Inlet Temperature Management

Inlet temperature is one of the most useful control variables on Roest machines because it describes the hot air being delivered into the roast more directly than bean temperature or exhaust temperature. This page explains how to use inlet temperature as a profiling tool, how to adjust it across batch sizes and machine types, and how to diagnose common outcomes when inlet values are too high, too low, or poorly timed.

Why Inlet Temperature Matters on Roest

Roest behaves like a compact, high-airflow roaster, so the hot air path from heater to bean mass is short and changes in inlet temperature can affect the roast quickly. Inlet profiling is therefore often treated as a primary control method, while bean temperature and RoR are used more cautiously because they are influenced by batch size, probe contact, bean shape, agitation, and sensor delay. For more on those limitations, see Bean Temperature Profiling and Rate of Rise Management.

Inlet temperature is not the same as “heat applied.” The same inlet reading can occur with different airflow, heater fan, exhaust, pressure, and power conditions. Higher airflow can increase heat-transfer efficiency while allowing lower inlet setpoints for equivalent color, roast loss, and roast duration; lower airflow may require higher inlet temperatures to reach similar timing. This interaction is covered in more detail in Airflow and Fan Settings, Pressure Management, and Heat Transfer Fundamentals.

Inlet profiles are generally easier to transfer than power profiles, but they are not plug-and-play. Machine variation, altitude, exhaust setup, voltage, heater fan settings, sensor calibration, and batch size can shift the same profile by seconds, minutes, or even prevent a roast from reaching first crack. Shared profiles should be treated as starting points, not recipes; see Profile Sharing and Starting Points.

Before Adjusting Inlet

A roaster needs an inlet sensor to run inlet profiles as intended. Older L100/S100 configurations without inlet temperature should use power profiles rather than ET profiles, especially above roughly 120–130g, because larger batches can compromise ET readings; machines from P11 onward were described as upgradeable with the inlet kit, and users should not repurpose the ET sensor slot for inlet measurement because it is a different sensor type and is not intended for high inlet temperatures 2 sources.

Batch size changes the meaning of every inlet curve. A 50g, 100g, 150g, 180g, and 200g batch can show different sensor behavior and require different inlet shape, RPM, fan, exhaust, and charge settings. Treat batch size as a profile-defining variable rather than an afterthought; detailed batch strategy belongs in Batch Size Scaling.

Flow direction matters. Counterflow and Ultra-style roasting are more efficient and usually require substantially lower inlet values than normal-flow L100/S100 profiles. Multiple contributors reported reducing inlet profiles by about 40–60°C when moving from normal flow to counterflow or Ultra-style conditions 2 sources.

Actionable Starting Points and Adjustment Rules

Use the following table as a starting framework, not as fixed targets. The same numeric inlet may roast very differently on two machines, so the first validation should be timing to yellow, timing to first crack, weight loss, color, and cup quality. curated Before applying the 180–200g guidance, check the rated batch capacity for the specific Roest model; on smaller S100/L100-style machines, these larger batches may be outside normal beginner operation and should be treated as advanced/off-label unless the unit and setup are rated for them, with the batch aborted or reduced if bean movement is poor or beans enter the chute/exhaust.

For fine tuning, change one thing at a time. A broad inlet offset of ±5°C is often enough to test direction; ±10–15°C is a larger profile move; ±20°C can shift a roast substantially and should be verified by cup and color. Denis gave two rough power-to-inlet heuristics: each 1% power drop can be about 3°C inlet, and a 2% power drop can be about 5°C inlet, depending on the setup 2 sources.

A combined control move can change inlet quickly: one reported adjustment of a 5–7% power drop plus a 5% air increase lowered inlet by about 18°C in roughly 50 seconds source. Use that kind of move deliberately near first crack; accidental power drops combined with fan increases can cool the roast more than intended.

Inlet Shape Through the Roast

A common Roest inlet shape is a climb, a short maximum or plateau, then a controlled decline into first crack or development. Denis described this as a “climbing phase,” then a “max value phase,” then a slow decline source. Christopher’s guidance for one profile was to place peak inlet between yellow and crack rather than applying too much heat at the start and too little later source.

The early roast should not be judged only by peak inlet. A low start can prevent harsh surface effects in small batches, but too soft a start can produce flat, weak, green, or low-intensity cups. Higher charge or higher early inlet can help coffees that taste flat, weak, or underheated in P1/drying, especially some washed, dense, or high-moisture coffees. For phase timing, see Drying and Maillard Phases and Charge Temperature Guidelines.

Approaching first crack, the inlet curve often needs to flatten or decline before the crack event. Several contributors warned against rising inlet too late into first crack because it can produce flicks, crashes, or over-aggressive development. Christopher described rising ET as a strong predictor of a flick and recommended a larger inlet cut when flicks appear source. Denis similarly warned that a profile not flat or declining well before first crack can create a mini-crash depending on the bean source.

Development management depends on process. Washed coffees are often discussed as needing a more decisive inlet reduction after first crack, while naturals or processed coffees may need different handling and can sometimes tolerate or need different late-heat behavior. Use Washed Process Roasting, Natural Process Roasting, and Development Time and Drop Decisions for process-specific guidance.

