Airflow and Fan Settings
Airflow on ROEST is not a single variable. The profile “fan” setting, heater/inlet fan setting, exhaust path, cooling fan behavior, batch size, drum RPM, and machine generation all interact, so copied percentages often fail unless the pressure and airflow environment are also comparable. This guide explains how to think about airflow settings, how to establish practical starting values, and when to adjust fan settings during roasting.
What the ROEST Fan Controls Mean
In normal ROEST profile editing, the fan percentage refers to the exhaust fan: it controls how much air is pulled out of the roasting chamber. The heater/inlet fan is separate; it sits before the heater/inlet and pushes air through the heating element into the drum, with older S100/L100 machines exposing this as an RPM value such as 3100–3400 rpm and newer machines using percentage-style heater fan settings 2 sources.
The bean-cooling fan is another separate airflow path. It can influence drum or exhaust pressure while roasting, especially on newer machines and Ultra-style airflow paths, so it should be treated as a roast variable rather than an unrelated convenience setting 2 sources.
Airflow should not be treated as equivalent to pressure. Pressure is a useful proxy for the balance between air pushed into the drum and air pulled out, but the measured value depends on probe location, exhaust installation, power, batch size, bean movement, and machine version. For detailed pressure measurement, manometer selection, and pressure targets, see Pressure Management.
How Airflow Changes Heat Transfer
Higher airflow at the same air temperature transfers more energy to the coffee, but the way this appears on the graph depends on profile type. In a temperature-based profile, increasing airflow can increase power draw because the heater must heat a larger volume of air; in a power profile or manual mode, increasing airflow can make temperatures drop because more heat is being removed at the same power setting source. Tom ROEST summarized the practical principle as “less airflow = less heat transfer,” with higher temperature needed to compensate if the same roast time is desired source.
Batch size changes the air-to-bean ratio. Small batches leave more empty chamber volume, so hot air can bypass the bean mass and move toward the exhaust; larger batches cover more of the inlet area and absorb or restrict airflow differently. This is why settings that work at 100g often do not translate to 150–185g, and why 50–100g sample profiles need their own approach 2 sources. For broader scaling guidance, see Batch Size Scaling.
Practical Setup and Starting Values
Use this section as the canonical workflow for setting airflow. The percentages below are not universal recipes; they are starting points to be checked on the specific machine, batch size, venting setup, and heater fan setting.
Setup Procedure
- Choose the batch size first. Do not tune airflow on 100g and assume it applies to 160–185g, or the reverse.
- Confirm the rear intake mesh, exhaust path, and chaff path are clean. A clogged rear mesh or exhaust path changes back pressure and makes roasts behave differently.
- Set a heater/inlet fan baseline. For many S100/L100 users this is 3100–3400 rpm; 3100 rpm is often discussed as a lower practical setting, while 3400 rpm is stock/default on many machines.
- Roast with beans, not just an empty chamber, when checking pressure. Pressure tests should use the preferred batch size because beans change the airflow and pressure environment.
- Use a manometer where possible, or a lighter/paper check as a coarse indicator. curated If using a lighter, flame, tissue, or paper as a pressure indicator, keep it outside the roaster openings and away from chaff, the exhaust stream, and the cooling tray; do not let paper or tissue be sucked into the machine, because a manometer or smoke pencil is safer and more repeatable. The goal is not a perfectly stable number; normal roasting pressure oscillates.
- Adjust the exhaust fan in small steps. Several contributors recommend changes of about 3–5% or 5–10%, not large jumps, unless intentionally testing a broad effect.
- Compare one variable at a time. Changing heater fan, exhaust fan, batch size, charge, and inlet curve together makes it impossible to know what caused the result.
