Checkweigher Rejecting Good Product: Calibration Guide

Checkweigher Rejecting Good Product: Calibration Guide

When your CWS-001 in-line checkweigher starts kicking good pouches, trays, and bags into the reject bin, you are not just losing product — you are losing trust in your line. I have spent the last twelve years crawling under checkweigher frames at snack, frozen food, and powdered-mix plants, and I can tell you that in roughly 85% of “false reject” callouts, the load cell is fine. The real culprits are vibration creeping in through the frame, misadjusted air supply on pneumatic rejecters, mechanical centering that has drifted, or environmental drift that nobody has re-zeroed in weeks. This guide walks you through a field-engineer-grade diagnostic and calibration procedure for the CWS-001, with the exact numbers, torque specs, and thresholds I use on the floor. Plan on 60 to 90 minutes start to finish if you go in order.

Symptoms Checklist: Confirm You Actually Have a False Reject Problem

Before you touch a screw, confirm the symptom is really false rejection and not a fill-weight problem upstream. Walk the line and answer the following. If you cannot tick at least five of these boxes, your issue is likely upstream of the checkweigher — go look at the filler or multihead scale first.

• Reject bin is filling with product that, when re-weighed on a calibrated bench scale, reads within ±2 g of the labeled target weight.
• Operator reports reject rate has climbed from a baseline of 0.3% to above 1.5% over the past 24 to 72 hours.
• The CWS-001 HMI “Live Weight” reading jitters more than ±0.5 g with no product on the belt.
• Rejects cluster at the start or end of a product batch, not evenly distributed.
• Machine has been moved, cleaned with high-pressure water, or had a belt replaced in the past two weeks.
• Compressed-air pressure at the checkweigher gauge reads below 5.5 bar or fluctuates more than ±0.3 bar.
• Frame is bolted to a mezzanine or shared structure with a pouch former, sealer, or carton erector.

Root Cause #1: Frame Vibration Coupling Into the Load Cell

This is the single most common cause of checkweigher false reject on the CWS-001, and it is the most misdiagnosed. The dynamic weighing belt sits on a flexure-beam load cell rated at 6 kg with a resolution of 0.1 g. That cell is so sensitive it will pick up the 12 to 18 Hz resonance of a nearby impulse sealer through a shared frame. The HMI averages it out visually, but the reject decision is made on a 25 ms window — the noise wins.

Diagnosis: Put the CWS-001 into Service Mode, scroll to Test > Load Cell > Live Graph, and clear the belt. With the belt running and no product, watch the graph for 60 seconds. A clean signal should sit flat within ±0.3 g. If you see a sine wave, a spike pattern, or any peak-to-peak amplitude above 0.8 g, you have a vibration coupling problem. Next, lay a dial indicator with a magnetic base on the frame cross-member where the load cell mounts. Anything above 0.02 mm of movement with the line running is too much.

Fix: Isolate the frame. The CWS-001 ships with four SGC-25 rubber-steel vibration pads; if yours are older than three years or compressed more than 1 mm, replace them. Torque the frame leveling feet to 18 Nm on each, and verify the frame is level within 0.2 mm per meter using a precision machinist’s level. Disconnect any conduit, air line, or cable that is hard-routed between the checkweigher and adjacent equipment — route them with a 200 mm service loop and cable ties instead. If the source is structural (a floor that bounces when the sealer fires), you will need a 6 mm steel sub-plate under the four pads, sized at least 400 × 600 mm. Re-check the Live Graph; you are looking for peak-to-peak below 0.4 g.

Root Cause #2: Pneumatic Supply and Rejecter Timing

The CWS-001 rejecter is a double-acting cylinder running on a 5/2 solenoid valve. It needs 6.0 to 6.5 bar at the machine inlet, measured at the FRL unit, with the line under full demand. If your air is weak, the cylinder stroke slows, the reject arm does not clear the belt in time, and good product gets clipped or swept. Worse, a slow return stroke leaves the rejecter partially extended into the lane and the next good pack trips it.

