.
The cup still gets the final vote.
What I am not concluding
I am not treating these three roasts as a clean experiment.
Roast 96 changed too many variables:
- warm start
- slightly smaller batch
- lower target temperature
- shorter total roast time
So it would be wrong to use it as proof of anything simple about cold versus warm starts.
I am also not treating Roast 97 as a formal calibration reference.
It matters because it is a useful record, and because it quietly marks the 100th roast event. But it still needs to be judged by drinking it.
What I will watch next
The next step is not to fix anything immediately.
The next step is to brew and taste.
For Roast 95 and Roast 96, I will be watching whether the small difference in weight loss is noticeable in the cup, especially given the lower target temperature on Roast 96.
For Roast 97, I will be watching whether the lower 13.7% weight loss drinks as:
- clean and light
- bright but pleasant
- thin or hollow
- underdeveloped
- or simply different from what I expected
If the Pacas Washed tastes good, I do not need to chase a number.
If it tastes thin or underdeveloped, then the next version probably needs more time.
Not a different philosophy.
Just a little more time.
Session note
This was a good closing session because it was not perfect.
Roast 95 was straightforward.
Roast 96 had a setup mistake.
Roast 97 became a quiet milestone and still landed lighter than expected.
That is probably a better picture of home roasting than a polished success story.
You plan.
You roast.
You notice the mistake.
You record the number.
You taste the coffee.
You learn something.
One roast at a time.