Batch Size Effects on Inlet Strategy

Small batches have less thermal mass, less stable probe contact, and more sensitivity to changes in airflow, inlet, and RPM. A 50–100g roast may need a gentler start or a later power/inlet peak than larger batches; several experienced users caution that BT and RoR should not be overinterpreted at those sizes. Christopher uses inlet/time for 100g or less because IT/BT at that batch size became too “wonky” at crack source.

Mid-size batches around 125–160g are often easier to manage than 50–100g when reliable data is desired, but they still need machine-specific offsets. Roest has been described as doing well from about 120g upward if the goal is useful data, while 50g can still taste good if data is ignored source.

Large L100/S100-style batches around 180–200g change airflow, sensor contact, ET behavior, and bean movement. They may need lower RPM, different pressure targets, and different inlet curves. Some 200g approaches use higher inlet values, but 200g also creates chute, jamming, cooling, and sensor limitations; it should be treated as a separate operating mode rather than a scaled 100g profile. See Drum Speed and RPM Settings and Batch Size Scaling.

Airflow, Heater Fan, and Pressure Interactions

Inlet setpoint, airflow, and pressure cannot be separated. Higher airflow can allow lower inlet setpoints for the same duration, color, and roast loss, while lower airflow can require higher inlet values. Tom Roest summarized the heat-transfer side directly: less airflow means less heat transfer, so a higher temperature is needed to make up for lower airflow if the goal is the same roast time source.

On temperature-based profiles, increasing airflow can increase power draw because the heater must heat a larger air volume; on power profiles or manual control, increasing airflow can make temperatures drop source. This means that changing fan behavior without adjusting inlet can make a roast diverge even if the inlet setpoints look unchanged.

Pressure adds another layer. More exhaust/fan tends to move pressure negative, and pressure behavior varies by machine, venting, heater fan, power, and batch size. Several contributors use lighter tests or manometers to understand whether a given fan setting is positive, neutral, or negative on their own setup. For measurement procedure and pressure targets, use Pressure Management rather than treating a fan percentage as universal.

Machine, Voltage, and Environment Variation

Shared inlet profiles need offsets. Contributors reported needing to reduce or increase all inlet points by 5–20°C to make another person’s profile hit similar timings. Denis reported generally reducing inlet 5–10–15°C compared with some US-made profiles to get similar roast times, while other users had to increase inlet values to match timing 2 sources.

Voltage affects heater behavior most directly, especially with power profiles. One contributor noted that 110V operation can fail to hit inlet values as quickly, and Denis suggested increasing all inlets by 10°C for 110V in one Ultra/profile-transfer context 2 sources. Socket voltage does not necessarily affect fan or RPM motors in the same way, since those are regulated by the PCB, but it affects heater power source.

Ambient temperature, humidity, altitude, and exhaust setup also matter. Inlet profiles can reduce some variation compared with power profiles, but they do not remove all environmental effects. Elevation and temperature probe calibration were specifically cited as reasons individual roasters can behave differently even if nominally set up the same source.

Troubleshooting Inlet Problems

Conflicts and Open Questions

CONFLICT (Unresolved): high-inlet approach vs lower-air-temperature caution. Some contributors report strong results from high-inlet, fast, larger-batch approaches. Sam described targets such as 370–380°C peak after yellow, 350–340°C at first crack, and 335–330°C at drop for a high-inlet approach, while Denis reported 370–380°C+ inlet roasts without roasty batches across six beans in one testing period 2 sources. Others, especially a_r_i_s, explored keeping air temperature below 250°C and cited research concerns around desirable aromatics degrading in the 235–250°C air-temperature range 2 sources. These are not interchangeable measurements or profiles; inlet probe position, roaster design, airflow, and roast duration all affect how the numbers should be interpreted.

CONFLICT (Unresolved): profile transferability. Denis has repeatedly argued that shared inlet/BT profiles do not transfer reliably because machines and coffees diverge, while Sorin has argued that BT/IT profiles can be adjusted and tailored rather than being impossible to share source. The practical synthesis is to use shared profiles as frameworks, then tune inlet offsets, pressure, RPM, and charge for the local machine.

CONFLICT (Unresolved): 200g sensor stability. Todd Johnson suspected that 200g roasts dramatically alter temperature reading stability and make BT curves hard to use, while Christopher Feran stated the opposite: that 200g is incredibly stable for BT and inlet readings source. Treat 200g as machine- and setup-dependent and validate with cup, color, weight loss, and bean movement rather than assuming either result universally.

Practical Workflow

For a new inlet profile, start with a known batch size and do not change batch size, RPM, fan, pressure, and inlet all at once. Warm up consistently, record charge/drum conditions, roast one batch, then evaluate first crack timing, weight loss, color, bean appearance, and cup. If the roast is broadly close, use an inlet offset of 3–10°C rather than rebuilding the shape.

If the roast is far off, first confirm that the profile matches the hardware mode: normal flow versus counterflow, L100/S100 versus Ultra, inlet sensor present versus absent, and intended batch size. Then check airflow and pressure before chasing inlet values. If the profile came from another machine, expect to adjust it; the goal is matching the roast outcome, not copying the numeric inlet curve.