Calibration and Environment

Calibration and environment determine whether two Roest profiles that look identical are actually comparable. This page covers the practical controls around measurement devices, room conditions, exhaust setup, voltage, altitude, and sensor interpretation so roasters can separate real roast behavior from measurement drift or site-specific effects.

Session Setup and Calibration Checklist

Use this checklist as the canonical setup routine before comparing profiles, sharing logs, or diagnosing roast-to-roast inconsistency.

Set up the room and exhaust before preheating.
Keep the roaster in a stable room, avoid open windows, and use a consistent exhaust path. The Roest manual clearance reported in discussion is at least 20 cm on all sides and 40 cm above the machine source. Wind, external fans, air purifiers, or changes in exhaust routing can change pressure and airflow behavior, so they should be treated as roast variables. curated Treat exhaust as a safety requirement as well as a calibration variable: roasting produces smoke and carbon monoxide, so use the manufacturer-recommended exhaust/ventilation setup, keep the exhaust path clear, do not block or restrict it, and do not roast in an enclosed unventilated room; a stable room is useful for repeatability, but it should not come at the expense of ventilation or fire/smoke safety.
Record the environment for every serious comparison.
Log room temperature, relative humidity, voltage at the socket, altitude/location, exhaust configuration, batch size, and profile mode. Community reports show that changes such as 20°C/50% RH versus 30°C/40% RH can materially change roast behavior for some users source. Humidity around the roaster can also drop during a session; one report measured 58% to 38% over two roasts source.
Warm the roaster consistently.
A full warmup is reported as 10 minutes, and other trusted guidance commonly uses 10–15 minutes before pressure or profile comparisons 2 sources. For drum-temperature-based starts, reported practices include heating until the drum is around 160°C, or using a specific drum-temperature end condition for repeatability 2 sources.
Stabilize the electrical load.
Avoid running other major appliances on the same circuit during power-profile work. Voltage affects available heater power, and a fixed power percentage is not necessarily the same roast energy if line voltage changes during the day source. If voltage fluctuates, a voltage stabilizer or conditioner may help, but it must be sized from the specific roaster nameplate/manual and use a continuous rating comfortably above maximum draw; many users choose 2 kW or higher for S/L machines where applicable. A discussed recommendation mentioned servo stabilization with ±3% regulation, but 1600 W may be insufficient for some S/L units; one user example was rated to 2200 W source.
Calibrate pressure measurement in the measuring position.
Place the manometer or dial where it will stay during roasting, then zero it without the Roest connected. The device should not be zeroed horizontally and then moved vertically; orientation changes readings 2 sources. Ensure the trier plug is tight, because leaks can pull false air and produce bad pressure values source.
If using a digital pressure meter, recheck zero often.
Several users found digital meters drifting after heat exposure or after a roast. The orange/yellow digital style was repeatedly reported as needing recalibration after each roast or before a second roast 2 sources. An analog -30 Pa to +30 Pa gauge was reported as easier to read and stable between sessions by one user source.
Use pressure values only from the same measurement location.
Drum/plug readings and exhaust-probe readings are not interchangeable; an exhaust probe can read about 5–7 Pa more negative than the drum/plug method source. The pressure targets themselves belong in Pressure Management.
Standardize color-meter preparation.
Clean the glass, use a consistent grind, and record the meter and calibration method. One workflow uses baking soda/natron as a calibration check around 230, then takes at least three readings, removes obvious high/low outliers, and averages the rest source. Grind size changes readings, so comparisons should specify grinder, burrs, and grind setting; very fine or espresso-range preparation was recommended for consistency source.
Treat moisture and water-activity numbers as device-specific.
Moisture readings should not be compared without naming the meter, calibration schema, device, and sampling protocol. The difference between ISO 6673 and ISO 1446 calibration schemas can be 1–1.4%, which is large enough to explain many apparent disagreements source. For chilled-mirror water activity on green coffee, calibration may be less critical in experience, but cleaning for dust and checking against a standard remain good practice source.
When a profile transfer fails, compare the calibration variables before changing the coffee.
Check voltage, batch size, inlet/ET/BT offsets, fan and heater settings, exhaust routing, pressure sign, and sensor location before concluding that the profile itself is bad. Shared profiles often need site-specific adjustment; this is covered further in Profile Sharing and Starting Points.

Environment Variables That Change Roast Behavior

Roest profiles are sensitive to the conditions around the machine because the roaster moves a lot of air and has low thermal mass. The most important environmental variables to record are room temperature, humidity, voltage, altitude, exhaust configuration, and whether outside air or wind can affect the exhaust.

A controlled room with no open windows is a repeatability advantage. One experienced user described roasting consistently at 21.5–23.5°C room temperature, 55–65% humidity, and no open windows source. Another reported that an open window at 16°C made even a 15-minute preheat feel too cold source.

Humidity is worth logging, but it should not be treated as a simple universal correction factor. Some practical reports say low or changing humidity alters timing and cup results; one user observed first crack moving toward 6 minutes when humidity dropped and associated that with burned results source. At the same time, a quoted simulation claim said relative humidity from 20% to 80% at ambient temperature did not affect simulated bean temperature after air was heated to 200°C, because the heated air became very dry source.

⚖️

Unresolved conflict

Community experience suggests room humidity can affect roast speed, pressure, or cup outcomes, while at least one cited simulation argues humidity has little direct effect on bean temperature once air is heated. The practical approach is to log humidity, control it when possible, and avoid assuming that another roaster’s humidity correction will transfer.