Starting-Point Table
| Use case | Heater / inlet fan | Exhaust fan starting point | Later exhaust adjustment | Notes |
|---|---|---|---|---|
| General S100/L100 starting point without a manometer | 3100 rpm | ~45% to yellow | 50–55% during Maillard, 55–60% around first crack | Offered as a “good place to start” for many users, not a universal pressure match source. |
| Denis-style 185g stock-machine approach | 3400 rpm | 35% | Ends around 58% | On his unit this ran from slightly positive/near-neutral early to negative in development; he cautioned the values are unit-specific source. |
| 180g/185g stepped exhaust example | Not specified | 45% | 50% at 150–160°C BT, 55% at 180°C, 60–65% at 190°C | Example for using less exhaust early and increasing later source. |
| 125g low-air profile example | 3100 rpm | 35% | 40% after color change, 45% into development | Used as a compact lower-air progression for inlet profiles source. |
| 100g profile with negative pressure throughout on one setup | 3200 rpm | 80% at charge | 75% at 110°C, 70% at 135–150°C, 65% at 160°C, 70% at 172°C, 75% at 185°C, 80% at 193°C | These were minimum values for one setup to maintain negative pressure through the drum; do not copy without checking the local pressure behavior source. |
| Small fan adjustments around yellow/FC | Existing baseline | Existing baseline | +5% after yellow, then +5–10% around 185–188°C or first crack | A conservative fan-step pattern for clearing moisture/chaff and moving more negative late source. |
Adjustment Map
| Observation | Likely airflow-related cause | Adjustment to test |
|---|---|---|
| Flame or paper is strongly sucked in; pressure is very negative | Exhaust is too high for that batch/setup, or external venting is pulling too hard | Reduce exhaust fan, check venting restrictions, or recalibrate pressure measurement. |
| Trier pushes out, hot air exits trier hole, room smell increases | Pressure is too positive; exhaust may be too low for smoke/chaff evacuation | Increase exhaust fan in small steps or reduce heater fan if appropriate; verify with manometer or flame. |
| Chaff remains dark/burned or room smells smoky | Insufficient evacuation, clogged exhaust path, or overly positive pressure | Clean exhaust and rear mesh; increase exhaust after yellow or near first crack. |
| Beans appear in chute/exhaust | Batch too large, airflow too high, RPM too high, or bean shape/size interaction | Reduce batch weight, reduce drum RPM, or reduce fan. For 200g, ROEST support text quoted in the community recommended RPM below 40 to reduce beans entering the exhaust source. |
| Roast becomes flat, thin, or underdeveloped when fan is high | Heat may be leaving the chamber or inner development may lag | Try lower exhaust, higher inlet/power, or a different batch size; compare by taste rather than graph alone. |
| Roast tastes heavy, brown, or overly developed with low fan/near-positive pressure | Too much heat retained or too much inner development | Increase exhaust slightly, shorten development, or lower inlet/power depending on the profile. |
Heater/Inlet Fan Baselines and Safety
Lower heater fan settings such as 3100–3200 rpm are widely discussed because many users found better cup quality, less roastiness, or more controlled pressure at lower inlet fan speeds. At the same time, the heater fan cools the heating element, and several contributors warn against going too low. ROEST has been reported as recommending not dropping below 3000 rpm, and several users report errors or shutdown behavior around 3000–3050 rpm on newer machines 2 sources.
Older machines were described as having lower default fan speeds such as 3200 rpm, while later machines were commonly discussed around 3400 rpm after heater failures led to more airflow for heater protection 2 sources. In practice, the safer documentation stance is to treat the heater fan as a machine-level calibration value: find a working setting, avoid repeated low-limit experimentation, and monitor errors or PCB temperature if changing it.
Newer Ultra/L200-style machines use heater fan percentages rather than the old RPM-style setting. The default heater fan ratio is commonly cited as 25%, while some users lowered it to 15–19% to match older pressure behavior or standard-mode profiles 2 sources. Ultra behavior differs enough that old S100/L100 airflow values should not be copied directly; see ROEST Ultra Guide.
Exhaust Fan Strategy by Roast Phase
Many successful approaches keep lower exhaust early, then add exhaust after yellowing or near first crack. The early low-exhaust phase is often used to avoid over-evacuating heat and moisture before the coffee has built enough momentum. Later exhaust increases help remove moisture, smoke, and chaff and move the chamber toward slightly negative pressure.
A common community pattern is near-neutral or slightly positive pressure through drying, then slight negative pressure approaching and through first crack. Several users target around 0 to low negative values before first crack and roughly -5 to -10 Pa late in the roast when measured in the drum, but the exact fan percentage needed varies dramatically by unit and venting setup 2 sources. For first crack timing and development decisions, see First Crack Management and Development Time and Drop Decisions.
Large fan jumps can create problems. A 30% to 100% jump was described as massive, and several users prefer 5% or 10% increments around yellowing and first crack unless they are deliberately stress-testing a profile 2 sources.
Batch Size Changes Everything
Small batches are more sensitive to airflow and fan changes. With 50–100g, bean temperature readings are less reliable, hot air can bypass the coffee more easily, and large fan changes can pull heat from the chamber quickly. For 100g, some users use higher exhaust to keep negative pressure and stable readings, while others use lower heater fan and lower exhaust for gentler sample roasting; both approaches require tuning on the specific unit.
