Commercial Coffee Drum Roaster Monthly Chaff Collector, Thermocouple & Drum Inspection Log

ContextAfter weighing and disposing of loose chaff, use a stiff brush to dislodge compacted fines from the bowl walls, lower cone, and any baffles. Oily chaff from medium-dark or dark roast profiles polymerizes into a resinous layer over time; if you encounter a sticky residue that the dry brush cannot dislodge, wipe the affected area with a cloth lightly dampened with isopropyl alcohol, then follow immediately with a dry wipe to prevent moisture retention. Never use compressed air inside the collector while any dry chaff remains—it aerosolizes combustible fines into the workspace and can create a flash-fire risk if an ignition source is nearby.

ContextShine a flashlight inside the empty, cool bowl and examine every weld seam, particularly around the inlet port and the lower cone junction where thermal cycling stress concentrates. Hairline cracks at weld seams start small but propagate quickly under repeated heating. Bluish or rainbow discoloration on stainless steel indicates the metal has reached 400–500°C—hotter than the intended design temperature—which points to airflow restriction causing heat to back up into the collector rather than exhausting normally. On mild steel units, dark rust-orange pitting near any weld joint is a structural concern; surface oxidation is cosmetic, but pitting that can be felt with a fingernail requires evaluation by a welder or equipment service technician within 30 days.

ContextPress the gasket firmly along its entire perimeter with your thumb. A serviceable gasket springs back noticeably when you release pressure and shows no surface cracking or brittle sections. A gasket that stays visibly compressed (retains the thumbprint shape) has taken a permanent compression set and is no longer forming an airtight seal. This allows both chaff fines and hot combustion gases to escape at the joint during roasting. Replacement gaskets for most commercial collectors run $15–$60 depending on the machine; note the condition in your log as Good, Soft (plan replacement), or Failed (replace before next roast). A gasket rated Failed should be treated as a safety item, not a cosmetic one.

ContextManually move the discharge valve or butterfly damper from fully closed to fully open and back. The motion should be smooth and consistent throughout, with the blade returning cleanly to the closed position without sticking at any point. Stiffness is almost always caused by chaff resin buildup on the pivot shaft; clear the pivot holes with a wooden skewer and brush the shaft clean. Inspect the damper blade edge-on: it should lie flat when closed. A warped blade that no longer forms a flush seal will allow reverse airflow through the collector during certain phases of the roast cycle, drawing smoke into the workspace and degrading exhaust efficiency. A visibly warped damper blade requires replacement—straightening thin sheet metal typically does not hold under thermal cycling.

ContextIf your collector design includes a screen or mesh insert to catch fines before the blower impeller, remove it and hold it up to a light source. Any visible hole or tear larger than the mesh opening allows chaff fines to reach the blower and, in gas-fired machines, potentially the burner zone. A screen that has distorted out of flat (from heat or over-forceful cleaning) no longer seats properly and should be replaced. Partial clogging—visible as opaque patches against the light—increases static pressure across the collector and measurably reduces airflow through the drum; this manifests in roast logs as a rising ET (environmental temperature) at the same heat setting over successive batches.

ContextDisconnect or open the duct section between the drum rear and the collector inlet—most commercial machines provide an access panel or quick-release clamp at this junction. Shine a flashlight into the duct and look for oily resinous buildup on the interior walls. Two to three millimeters of sticky coating after a month of heavy dark-roast production is typical and should be brushed clean; a partial blockage (more than 30–40% of the cross-section obstructed) restricts drum airflow enough to push environmental temperature readings 5–15°C above historical baseline at the same heat setting, producing inconsistent profiles before any alarm or safety device responds. Rate the buildup: None, Light (< 2mm), Moderate (2–5mm, clean now), or Heavy (> 5mm, investigate frequency of dark roast runs).

ContextMany drum roasters route both the roasting drum and the cooling tray through a shared exhaust blower and chaff collection system. Inspect the cooling tray's duct connection point for accumulated chaff, particularly if your production includes natural-process or honey-process coffees, which shed significantly more chaff during cooling agitation than washed coffees. A blockage at or just downstream of the cooling tray junction can cause chaff to back up into the tray itself, where contact with warm or hot beans presents a smoldering risk. Use your flashlight to check at least 30 cm downstream of the tray outlet.

