Roest L200/Ultra Guide
The Roest L200/Ultra changes how familiar Roest profiles behave, especially when using counterflow. This guide explains the practical differences that matter when moving from older Roest workflows, adapting profiles, reading sensors, and deciding how to approach batch size, pressure, and drop decisions on the L200/Ultra.
What Changes on the L200/Ultra
The L200/Ultra should not be treated as a direct continuation of older Roest behavior. Counterflow changes the roaster enough that older profile assumptions often need to be rebuilt rather than copied point-for-point source. The Ultra also has different internal parts and probe locations than older models, so temperature readings and profile behavior can differ even when the roast direction appears familiar 2 sources.
Counterflow is the defining operational change. It tends to roast faster and transfer heat more efficiently into the beans, which is why equivalent profiles usually require lower inlet temperatures than normal flow. Several experienced users report needing roughly a 40–60°C reduction in inlet targets when converting from normal/L100-style profiles to counterflow; one L100 Ultra comparison achieved the same roast time and weight loss after reducing inlets by 43°C and using 20% less power 2 sources.
Counterflow profiles do not begin in counterflow immediately. The machine starts in the regular direction and then switches, because starting directly in counterflow can leave coffee stuck in the hopper 2 sources. Opening or using the viewport to inspect beans means the roast is no longer in counterflow mode source.
Canonical Starting Workflow for L200/Ultra Counterflow
Use this section as the default starting framework when adapting an existing profile to L200/Ultra counterflow. More detailed treatment of individual variables belongs in Inlet Temperature Management, Batch Size Scaling, Pressure Management, Bean Temperature Profiling, and Development Time and Drop Decisions.
| Decision | Practical starting point | Notes |
|---|---|---|
| Convert a normal-flow or L100-style profile | Reduce inlet targets by about 40–60°C | 40–50°C is specifically suggested for 50g counterflow; larger batches may still fall in the 40–60°C range depending on the source profile and batch size 2 sources. |
| If the converted roast runs too fast | Lower inlets further by about 10°C and retest | One reported first counterflow attempt at -40°C still ran too fast, after which the inlet targets were dropped another 10°C source. |
| Avoid over-cooling the start | Do not make the start excessively cold just to force a fast roast | A cold start was not found beneficial for fast roasting in this context source. |
| Initial batch range for serious counterflow testing | 180–200g is a common practical range; 50–100g is more sampling-oriented | A shared profile was described as 180–190g, and Tom Roest shared a 200g counterflow profile 2 sources. |
| Small-batch reading | 100g reads well in standard mode; above roughly 80g can be stable and accurate, with 50g noted in counterflow | This is an Ultra-specific observation about BT stability rather than a general rule for all models source. |
| Drum speed | Start near 55–60 rpm if no other constraint is present | This is a reported working range, not an official universal setting source. |
| If beans stick or jam in counterflow | Try reducing drum speed before changing the entire profile | Reducing drum speed was reported to solve a counterflow sticking issue without changing flavor source. |
| Drop decision | Prefer manual drop based on BT, color, and cup feedback rather than relying only on first crack timing | First crack timing and development time can vary because crack detection and audible crack count are not always consistent source. |
| Optional short development windows | For one Ultra profile, naturals were suggested at 25–35 seconds and washed coffees at 35–45 seconds | Treat these as profile-specific starting points, not universal drop rules source. |
A practical first test is therefore: select a known 180–200g profile or adapt an existing normal-flow/L100 profile; reduce inlet targets by roughly 40–60°C; keep the profile shape only as a first approximation; roast manually; then adjust from cup results, BT readings, color, and weight loss. The profile shape may need more than a uniform temperature offset, because some users found their existing inlet-profile shapes did not map cleanly to counterflow even after large temperature reductions source.
Reading BT, Reverse BT, and First Crack
The L200/Ultra uses different BT sensing behavior across orientations. Community discussion describes two BT sensors: one for counterflow and one for normal flow, because one sensor cannot read BT correctly across both spinning orientations source. This is why profile comparisons should not assume that a BT value from normal flow, counterflow, and older Roest models represents the same physical measurement.