For 150–185g, the bean pile covers more of the inlet area, changes pressure behavior, and generally gives more reliable BT data than 50–100g on older machines. Denis6004 repeatedly warns that learning a machine at 100g does not teach the same behavior as 160–185g, and vice versa source. For full batch-size strategy, see Batch Size Scaling.
At 200g, airflow and RPM constraints become more mechanical. Beans can enter the exhaust or chute, especially with higher airflow, high RPM, larger beans, or developed beans that have become lighter. ROEST support guidance quoted in the community recommends slower drum speed for 200g to avoid beans getting stuck, while some users prefer staying closer to 170–185g on S100/L100 for reliability 2 sources.
Sensory Effects and the Main Disagreement
Community tasting reports repeatedly connect airflow with cup character. Lower fan or near-neutral/positive pressure is often described as giving more body, sweetness, intensity, and heavier fruit; higher airflow or more negative pressure is often described as cleaner, brighter, thinner, more acidic, or more floral. Individual comparisons also report the opposite depending on batch size, machine, inlet curve, and roast duration, so these should be treated as tendencies rather than laws.
Some controlled or semi-controlled comparisons suggest airflow has little direct flavor impact when temperature, roast time, color, and loss are matched. Tom ROEST stated that pressure differences in the sample roaster are minuscule and that airflow’s effect is small to none if temperature is adjusted to make heat transfer identical; Christopherferan reported near-identical cup profiles in calibrated pressure comparisons 2 sources. Other community experiments report strong sensory differences from 5% fan changes or from 30% versus 65% airflow, including differences in sweetness, acidity, body, aroma, and dryness 2 sources.
The practical resolution is to use airflow as a controlled roast-design variable, not as an isolated magic setting. Match batch size, inlet curve, development target, and pressure range as closely as possible before attributing flavor differences to fan percentage alone. For theory, see Heat Transfer Fundamentals; for taste evaluation, see Cupping and Sensory Evaluation.
Venting, Environment, and Maintenance
The exhaust installation can change pressure and roast behavior. curated Treat roast exhaust as a safety issue as well as a profile variable: roasting produces smoke, irritants, and some carbon monoxide even on electric machines, so vent outdoors or ensure appropriate local ventilation, keep exhaust paths unobstructed, and use smoke/CO alarms where roasting indoors. Tube length, bends, inline fans, open windows, wind, exhaust restrictions, AirWave-style filters, and clogged mesh can all change the effective draw. A direct vent can work well, but external fans or filters should be dialed in against pressure behavior rather than simply attached and assumed neutral 2 sources.
The rear intake mesh and exhaust path are part of the airflow system. A clogged mesh changes back pressure and can make roasts differ; regular cleaning and vacuuming of the rear mesh, exhaust tube, and fan/cyclone area belongs in the normal workflow 2 sources. See Maintenance and Cleaning.
Idle cooling fan settings are contested. Some users turn the cooling fan idle to 0% to simplify pressure behavior, but Tom ROEST warned that setting idle fan to 0 while roasting can blow roasting gases out of the cooling tray, and other users prefer leaving idle fan active for room ventilation or consistency 2 sources. Treat the cooling fan idle value as part of the airflow setup and do not change it casually mid-profile.
Key Takeaways
Airflow settings cannot be copied reliably without matching batch size, heater fan setting, exhaust path, and pressure behavior. Start with conservative fan steps, measure or at least check the pressure trend with beans in the roaster, and change one variable at a time. Once a machine’s neutral and slightly negative ranges are known, airflow becomes easier to use as a practical tool for heat transfer, chaff/smoke evacuation, and cup-style tuning.
