Pressure Management

Pressure management on Roest is the practice of observing and shaping the pressure environment around the coffee while roasting. The main reader question is practical: how to measure pressure, how to translate readings into fan and heater settings, and how much pressure control is useful before it becomes noise. This page focuses on drum/exhaust pressure behavior, measurement discipline, target ranges, and troubleshooting; detailed fan theory, heat transfer, and model-specific roasting profiles are covered in linked pages.

What the Pressure Reading Means

Roest pressure readings are gauge or differential pressure readings: zero is atmospheric reference, and the displayed value represents a pressure difference rather than absolute atmospheric pressure. Tom Roest described the measurement as gauge pressure, specifically the difference between the drum and the nipple on the back plate 2 sources. In practice, positive pressure means the roaster is pushing more air into the measured space than is being evacuated at that point; negative pressure means the exhaust side is drawing relative to atmosphere.

Pressure and airflow are related but not identical. Changing exhaust, heater fan, temperature, power, batch load, bean size, venting, tilt, and even external air movement can change the reading. Tom Roest explicitly cautioned that pressure and airflow are not the same thing, noting that counterflow or tilt can make pressure more negative while airflow is lower because the inlet is blocked source. For broader airflow behavior, see Airflow and Fan Settings and Heat Transfer Fundamentals.

Measurement location matters. A drum/trier measurement, an exhaust measurement, and the Ultra’s integrated sensor can show different values for the same roast. Christopher Feran reported that an exhaust probe reads more negative than the drum/plug method, with a typical difference around 5–7 Pa source. Denis6004 also reported an Ultra digital sensor reading about 8 Pa when his analog trier meter read about 0 Pa source. Pressure targets should therefore be interpreted in the context of where they were measured.

Canonical Dial-In Procedure

The most useful pressure work is not chasing a perfectly stable number. It is mapping a given roaster, batch size, venting setup, and profile so the operator knows which fan settings produce neutral, slightly positive, and slightly negative conditions during a real roast.

Use the target batch size and beans. Pressure tests should be performed with beans while roasting, using the preferred batch size, because empty-drum readings and bean-loaded readings differ 2 sources.
Use a pressure instrument that resolves Pascals. A useful manometer should display Pa directly or at least 0.001 kPa increments, because the relevant Roest range is often only a few Pascals. A 0–50 Pa, 0–60 Pa, -30 to +30 Pa, or similar low-range meter is more useful than a coarse kPa-only meter 2 sources.
Mount and zero the meter consistently. Keep the meter in the same orientation for calibration and roasting, zero it with the tube detached, and do not move it after zeroing. Some meters drift with heat or orientation, so repeated zeroing may be necessary 2 sources. For environmental and calibration variables, see Calibration and Environment.
Seal the trier port properly if measuring there. A loose plug can pull false air, and beans can lodge near the trier hole and distort readings. Several users used silicone plugs, stainless straws, drilled plugs, or custom inserts; the critical requirement is a tight seal and a sampling point that does not let beans obstruct the port 2 sources. curated Use only heat-resistant materials at the trier port, such as metal or high-temperature silicone, and keep ordinary plastic manometer tubing outside the hot zone; if a plug, tube, or insert softens, smells, or deforms, stop and replace it with a heat-safe setup.
Record pressure at repeatable roast landmarks. Useful checkpoints are charge or early drying, yellow, around 175–190°C BT, first crack, and drop. Pressure changes with power and temperature, so readings should be taken under real roasting power rather than treated as static fan calibration 2 sources.
Adjust exhaust in small increments. A 30% to 100% fan change is a massive jump; smaller changes such as 3–5%, 5%, or 10% are the recurring practical adjustment size 2 sources. Roest’s manual control can change fan in 5% increments while roasting source.
Build a machine-specific pressure map. Do not copy another machine’s fan percentages blindly. The same nominal fan setting can be neutral on one machine and negative or positive on another because of exhaust setup, voltage, batch size, altitude, heater fan, and hardware variation 2 sources.

Starting Pressure Targets

These targets are for drum or trier-style measurements unless otherwise stated. They are starting ranges, not universal laws.

Roast stage	Practical starting target	Notes
Start to yellow	near neutral to slight positive, roughly 0 to +5 Pa	Denis6004 commonly described 0–5 Pa to yellow, with yellow around 150–155°C BT for his 180–185g batches 2 sources.
Yellow to late Maillard	neutral to low negative, about 0 to -5 Pa	Several reports converge on moving from neutral/positive early to slightly negative after yellow rather than staying positive the entire roast 2 sources.
First crack / development	low negative, often about -5 to -10 Pa	Denis6004 gave -5 to -10 Pa from yellow to drop and described -10 Pa as a personal upper threshold at drop 2 sources.
100g batches	often more negative early, around -2 to -6 Pa in Denis6004’s approach	Smaller batches behave differently and are more sensitive to airflow and heat loss; 100g should not be scaled directly from 150–185g settings source.
Excess negative warning zone	about -20 Pa and beyond may be too much for many drum/trier targets	Denis6004 described -20 or -25 Pa as too much for him, while other users have liked more negative profiles, so this is a caution zone rather than a hard limit 2 sources.