Voltage, Power, and Electrical Stability

Voltage primarily matters because it changes available heater power and makes power percentages less portable. One report gives 1650 W at 115 V and 1578 W at 110 V, meaning the same nominal 70% power can represent different actual wattage depending on supply voltage source. Another user reported power dropping from 227–228 V to 220 V or lower during roasting and added a voltage stabilizer to keep the supply more stable source.

Voltage does not appear to affect every motor in the same way. One explanation states that socket voltage affects the heater, while exhaust and RPM motors are regulated from the PCB and should not directly depend on socket voltage source. This distinction is important: unstable voltage may change heat delivery without changing displayed fan or RPM settings.

For power profiles, stable voltage is especially important because the profile commands a percentage rather than a target inlet temperature. Inlet profiles can compensate for some voltage and ambient variation through feedback, but they do not eliminate every site-specific difference. For more on when to use power versus inlet control, see Power Curve Strategies and Inlet Temperature Management.

For P3000 electrical planning, manufacturer guidance in the discussions listed the EU three-phase requirement as 16 A at 400 V, with 10.24 kW and 16 A per phase 2 sources. curated P3000 three-phase electrical work should be specified and installed by a qualified electrician according to the local Roest documentation and local electrical code; users should not adapt plugs, phases, breakers, or neutral/ground connections by trial and error.

Altitude, Air Density, and Exhaust Conditions

Altitude changes air density and therefore convection, pressure, and cooling behavior. Less dense air at elevation was described as less efficient for convection source. Users at higher elevations reported needing different heater fan or inlet settings; one report attributed Roest support guidance to raising heater fan to 3800 rpm or more above 1200 m, while another roaster reported increasing the hottest inlet point by 10°C at 1200 m+ and 15°C at 1500 m+ 2 sources.

These values should not be copied blindly. Pressure, airflow, and heat transfer are also affected by batch size, exhaust routing, voltage, and the specific unit. For altitude-related airflow and pressure adjustment, use Airflow and Fan Settings, Pressure Management, and Drum Speed / RPM Settings.

Exhaust routing can change measured pressure and roast behavior. Angled tubes, upward venting, outside restrictions, wind, and external fans have all been reported to affect pressure. A ventilator blowing in front of the Roest was reported to increase positive pressure inside the machine substantially during roasting source. Wind and changing outside air are therefore not background noise; they are part of the roast environment.

Sensor and Measurement Comparability

Roest sensor readings are not universal constants. ET, BT, inlet, drum temperature, pressure, moisture, and color all depend on where and how they are measured.

ET can change with batch weight, bean volume, bean size, and whether beans approach or touch the probe. Larger batches can partially cover or influence ET behavior, and at 160–185 g one user argued ET profiles may not work because the ET sensor can be partially covered by beans source. ET also changes with load: the same power profile at 100 g, 150 g, and 170 g can produce different ET readings source. For roast interpretation, see Bean Temperature Profiling and Rate of Rise Management.

BT is also batch-size dependent. At 100 g, several users treat BT and RoR cautiously because the probe may not be well immersed in the bean pile; larger batches such as 150–200 g often produce more normal-looking BT event temperatures 2 sources. This is one reason profile transfers should keep batch size fixed; see Batch Size Scaling.

Pressure readings must be checked for sign and port orientation. A reversed pressure dial can make a user believe +10 Pa is actually -10 Pa, because one port is positive and one is negative source. Some dials also show negative on the opposite side from what the user expects, so the sign convention should be verified before drawing conclusions.

Calibration Troubleshooting

Symptom	Likely cause	Fix
Manometer reads differently when moved	Orientation-sensitive calibration	Zero the device in its working position and do not move it during the roast source.
Digital pressure meter reads 0 before roast but not after	Heat-related drift	Recheck zero after one roast; if it drifted, recalibrate before using the next reading source.
Pressure values are wildly positive or negative	Reversed ports, wrong unit, leak, or exhaust restriction	Confirm Pa rather than kPa, verify positive/negative ports, check trier plug fit, and inspect exhaust path 2 sources.
Profile runs faster or slower than another machine with the same settings	Voltage, altitude, unit variation, batch size, or calibration mismatch	Record socket voltage, altitude, batch size, sensor offsets, and fan/heater settings before modifying the profile.
Color readings drift lower or higher	Dirty glass, grind-size change, sensor distance, or calibration change	Clean the glass, use the same grind and sample prep, calibrate/check against the same reference, and average multiple readings.
Moisture numbers disagree between users	Different meter, ISO schema, calibration, sample temperature, or range limit	Report the meter, calibration schema, sampling protocol, and sample temperature; do not compare raw values without that context source.
First crack is inaudible or not detected	Coffee/process behavior or machine detection limits	Inspect beans and ET/BT trends rather than relying only on crack detection; naturals and high-moisture coffees may crack faintly or not be registered 2 sources.
Back-to-back roasts drift even with same profile	Drum temperature or residual heat mismatch	Use a consistent cooldown or between-batch protocol; see Cooling and Between-Batch Protocol.

When to Stop Measuring

Pressure, color, moisture, and sensor checks are tools for calibration, not permanent distractions. Several experienced users treat a manometer as useful for finding a unit’s fan and pressure range, then stop using it once repeatable settings are established 2 sources. The same principle applies to environmental logging: record enough to understand the machine, then keep the variables that actually move results under control.