Experimental Notes on Exhaust, Inlet, and Surface Color
When comparing exhaust settings, treat single-batch results cautiously. In one comparison, 60% exhaust appeared to give a lighter bean exterior than 90% exhaust at similar inner development, even though the lower-exhaust roast ran longer; however, the author explicitly noted this was based on only one batch of each condition. In later multi-batch testing, the same tester reported that simply roasting longer with lower inlet temperatures did not make the outside lighter. Use replicated batches before concluding that a fan or inlet change is causing a surface-color difference. source
A practical interpretation offered in the discussion was that higher airflow renews the hot air around the beans more often. At the same measured inlet temperature, this can change effective heat transfer to the coffee because the air contacting the beans is refreshed more frequently; the inlet value describes the air reaching the beans, not the heater power required to maintain that condition. source
Lower airflow may also make some roasts appear more stable in PID-power behavior, but this should be evaluated together with roast outcome, pressure environment, and repeatability rather than treated as automatically better. source
For profile-matching experiments, first crack timing can be difficult to match exactly. If the goal is to isolate the effect of airflow or inlet, document the unavoidable FC mismatch and compare multiple batches instead of relying on a single matched curve. source
Community Notes: Lower Heater Fan and Staged Exhaust Air
A community-reported adjustment that produced a large perceived change was reducing the heater/inlet fan from 3400 RPM to 3100 RPM. This is worth treating as a heat-transfer and pressure-environment change, not merely an airflow-number change, because heater fan, exhaust fan, and exhaust restriction interact. source
One suggested exhaust-air starting pattern was to run relatively low air, around 50–55%, until yellowing, then increase to about 65% at first crack. This should be interpreted as a practical starting point rather than a universal recipe; machine generation, exhaust setup, batch size, pressure, and coffee density can all change the result. source
Another user described 75% fixed airflow as an older personal go-to setting for washed Janson Geisha, while noting interest in comparing it against different airflow approaches. This supports the broader point that fixed airflow can be a repeatable personal baseline, but it should be validated against cup results and the machine’s pressure/airflow environment. source
When changing fan settings, also check whether the drum remains slightly negative relative to the room/exhaust setup. A manometer is the preferred tool; one community member mentioned using a lighter as an informal check when no manometer is available, but any open-flame method should be treated cautiously around chaff, smoke, and hot equipment. source
Practitioner Batch-Size Note
Christopher Feran reported that he no longer uses 100 g batches in the L100. His current approach is either 50 g using inlet/time control, or larger batches around 150–160 g. Treat this as a practitioner-specific profile preference rather than a universal recommendation. source
Reported Default Exhaust During Preheat
One reported ROEST setup used a preheat temperature of 200°C AT with the default exhaust set to 3400 RPM. Treat this as a reference point for comparing preheat and exhaust conditions rather than a universal profile requirement, since airflow behavior can vary with machine setup and pressure environment. source
Community-Shared BT/PWR Profile Example
a_r_i_s shared a ROEST BT/PWR profile link as a possible profile to try: https://front.roestcoffee.com/sharedprofile/dc6969db-3b11-4609-9563-7f403149e753/ . They cautioned that the roast was run at sea level, on a power unit that seemed more powerful than most, and with the heater fan at 3100 RPM. Treat this as a transferable starting point only after checking whether your machine, elevation, power delivery, and fan behavior are comparable. source
In the same discussion, another community member said profiles around 150 g worked best for them, mainly using pure Inlet profiles but also Power profiles. They also referenced using 3200 RPM. This reinforces that batch size and fan RPM choices are user- and setup-dependent rather than universal defaults. source
The profile context was mostly natural-process coffees from Ethiopia, Panama, and Colombia, according to a_r_i_s. When adapting the shared profile, keep in mind that process and origin may affect how closely another coffee follows the same heat and airflow assumptions. source
Drum RPM and Bean Carryover Anecdote
A community member reported that, in their setup, running around 35–40 rpm did not send beans into the referenced upper area, while 40–45 rpm did. Treat this as a context-dependent anecdote rather than a universal threshold: bean movement and carryover can depend on batch size, airflow, roast phase, machine setup, and where exactly the beans were observed. source
Field Note: Air Calibration and Uneven Crack
One user reported an El Salvador Geisha 300h macerated coffee that appeared uneven at crack. Their planned adjustments were to reduce preheating, roast a smaller 80 g batch, and possibly increase air calibration from a current value of 0.93. Treat this as an anecdotal troubleshooting example rather than a general rule: unevenness at crack may involve preheat, batch size, airflow calibration, bean processing, and profile design together. source
Probe Reading Context
Christopher Feran described the machine probe as most nearly measuring exhaust, which is useful context when interpreting temperature readings alongside exhaust fan and airflow changes. Treat this as a practical interpretation of the probe signal rather than a full sensor-specification statement. source
RPM and Bean Movement Check