In counterflow, the reverse BT reading may be more bean-representative than the standard BT reading. One L200 user reported BT reverse readings around 200°C at first crack, about 9–10°C lower than the main roast readings, and interpreted the reverse probe as being better positioned on the beans in that mode source. Another 185g counterflow reference reported first crack around 209–210°C, reinforcing that first-crack temperature depends on the sensor and mode being referenced source.
Because first crack can be inconsistent across coffees, operators, and detection methods, development time should be handled cautiously. Some beans crack quietly or show low crack counts, and different users may mark first crack at different moments; for repeatable control, BT-based manual dropping is often preferred over automatic drop conditions tied to first crack source. For a deeper treatment, see First Crack Management and Development Time and Drop Decisions.
Pressure, Airflow, and Fan Interpretation
On the Ultra, the pressure sensor is described as being in the exhaust, which makes pressure logging useful regardless of roast direction source. Pressure logging itself has been reported as intuitive, but pressure should not be treated as the same thing as airflow 2 sources.
Tom Roest’s framing is that negative exhaust pressure is primarily about ensuring smoke exits through the exhaust, while heat transfer should be thought about in terms of airflow rather than pressure alone source. Counterflow or tilting can make pressure more negative because the inlet is more blocked, even while actual airflow is lower source.
The integrated pressure meter can still be used practically: adjust exhaust fan to reach the desired pressure environment, while judging results by roast behavior and cup quality 2 sources. For detailed fan and pressure setup, defer to Airflow and Fan Settings and Pressure Management.
Batch Size, Counterflow Capacity, and Sticking
The Ultra’s useful batch range is broader than older 100g-centered expectations. Community discussion describes upgraded Roest capability from low 20–50g batches up to 200g and sometimes beyond source. On Ultra counterflow, 180g+ Ethiopian batches have been reported without issues, and some users report that 220–230g can work with some coffees 2 sources.
200g+ counterflow capacity is coffee- and machine-behavior-dependent. Some experienced users report 200g counterflow as workable, and Sorin states that Ultra counterflow can easily go up to 230g; however, another reported 200g as not working well in counterflow and nearly blocking the chamber/drum blades 2 sources. The practical conclusion is to treat 200g as a realistic working size but not a guaranteed safe maximum for every coffee, especially where small beans, high charge mass, or drum-direction transitions increase the chance of blockage.
The machine is designed to detect a blockage and reverse the drum to try to release it, but repeated jamming should still be avoided for both machine care and coffee quality source. curated If a counterflow batch sticks or jams, do not try to clear beans with fingers or tools while the roaster is running or hot; stop the roast and follow the manufacturer’s clearing/cooling procedure before intervening. Counterflow dumps may also be chaffier than expected, so discharge behavior should be watched during early profile work source. curated During early counterflow tests, keep the chaff path, exhaust, and collection areas clean and monitor discharge closely; unusual chaff buildup or smoke backup should be treated as a reason to stop and inspect before continuing. Broader batch-size planning belongs in Batch Size Scaling, and drum-speed-specific adjustments belong in Drum Speed RPM Settings.
Power and Heat Application
Ultra counterflow applies heat quickly and powerfully. A flatter power application has been suggested when the roast is too aggressive, especially in L200/Ultra counterflow contexts source. One comparison from the same location and exhaust setup reported 100g Ultra counterflow roasting in just over 4 minutes at 48% peak power, while an L100+ needed 75–80% power for a 5:30–6:00 roast at the same batch size source.
Lower inlet temperatures and less stress on the heating element have been associated by experienced users with cleaner flavor, longer shelf life, and better results, but those benefits depend on still carrying enough heat through development source. If the roast loses momentum late, the first adjustment should be made within the profile framework rather than simply restoring old L100-style heat levels. See Power Curve Strategies, Rate of Rise Management, and Heat Transfer Fundamentals.
Profile Migration and Software Notes
Creating counterflow profiles from logs has received explicit software attention. Version 1.14, dated 13 February 2026, was reported to improve support for creating profiles from logs for L200 Ultra counterflow mode and P3000 source. The web portal settings page was also described as the place to find the latest updates and descriptions source.
The Ultra UI has more complexity than older workflows, but the new touch screen and WiFi chip have been reported positively in use 2 sources. For shared profiles, use Profile Sharing and Starting Points, but treat any imported profile as a starting structure rather than a finished recipe; the L200/Ultra’s counterflow behavior, probe locations, and batch-size sensitivity all require taste-based verification.