ContextWith the roaster powered on and running at idle (no batch loaded, drum rotating, blower active), hold a strip of lightweight paper—approximately 3 cm × 15 cm—near the chaff collector discharge outlet with the lid removed. The strip should deflect steadily and noticeably toward the collector inlet, indicating the blower is drawing air through the collection path. Sluggish or absent deflection at idle suggests a blower impeller packed with compacted chaff fines (a separate service item) or an unsealed duct joint that is bypassing the collector. This test is a quick sanity check, not a calibrated measurement, but it catches gross airflow failures that instrument readings alone often miss until they become severe.

ContextUnscrew or release the BT probe from its drum port and examine the tip under bright light. A light tan or gray film is normal oxidation and has minimal effect on accuracy. A thick black carbon coating acts as thermal insulation and causes the probe to read 3–8°C below actual bean temperature—silently compressing your measured roast development time without changing anything on screen. Clean the tip gently with fine 400-grit sandpaper or a stainless steel wire brush, then wipe with a dry cloth. Inspect the sheath for bends, dents, or punctures; a physically damaged sheath on an ungrounded thermocouple junction allows EMI noise to enter the millivolt signal, producing the irregular 2–5 second temperature spikes that experienced roasters describe as a 'jumpy curve.'

ContextThe BT probe must be inserted precisely to the depth specified in your roaster's service manual—typically 40–80 mm past the faceplate interior wall for machines in the 15–30 kg range, though this varies significantly by manufacturer and drum diameter. At the correct depth the tip samples the moving bean mass; at even 10 mm too shallow it measures the air layer near the faceplate, which can run 20–30°C above true bean temperature during the onset of first crack. Mark the correct insertion depth on the probe sheath with a ring of high-temperature paint pen or a small crimp mark so you can confirm correct seating visually every time you reinstall the probe, without measuring from scratch.

ContextSet your multimeter to resistance (Ω) mode and disconnect the thermocouple leads from the controller terminal block—note the polarity markings before disconnecting. A Type K thermocouple in good condition will read in the range of 0–300 Ω at room temperature depending on wire length and gauge. An open-circuit reading (OL or infinite resistance on the display) means the internal junction has failed; this probe must be replaced before the next roast session, as the controller will display either a fault code or a frozen/erratic temperature that provides no usable feedback. A reading that has climbed more than 100 Ω above a previously documented baseline indicates progressive internal wire degradation from sustained high-temperature exposure; plan replacement within 30 days.

ContextPlace your calibrated reference probe and the roaster's BT probe side by side at ambient room temperature; they should agree within ±1.5°C. Then perform a brief test fire—no beans loaded—and compare readings at 100°C and again at approximately 200°C as the machine warms. A consistent offset (e.g., the roaster BT always reads 4°C below the reference) can be corrected in software on most roaster control platforms and must be logged as a software compensation value. A growing offset compared to prior months' log entries signals accelerating probe drift. In practice, Type K probes used in daily commercial production typically require replacement every 12–18 months; probes that reach a 5°C cumulative drift from original baseline are approaching the end of useful service.

ContextWalk the entire cable run from where it exits the drum housing to where it terminates at the controller terminal block. Look for any section where the outer insulation jacket shows discoloration, cracking, brittleness, or abrasion against the machine frame. Thermocouple extension wire uses specific alloy conductors matched to the thermocouple type—yellow jacket for Type K—and must never be spliced with ordinary copper wire. A copper splice creates a parasitic thermocouple junction at each connection point that introduces a temperature-dependent offset, shifting readings by an amount that changes with ambient workshop temperature and is extremely difficult to diagnose. If you find any copper splice in the thermocouple circuit, that entire cable segment must be replaced with matched-alloy extension wire before the machine returns to production.

ContextThe ET probe typically mounts through the rear panel or side wall of the drum housing, pointing into the roasting zone. Remove it and clean the tip exactly as you did the BT probe—carbon buildup affects both equally. Reinstall it and verify the compression fitting, lock nut, or clamp is tight enough that you cannot shift the probe depth by hand. A loose ET probe that gradually migrates deeper or shallower between roasts is one of the harder sources of profile inconsistency to diagnose, because the shift is gradual and the reading changes without triggering any alarm. The symptom is usually a batch-to-batch variation in ET at a fixed heat setting that appears random but actually correlates with whether the machine was bumped or moved recently.

ContextAt each terminal block or plug-and-socket connector along the thermocouple circuit, look for green or white oxidation on the terminal faces. Thermocouple circuits operate in the millivolt range; even partial oxidation on a terminal adds resistance that the controller interprets as a temperature change. Clean visibly oxidized terminals with a pencil eraser or an electronics contact cleaner spray, then allow to dry fully before reconnecting. Tighten all terminal screws until snug, using a flat-head driver—loose screws are the single most common cause of intermittent 'spiky' roast curves where the BT or ET reading jumps erratically for 2–5 seconds, then settles. Do not overtighten: cracked terminal block bodies are a frequent result on older machines with plastic housings.