A community note suggested calibrating drum RPM so the beans are just moving but not jumping. Treat this as practical setup guidance rather than a fixed RPM value: the correct setting may depend on batch size, bean density, and the specific ROEST airflow environment. source
Diagnostic: Test for Low Airflow with a Smaller Batch
If a roast seems constrained late in the profile, one practical test is to either increase airflow in the later stages or repeat the same profile with a smaller batch, such as 90 g. If the smaller batch produces better results on the same profile, that can indicate that airflow was too low for the original batch size or roast conditions. source
Community Profile Example: 3100 RPM Heater Fan at Sea Level
A trusted community member shared a ROEST Front profile and noted that it uses a 3100 RPM heater fan setting at sea level. They advised charging at 180°C drum temperature, marking first crack at 192°C, and adjusting development time to taste. They also said the roast should reach 180°C at about 3:30. Treat these values as profile-specific guidance rather than universal targets, especially because heater fan behavior can vary with altitude and machine setup. source
Shared links from the same post: roast profile: https://front.roestcoffee.com/sharedprofile/c93e040c-73c7-4528-9b1c-845f538eecd1/ and associated BBP: https://front.roestcoffee.com/sharedprofile/8b635574-e891-4abf-8c35-dfce0e64f04c/ . source
Chaff Path and Drum Enclosure as Airflow Variables
When comparing ROEST to other compact roasters, users noted that the chaff path and drum enclosure can change the airflow environment substantially. One comparison was made to the Kaleido Strom 30, which was described as appearing to use a perforated drum with chaff falling downward, similar in concept to designs where a chaff collector sits below the drum. This was discussed as a design where chaff may remain within the broader roasting chamber rather than being immediately pulled away from the hot-air stream. source
A related design concern is that an unsealed or more open drum means the roaster must control not only the bean mass but also a larger surrounding air volume. Users suggested this can introduce more empty space around the beans and may allow temperature differences between upper and lower areas of the roaster. These observations are useful when comparing profiles across different machines: even if two roasters look similar externally, their chaff separation path, drum sealing, and internal air volume may make the same fan or heat settings behave differently. source
By contrast, ROEST was described as pulling chaff away from the hot air, which is one reason its exhaust/chaff path should be treated as part of the airflow system rather than as a separate housekeeping feature. source
Airflow Direction
ROEST airflow has been described as an inverted flow path: air enters from the top of the roasting system and exits at the bottom. This is useful context when thinking about how exhaust fan settings, drum agitation, and bean movement interact, because the fan percentage should be interpreted within this top-to-bottom air path rather than assumed to behave like a conventional bottom-to-top setup. source
Reported Airflow Path
A trusted user description states that air enters from the top round hole and exits through the bottom chaff collector. This is useful when reasoning about exhaust path, filtration, and where airflow restrictions or chaff buildup may affect the roast environment. source
Live Adjustment During a Roast
Heat and air are not necessarily fixed once a roast is underway: operators can change both heat and air during roasting. This means a saved program/profile should be treated as a starting framework rather than a completely hands-off constraint when active roast conditions require intervention. source
Exhaust Pressure Reference After Yellow
A practical reported reference point is that, when measuring exhaust pressure after yellowing, the reading may be around 80–140 Pa negative pressure. This should be treated as a pressure-environment reference rather than a universal fan percentage, because fan settings, batch size, machine generation, exhaust path, and airflow restrictions can all change the relationship between profile fan percentage and measured pressure. source
Visual Clues of a More Air-Roasted Result
A trusted contributor described a sample as looking “more air roasted” when the beans showed a more uneven exterior, a not-as-dark exterior color, signs of more internal pressure such as bigger cracks, and more “typing”/tipping-like marks, especially in naturals or processed coffees. Treat this as an observational cue rather than a standalone rule, because the cited comparison was clarified as a test rather than a matched roast, comparing 20 seconds vs 60 seconds of development time. source
Small-Batch Profile Adaptation Note
In a discussion about adapting a 100 g ROEST profile from a stock machine with a 3400 rpm heater, denis6004 emphasized not simply increasing airflow when moving between units. He advised finding the appropriate values for the specific machine and keeping the low-to-high fan change modest: roughly 31% to 40%, or about a 9–10 percentage-point delta, rather than “bumping airflow more.” source
For context, the referenced 100 g profile targeted first crack at 6 minutes or later, 40–50 seconds of development, and about 10–12% weight loss. Denis6004 said the profile was built by Sam at CornerStone. source
Denis6004 also cautioned that he no longer favors 50–100 g roasts for brewing/tasting, saying that from his perspective 100 g roasts do not reach the taste quality of 150–180 g roasts. Treat very small ROEST batches as sample-roasting compromises rather than assuming they will translate directly to larger, preferred batch sizes. source
He also noted that a 116 g Ethiopia example may roast longer than other coffees, reinforcing that bean behavior can change timing even when batch size is similar. source