For batch-specific scaling, see Batch Size Scaling. For how these targets interact with first crack and drop, see First Crack Management and Development Time and Drop Decisions.

Interpreting Pressure Changes

Increasing the profile fan/exhaust generally pulls the system more negative; decreasing exhaust generally moves it toward neutral or positive. Heater fan and power can shift the entire pressure environment, so the same exhaust percentage can behave differently at different power levels. Denis6004 measured 40% air at 19% power near 0 Pa and 40% air at 75% power around -12.5 Pa in one empty heated test source. He also summarized one roast as roughly +5 to 0 Pa from start to 150/160°C BT, +3 to -1 Pa from 150 to 190°C BT, and -6 to -12 Pa above 190°C source.

Small pressure oscillations are normal. The drum contains moving beans while air is being pushed in and pulled out, so the reading may bounce several Pascals even when settings are unchanged. Denis6004 reported variations around 2–5 Pa while roasting and cautioned that drum pressure “to the dot” is not the point; the useful goal is staying in a practical range 2 sources.

Batch size strongly changes the pressure map. Lower batches such as 50–100g do not behave like 150–200g, and users repeatedly reported different neutral points and different sensory results between 100g and 185–200g. Denis6004’s 100g approach was around -2 Pa moving toward about -6 Pa, while his broader 150–185g range was closer to +2 to -10 Pa from start to finish 2 sources.

Sensory and Roast Effects

The community has used pressure as a practical tuning variable, but the mechanism is not fully settled. Within acceptable ranges, it is best treated as an optimization layered on top of the profile’s main intent rather than the primary design variable. Asamimasa framed pressure as secondary to the overall profile intent within rough ranges of acceptability source.

Near-neutral or slightly positive pressure early in the roast is often associated with more internal development, greater bean expansion, more body, and sometimes more sweetness. Too much positive pressure, however, is associated in several reports with smoky or roasty room aroma, stuck beans at the trier, pushed-out triers, overly puffy beans, faster staling, or an overdeveloped interior. Denis6004 warned that strongly positive pressure such as 30–10 Pa in the drum can reduce suction, keep smoke/chaff in the chamber, push the trier out, and create fire risk with some naturals and chaff source.

Low negative pressure after yellow and into development is commonly used to evacuate smoke, gases, and chaff without making the roast excessively high-air. It is often described as cleaner and brighter, but excessive negative pressure or high exhaust can also pull heat, reduce body, increase acidity, or produce brown/roasty notes depending on the profile and power application. Sorinl described a higher-fan comparison as more acidic and discreet with lower sweetness, while the lower-fan roast was more aroma-intense, sweeter, and deeper source.

Measurement Methods

Manometer

A manometer is the preferred tool when dialing in a machine. It should read in Pascals or fine kPa increments, and the operator should confirm that positive and negative direction are understood for the particular port connection. Denis6004’s analog dial was intentionally connected so that under 0 on his dial corresponded to positive pressure and above 0 corresponded to negative pressure; this sign convention depends on which port the hose is connected to 2 sources.

Cheap digital meters can drift or clamp near zero. Reports include meters that only read kPa, meters that needed recalibration after each roast, and meters whose readings changed with orientation or heat. Analog low-range gauges were favored by several users because they were easier to read and did not require recalibration between sessions in some setups 2 sources.

Lighter or Paper Test

The lighter or paper test is a quick qualitative check. A flame or thin paper drawn into the trier hole indicates negative pressure; hot air or beans coming outward indicates positive pressure. curated Treat the lighter test as a brief orientation check only: keep any flame brief, outside the trier opening, away from chaff and exhaust, and do not leave the trier port open longer than necessary; a strip of thin paper or tissue is a safer qualitative check when smoke or chaff is present. Denis6004 recommended pulling the trier during a roast and reducing exhaust until near zero if the flame is pulled in too much, but several users cautioned that removing the trier itself can affect readings and temperature behavior 2 sources. curated Removing the trier during a roast can release hot air, chaff, or beans, so use heat protection and keep your face and hands out of the path.

The lighter test is useful for orientation but not a replacement for a stable pressure map. Christopher Feran reported that his manometer showed neutral around 50% while the lighter showed neutral around 65–70%, illustrating that qualitative tests can disagree with instrument readings source.