ContextIf your machine includes a post-drum exhaust thermocouple, run a brief test fire without beans and note the exhaust temperature at a consistent heat setting (e.g., full heat at the end of a 5-minute warm-up). Compare this reading to the same measurement from your previous three monthly logs. An exhaust temperature more than 15°C above the historical average at the same heat setting suggests restricted airflow through the drum—check chaff duct blockages and drum end-seal gaps first. An exhaust temperature more than 15°C below the historical average more commonly indicates a thermocouple fault in the exhaust probe itself or an unexpected air leak (a failed seal pulling ambient air into the exhaust path and diluting the hot gas stream).

ContextHold a 300 mm metal ruler flat against the drum barrel surface at multiple positions around its circumference and along its length. Any gap between the ruler and the drum surface greater than 1–2 mm over that span is meaningful distortion from thermal stress. Localized patches of extreme darkening or scaling surrounded by normal gray metal indicate the drum experienced a localized overheat event—possibly from a batch held at full heat past the intended end-point, a cooling failure, or an airflow blockage that concentrated heat in one zone. A measurably warped drum causes uneven bean agitation (you may notice audible irregular rattling during rotation), inconsistent roast evenness across the batch, and in severe cases, contact with the drum housing that produces metal scoring.

ContextRotate the drum slowly by hand one full revolution while shining a flashlight through the drum opening, pausing every quarter-turn to examine each flight (fin) against the beam. Flights should be straight, evenly spaced, and firmly attached at both the leading and trailing ends. A bent or twisted flight concentrates beans in one zone during rotation, causing uneven first crack—the cup result is a batch with a wide spread of development levels. Scrape carbon buildup from flight surfaces with a wooden or hard-plastic scraper; avoid metal tools, which can score the drum wall and create crevices that accumulate roast oils. Examine every weld joint where flights attach to the drum wall; a visible crack in a weld joint is a safety item requiring qualified weld repair before the machine returns to production.

ContextInsert feeler gauge blades into the gap between the rotating drum face and the stationary faceplate at the 12 o'clock, 3 o'clock, 6 o'clock, and 9 o'clock positions. Record all four measurements in your log. The manufacturer specification for this gap is typically 1.0–2.5 mm on most commercial machines; consult your service manual for the exact tolerance. A gap tighter than minimum spec risks the drum contacting the faceplate as it expands thermally during roasting—you will hear this as a periodic metallic scraping in the first 10–15 minutes of a roast. A gap wider than maximum spec allows combustion gases and chaff to escape forward around the drum, producing visible smoke at the front of the machine and delivering chaff into the operator area—a dual fire and air-quality hazard.

ContextClose the front door or faceplate fully and attempt to open it without releasing the latch; it must resist firmly with no give. Inspect the latch mechanism for visible wear on the catch surface, bent components, or weakened spring tension—the spring should return the latch to the locked position with a definitive snap, not a sluggish return. On machines using a bolt-and-flange closure rather than a spring latch, check each bolt for proper torque and replace any that turn freely without significant resistance. A front door that partially opens during a roast allows beans to spill into the burner area and creates an immediate fire condition; this is one of the few faults that warrants stopping a roast in progress and dumping the batch before it reaches the cooling tray.

ContextWith all guards reinstalled except any minimum access needed for observation, turn the drum through one complete revolution slowly and deliberately. Smooth, consistent rotational resistance with no binding is normal. A tight spot that recurs at the same angular position every half-revolution indicates a bearing fault or a drum that has developed an ovality—an egg-shaped cross-section from repeated thermal expansion and contraction without proper support. A gritty, grinding sensation under the hand indicates metal-on-metal contact, most commonly from a bearing that needs lubrication or has reached the end of its service life. Even faint anomalies should be documented with a description in your log; bearing deterioration is gradual, and month-to-month comparison of your written notes is the most reliable early-warning system available without specialized instrumentation.

