Drum Speed / RPM Settings
Drum speed on Roest controls the agitation paddles rather than a rotating drum, but the setting still has a large effect on bean movement, heat transfer, sensor readings, and whether beans are carried into the exhaust path. The practical question is not “what RPM is best,” but which RPM range matches the batch size, hardware generation, bean size, process, and profile style. This page gives starting points and adjustment logic for drum-speed decisions; detailed airflow, pressure, and defect work should be handled through Airflow and Fan Settings, Pressure Management, and Roast Defects Troubleshooting.
What Drum Speed Changes on Roest
On Roest, the drum itself is described as fixed while the internal propellers or paddles move the coffee, so “drum RPM” is best understood as main motor agitation speed rather than a conventional rotating-drum speed source. Higher RPM throws more beans into the hot air stream and generally increases heat transfer, while lower RPM leaves the bean pile more stacked and reduces energy transfer efficiency 2 sources.
Because of that, RPM is not a minor setting. Multiple discussions treat main motor speed as more influential than airflow changes for roast behavior, especially because agitation changes both hot-air exposure and how the bean pile interacts with the inlet area 2 sources. RPM also interacts with Batch Size Scaling: a 100 g batch, a 150 g batch, and a 200 g batch do not expose the probes, inlet, and exhaust path in the same way.
Starting Settings and Adjustment Rules
Use the table below as the canonical starting point. These are not universal profile values; they are safe places to begin before tuning inlet, power, airflow, and development time.
| Batch / use case | Practical RPM starting point | Adjustment logic |
|---|---|---|
| 50 g samples | Start around 55 RPM if following a fast sample-style ET profile; try 45 RPM if beans are flying excessively or if the profile is too fast. BT readings should not be trusted heavily at this size 2 sources. | Use 50 g mainly for testing and sample work. Judge by ET/inlet behavior, first crack, color, and cup rather than BT. |
| 100–120 g | Start at 55 RPM. 60–65 RPM can be used for faster, more energetic roasts when the bean shape tolerates it; 45–50 RPM can improve BT contact but may slow or unevenly mix the roast 2 sources. | If the roast is slow or underdeveloped after lowering RPM, increase heat or return toward 55. If BT is noisy or appears air-influenced, consider lowering near first crack rather than running low RPM for the whole roast. |
| 125–150 g | 50–55 RPM is the normal starting range. 55 RPM is explicitly described as fine for 150 g, and 50 vs 55 RPM is often close at 125 g 2 sources. | Use 55 when more energy transfer or evenness is needed. Use 50 when taste trends toward too much heat, roastiness, or bitterness. |
| 160–185 g | Start around 50–55 RPM, then inspect the chute and cup. 185 g at 55 RPM has been reported without beans in the exhaust chute, but above roughly 180 g becomes more constrained by bean size and chute behavior 2 sources. | If beans reach the chute, lower RPM before increasing batch size further. If the cup becomes flat or too “drum-like,” test a higher RPM with adjusted inlet. |
| 200 g on S100/L100-style machines | Start at 40–45 RPM if following Roest-style guidance; move toward 35–39 RPM if beans enter the exhaust or chute 2 sources. | Inspect the exhaust/chaff path after early tests. If beans still leave the drum after lowering RPM, reduce fan slightly rather than pushing RPM higher. |
| 200 g on Ultra / L200-style machines | Start lower than an equivalent older-machine setting if the real agitation looks faster. For small beans that stick, reducing by about 5 RPM has been reported to solve the issue in counterflow without losing roasting performance source. | Visually confirm bean movement. Hardware generation and motor control can make the same displayed RPM behave differently. |
| P3000 small batches | For P3000, 1 kg is described as fine with slightly lower RPM, while Tom’s quoted optimal ranges were 40–50 RPM at 1 kg and 55–65 RPM at 3 kg 2 sources. | Do not transfer L100/S100 RPM assumptions directly to P3000. Use P3000-specific airflow and pressure behavior. |
The first adjustment after changing RPM should usually be roast time and heat-transfer observation, not cup interpretation alone. Lowering RPM often slows the roast: one 100 g comparison found 55 RPM roughly 2 minutes faster than 35 RPM, and another user saw a roughly 30 second increase after lowering RPM 2 sources. If RPM is lowered and the profile becomes too slow, adjust inlet or power deliberately rather than assuming the same profile will translate.
Batch Size and Bean Movement
Small batches expose the Roest’s BT probe and hot-air path differently from larger batches. At very low charge weights, the probe can spend more time reading air than beans, and 50 g BT readings have repeatedly been described as unreliable 2 sources. At 100 g with 55 RPM, the BT probe may read a mixture of air and bean temperature, so a smooth BT/RoR curve is not necessarily proof of better bean development source.
Larger batches improve bean coverage but create different constraints. At 200 g, the bean pile can cover the inlet, making convection still strong even at lower RPM; the goal shifts from lifting beans into the airflow to keeping a full drum well mixed source. The same fullness also increases the risk of beans reaching the exhaust, chute, trier opening, or ventilation exit, especially as beans expand later in the roast.
Bean size and shape matter. Small, round, or irregular beans are repeatedly linked with sticking or jamming, while larger beans at high batch weights may need lower RPM to avoid the chute. Jams are not purely a batch-size problem; they have been reported at both high and lower RPM when very small beans or awkward shapes are involved 2 sources.
High RPM: When It Helps and When It Fails
Higher RPM is useful when the roast needs more hot-air exposure, faster heat transfer, or better mixing. Several contributors report more even roasts or faster progress at higher RPM, especially in smaller batches or dense washed coffees. A 100 g setting of 65 RPM has been described as safe by one source, and 65 RPM is commonly discussed for fast or high-energy profiles source.