Hardware and Model Notes

On L100/S100-style machines, pressure is commonly mapped through the trier/plug method or inferred by flame/paper tests. The old S100/L100 heater setting is discussed in RPM terms, often with stock at 3400 rpm and a practical lower boundary around 3100 rpm on common units; lower settings may trigger errors or risk heater stress 2 sources. For more on heater fan and RPM settings, see Drum Speed and RPM Settings and Airflow and Fan Settings.

On Ultra/L200, the integrated pressure sensor location differs from the older trier method. Christopher Feran described the Ultra pressure sensor as being in the exhaust, while the P3000 sensor sits between the coffee pile and inlet source. This means Ultra pressure values should not be copied directly from older L100/S100 trier readings without mapping the offset on that machine. See Roest Ultra Guide.

On P3000, the system is described differently: Tom Roest characterized it as a sealed positive-pressure system where positive pressure drives airflow, and noted that pressure is measured upstream of the beans in P3000 versus downstream of the beans in Ultra 2 sources. P3000 pressure numbers therefore should not be interpreted as if they were L100/S100 drum/trier numbers.

Troubleshooting Pressure Problems

Symptom	Likely cause	Fix
Reading stays at zero	Meter may be in coarse kPa units, clamped around zero, incorrectly zeroed, leaking, or blocked	Use a Pa-capable meter, zero with tube detached, check plug seal and tubing, and verify response by applying gentle pressure before roasting.
Readings drift between roasts	Heat, orientation, or meter instability	Keep the meter in the same position, zero before each roast if needed, and consider an analog low-range gauge if the digital meter drifts 2 sources.
Beans lodge at trier or trier pushes out	Pressure trending too positive, too low exhaust, or plug not occupying enough trier-hole depth	Increase the airflow profile by about 10–20% in the affected range or improve the plug; this was specifically suggested for stuck beans around 185g normal beans source.
Strong room smell, smoke not evacuated	Too positive or exhaust path not pulling enough	Increase exhaust modestly, check that the exhaust has free access to ventilation, and inspect chaff path and external venting. Positive exhaust pressure can indicate pressure buildup at the exhaust and smoke remaining in the roaster source.
Pressure is much more negative than expected	High exhaust, external fan/vent suction, cooling fan state, tilt/counterflow, or sensor-location offset	Remove external variables, test with the same batch and venting, and compare to hose-free or known baseline pressure. External fans and venting can materially change readings 2 sources.
Identical profiles give different pressure	Bean mass, BBP timing, heater fan oscillation, venting, cooling fan state, or measurement drift	Confirm the fan actually follows the profile, avoid roasting while cooling, repeat zeroing, and inspect pressure tubes and sensors. Kash5413 reported identical profiles alternating between -6/-8 Pa and +3/+5 Pa, with warm-up/BBP readings staying stable until bean mass was introduced source.
Chaff is burned or not collected	Fan/exhaust may be too low, or air path may be restricted	Empty the chaff tray before a roast and inspect chaff afterward; clean chaff and good evacuation indicate enough exhaust, while little or burned chaff suggests insufficient fan/exhaust for that roast 2 sources.

For cleaning and exhaust maintenance, see Maintenance and Cleaning. For chaff, smoke, tipping, and underdevelopment symptoms, see Roast Defects Troubleshooting.

Meaningful Disagreements

CONFLICT (Unresolved): pressure as a roasting variable versus airflow as the real control. Tom Roest argued that the pressure differences discussed in the sample roaster are minuscule and that, for heat transfer, he thinks in terms of airflow rather than pressure; he also reported P3000 testing where airflow effects were small to none when temperature was adjusted to make heat transfer identical 2 sources. Several experienced community members nevertheless use pressure measurements as a practical calibration tool for exhaust, machine matching, smoke evacuation, and cup-style tuning. The most conservative synthesis is to treat pressure as a useful proxy and setup tool, not as a standalone explanation for roast quality.

CONFLICT (Unresolved): positive early pressure versus always-negative pressure. Some contributors prefer slightly positive or neutral pressure until yellow or even near first crack, then a modest move negative for smoke and chaff. Others prefer staying negative but close to neutral from the start. Christopher Feran stated both a negative-through-roast preference in one discussion and later described running slightly positive until yellow/halfway, then neutral or slightly negative, then up to -10 at crack 2 sources. This is best handled as profile intent: use the starting targets above, cup the result, and adjust by small fan changes rather than treating either philosophy as universally correct.

Practical Takeaways

Pressure management is most useful during setup, profile transfer, troubleshooting, and batch-size changes. Once a machine’s neutral and low-negative fan ranges are known for the preferred batch sizes, many experienced users stop roasting with a manometer and simply keep the established fan map.

The most repeatable workflow is to map pressure with beans, in the actual venting setup, using Pa-scale readings; aim near neutral early and modestly negative later; avoid large fan jumps; and evaluate the results by cup, weight loss, color, and visible bean condition. Pressure is a guide, not the roast itself.