The roaster is a 1987 Probat P25. We bought it second-hand in 2014 from a roastery in Fitzroy that was upgrading to something newer and more digital. We paid less for it than you’d pay for a decent used car. It has a crack in one of the sight glasses that we’ve repaired three times. The temperature probe runs about four degrees hot and we’ve learned to read for it.
We have never seriously considered replacing it.
The argument for keeping old equipment
New roasters are faster to profile, easier to replicate, more consistent between batches. These are real advantages. They’re also, to varying degrees, irrelevant to what we’re doing.
A roaster that makes decisions based on real-time data curves is only as good as the person interpreting the curves. The machine doesn’t taste the coffee. The machine doesn’t know that the humidity in the roastery today is lower than usual, that the green came in slightly wetter than the moisture reading suggested, that the first crack on this particular Yirgacheffe tends to arrive eight seconds ahead of where the profile predicts it.
That knowledge lives in the person, not the equipment.
What eleven years does to your hands
By the fourth year, you stop relying on the sound of first crack to guide you. You start smelling it before you hear it. By the seventh year, you’re pulling samples every forty-five seconds not because you need to but because your body does it automatically while your attention is somewhere else. By the eleventh, the machine is an extension — not a tool you operate but something closer to an instrument you’ve learned to play.
You can’t buy that on a spec sheet.
We once had a visiting roaster from a well-funded Melbourne outfit spend the morning with us. They watched our operator work the P25 for two hours, asked a lot of questions, then said: “I couldn’t do this on our machine.” They meant it as a compliment to our process. We took it as a compliment to our roaster.
The actual cost of upgrading
The roasters we’d consider — the ones that would genuinely produce a different outcome rather than just a more comfortable experience — start at around $80,000. We could absorb that. The question isn’t whether we can afford to upgrade. It’s whether upgrading would make the coffee better.
So far, the answer is no. The day the answer changes, we’ll buy the machine. Until then, the crack in the sight glass stays where it is.
Coda
We’re aware this sounds like rationalisation. You keep the old Probat because it’s what you know, and you build a story around knowing it. That’s probably partly true.
But: the coffee is still exceptional. The drum is still the same drum. The window is still the same window. If it’s not broken — and it very clearly isn’t broken — we have no interest in fixing it.