The risk is that high RPM can become too aggressive. Reports around 65–70 RPM include false crack-detection noise, propeller or bean sticking, beans being dragged rather than mixed, and internal overdevelopment or burned pockets in some coffees 2 sources. High RPM can also make the BT and RoR plots misleading because more air reaches the probe, especially in small batches.
Use high RPM as a controlled profile tool, not as a universal quality upgrade. It is more defensible for 100–125 g or specific high-density washed coffees than for full 200 g batches, unless the machine, bean size, and exhaust behavior have been validated.
Low RPM: When It Helps and When It Fails
Lower RPM can improve probe contact, reduce beans entering the chute, and shift the roast toward a more stacked bean pile. It is often the first practical fix for 200 g batches that throw beans into the exhaust. It can also change sensory balance: some comparisons found lower RPM roasts slower, darker, more developed, or cleaner, depending on the profile and coffee 2 sources.
Lower RPM is not automatically more even. Several 100 g tests at low RPM were reported as uneven, underdeveloped, or stalled, while other tests showed only small time changes. A very low RPM can reduce heat transfer enough that the profile no longer reaches the intended crack behavior 2 sources. For small batches, lowering RPM to improve BT readings can therefore compromise mixing and cup quality.
RPM, First Crack, and Sensor Readings
RPM affects both the actual roast and the way Roest reports the roast. Increasing RPM near first crack can create an RoR flick or apparent BT movement that may partly reflect more hot air around the probe rather than a true bean-temperature change source. The Roest team was also quoted as recommending lowering RPM to around 50 near first crack for better readings on 100 g roasts source.
For high-RPM or large-batch roasts, automatic first-crack detection can be less dependable because bean impacts and machine noise can be mistaken for cracks. Manual marking is often used when RPM, batch size, or bean noise makes the microphone unreliable; see First Crack Management for marking strategy.
Hardware and RPM Granularity
Displayed RPM does not always equal the same real agitation across machines. Multiple contributors reported machine-to-machine differences, including one case where one operator’s 55 RPM visually matched another’s lower setting, and another report that RPM may not be consistent across machines or voltage source.
older support guidance described RPM as operating in broad intervals, including 20–28, 29–33, 34–39, 40–49, and 50–65, with no visible difference between some adjacent settings. Later Ultra/L200 discussions describe newer motors and finer RPM control, including 1 RPM increments on Ultra and better granularity on the 200 series 2 sources. Treat RPM numbers as machine-specific unless the same hardware generation and motor behavior are confirmed.
This is especially important when sharing profiles. A profile that works at “55 RPM” on one unit may need 45, 50, or 55 on another to reproduce the same bean movement. For profile transfer, see Profile Sharing and Starting Points and Calibration and Environment.
200 g RPM Guidance Conflict
200 g guidance appears in two practical forms. One side cites max 45 RPM or 40–45 RPM for 200 g, while another recommends setting RPM below 40 to minimize beans going into the exhaust 2 sources. In practice, both can be reasonable depending on bean size, hardware generation, airflow, and whether the operator is seeing beans in the chute.
The safest operational resolution is to begin near the lower end of the accepted range, inspect the exhaust and chaff collector, and adjust upward only if mixing, chaff removal, or cup quality require it. If beans appear in the chute or exhaust, lower RPM first; if the roast then becomes too slow or flat, compensate with inlet or power rather than immediately returning to high RPM.
Troubleshooting by Symptom
| Symptom | Likely RPM-related cause | Practical fix |
|---|---|---|
| Beans in chute, exhaust, or chaff collector | RPM and/or exhaust airflow is lofting expanded beans too high, especially at 180–200 g. | Lower RPM one effective step; for 200 g, move toward 35–40 before reducing fan. curated Treat whole beans in the exhaust or chaff collector as a fault condition, not just a profile variation: clear the beans and chaff, confirm the exhaust/chaff path is unobstructed, and avoid repeated roasts while beans are accumulating there. |
| Roast slows, cracks weakly, or fails to crack after lowering RPM | Lower agitation reduced heat transfer. | Raise inlet or power, or return closer to 50–55 RPM. Do not assume the same profile works unchanged. |
| BT/RoR becomes noisy or flicks after RPM changes | Probe is reading more air, or the machine reacts to agitation changes. | Avoid large mid-roast RPM jumps unless deliberately profiling them. For 100 g, consider lowering near FC only for readings. |
| Small batch looks uneven at low RPM | Beans are not being mixed or lofted enough. | Increase RPM toward 55, or increase batch size to improve probe coverage and mixing. |
| High-RPM roast tastes flat, burned inside, or overly developed | Excessive hot-air exposure and fast internal development. | Reduce RPM, reduce late energy, or shorten development; be especially careful with naturals and low-density coffees. |
| Stuck beans or motor reversals | Bean size/shape catches in the propeller/door/exhaust geometry, compromising agitation. | Lower RPM by about 5 on machines with fine control, inspect the sliding-door lip if recurring, and review Maintenance and Cleaning. curated If repeated motor reversals or stalls do not clear promptly during a roast, abort/cool the roast and inspect the machine after it is safe rather than continuing the profile. |
Relationship to Other Controls
RPM should be tuned alongside inlet, power, airflow, and pressure rather than in isolation. Higher RPM can allow lower inlet temperatures for similar heat transfer, while lower RPM may require higher inlet or power to preserve timing. For inlet-specific tuning, use Inlet Temperature Management; for power-shape decisions, use Power Curve Strategies; and for interpreting BT behavior, use Bean Temperature Profiling.
When changing RPM, evaluate the result by cup, color, weight loss, and defect inspection rather than by graph shape alone. A smoother or more believable BT trace can come from probe contact changes, not necessarily from a better roast.