ContextOn belt-driven roasters, press the center of the drive belt's longest unsupported span with moderate thumb pressure. A correctly tensioned belt deflects approximately 10–15 mm; more than 20–25 mm deflection indicates a belt that is too loose and will slip under load, producing inconsistent drum RPM that manifests as profile variability before any mechanical warning sign appears. Inspect both the flat (inner) and ribbed or outer faces of the belt for cracking, glazing, fraying at the edges, or embedded chaff debris. A glazed belt surface—shiny and slick rather than the normal matte texture—means the belt has been slipping and heating; it will not regain grip and must be replaced. Belt replacement on most commercial roasters takes 30–60 minutes and typically costs $20–$80 in parts. On chain-driven machines, check chain sag with a straightedge and look for stiff links, corrosion, or elongation by comparing the chain pitch against a new reference link.

ContextExamine each tooth on the drive sprocket or gear for the 'shark-fin' wear pattern—one flank of each tooth worn to a sharp point while the other remains relatively flat. This occurs because load is applied in one direction only; it is normal and progressive. When more than 25–30% of teeth show this profile, the sprocket should be replaced; a tooth that shears under load during a roast locks the drum immediately. Sight along the alignment between the motor pulley and the drum pulley: they should lie in the same plane with no visible lateral offset. Even 3 mm of misalignment places a continuous side-load on the motor shaft bearing and causes the belt or chain to track toward one edge, accelerating both the belt wear and the bearing deterioration simultaneously.

ContextConfirm first whether your roaster has grease-serviceable bearings (Zerk fittings) or sealed-for-life units—the service manual will specify. For grease-serviceable bearings, use only a high-temperature grease rated to a minimum of 200°C; suitable options include Mobilgrease XHP 222 or equivalent NLGI Grade 2 high-temperature formulations. Standard workshop grease softens and migrates at temperatures above 120–150°C, seeping out of the bearing housing and into the drum area where it attracts chaff and can smear onto the belt drive. Two pump strokes of a hand grease gun per fitting is sufficient for most commercial roaster bearings; more than three pumps risks forcing grease past the bearing seal and into the drum interior. Record the grease product name, volume, and date in your monthly log so that any future bearing failure can be evaluated against the lubrication history.

ContextAll numerical findings from this inspection—feeler gauge readings at four positions, BT and ET probe offsets against the reference thermometer, chaff weight, belt deflection measurement, bearing grease quantity, and any resistance readings—must be entered in the log before the machine is powered back on for production. A verbal note or mental record is not a maintenance log. The signed log provides three simultaneous functions: it constitutes an auditable maintenance trail for insurance and regulatory review, it builds a multi-month trend dataset that enables predictive maintenance decisions, and it serves as handoff documentation so that a different technician performing next month's inspection understands what was found and done this month. Use consistent units throughout—millimeters, not inches; degrees Celsius, not Fahrenheit—to keep trend analysis reliable across entries.

ContextPull the log sheets from the last three monthly inspections and lay them alongside the current sheet. For each probe—BT, ET, and exhaust—note whether the measured offset from the reference thermometer has been stable, gradually worsening, or accelerating. A probe whose offset is increasing monotonically (each month measurably worse than the previous) is in active degradation. A probe that has drifted more than 5°C cumulative from its original documented baseline should be recalibrated by a thermocouple service provider or replaced within 30 days even if it has not yet triggered a controller alarm, because 5°C of BT error at first crack corresponds to approximately 15–20 seconds of development time error—enough to shift a roast from the intended profile to a noticeably different cup outcome without any change appearing on the roast curve display.

ContextAt the bottom of the monthly log, list every finding that is outside spec or trending in the wrong direction. For each item, record: the specific finding with the measured value, the severity classification (Safety / Quality / Monitor), the required resolution, and the name of the person responsible. Safety items—cracked welds, failed gaskets, open-circuit probes, drum-contact clearances—must be resolved before the machine returns to production; document the serial number of any part ordered and the expected delivery date. Quality items affect roast consistency and should be resolved within 30 days; they allow production to continue but should be flagged to the head roaster so profile adjustments can be made if needed. Monitor items are within spec but trending; they require no action this month but should be re-evaluated next cycle.

ContextThe inspecting technician must sign and date the completed log personally. A countersignature from the facility manager or head roaster is good practice for operations carrying commercial property or manufacturing liability insurance—many policies require demonstrated oversight of equipment maintenance, not just the inspection itself. File completed logs in chronological order and retain for a minimum of 24 months; some insurers and local fire authority inspectors request records going back 12–18 months at audit. Paper logs should be stored in a fire-resistant folder or scanned to a backed-up digital location—a folder destroyed in the same fire it was meant to document provides no protection. Digital logs are acceptable provided they are stored on a backed-up platform and not solely on the roastery floor's local desktop.