Diagnosis: At the FRL gauge, with the line running, read pressure. Below 5.5 bar, you have a supply problem. Above but fluctuating more than ±0.3 bar, you have a regulator or dryer problem. Go to Service > Rejecter > Timing and run the “Stroke Test” ten times. The CWS-001 should log extend time of 80 to 110 ms and return time of 90 to 120 ms. Anything over 150 ms on either stroke is your false-reject culprit.

Fix: First, drain the FRL bowl — water in the cylinder ruins seals and slows the stroke. Replace the inline filter element if it is discolored. Set the regulator to 6.2 bar with the line at full demand. If stroke time is still high, remove the cylinder and inspect the piston seal; on the CWS-001 OEM cylinder (Bimba EF-25-50), the seal kit is good for roughly 8 million cycles, or about 14 months of single-shift operation. In the Rejecter Timing menu, set Pre-fire Delay to the value shown on your belt-speed decal — for a 35 m/min belt it is typically 12 ms, for 45 m/min it is 8 ms. A 2 ms error here is enough to push good product into the reject chute.

Root Cause #3: Load Cell Drift and Zero Offset

The CWS-001 uses a 6 kg strain-gauge load cell with a factory-trimmed zero offset. Cells drift. Thermal expansion of the flexure, belt tension changes, and accumulation of product dust on the weigh table all shift the zero point. A drift of just 1.2 g on a 250 g target is enough to push a perfectly good pouch past the lower reject threshold.

Diagnosis: From the HMI, go to Calibration > Zero Check. The display will show current zero offset. Anything outside ±0.4 g requires a re-zero; anything outside ±1.5 g requires investigation before re-zeroing (you are chasing a leak or contamination, not a calibration). Next, place the 500 g calibration weight (PN CWS-CAL-500) dead-center on the weigh table. Span reading should be 500.0 ±0.5 g. If it is off by more than 1.0 g, span calibration is needed.

Fix: Run Zero Calibration with the belt running and warm (let it idle 10 minutes first to thermally stabilize). The procedure is Calibration > Zero > Auto, then wait 30 seconds. Span Calibration is Calibration > Span > Place Weight > Auto. Use only the OEM calibration weight — a generic 500 g lab mass will be within tolerance but the OEM weight is certified to OIML R111 Class M1, which is what your audit trail expects. After span, re-verify zero. The two interact; iterate twice. Document the calibration date in the HMI; the CWS-001 stores the last 32 events in Calibration > Log.

Root Cause #4: Product Centering and Transfer Geometry

On a dynamic checkweigher, the product has to enter the weigh table centered, flat, and stable. The CWS-001 weigh table is 420 mm wide and 350 mm long, with a 90 mm infeed transition and a 90 mm discharge transition. If product touches the side guides, rides a transition plate, or bounces on entry, the load cell reads the impact as part of the weight. A pouch that lands 15 mm off-center can read 3 to 5 g heavy due to lateral force coupling.

Diagnosis: Run ten sample packs through at production speed and watch them. Look for: (a) pouch touching either side guide, (b) pouch pitching (front edge up or down) on entry, (c) pouch bouncing vertically on landing, (d) pouch riding the discharge transition. Any of these will cause checkweigher false reject even with perfect calibration. Measure side-guide gap; it should be product width + 8 to 12 mm per side. Measure infeed timing belt to weigh-belt height differential; it should be 0.0 to +0.3 mm (infeed never lower than weigh belt).

Fix: Loosen the side-guide hex bolts (8 mm), set the gap to product width + 10 mm per side, and re-torque to 4 Nm. Check the infeed transition plate — if you can slip a 0.5 mm feeler gauge under a pouch as it transfers, the plate is too low; shim it with the CWS-SHIM-05 kit. Belt tension should deflect 4 to 6 mm under 10 N of finger pressure at mid-span. Tighter belts couple motor vibration into the cell; looser belts let product slip and skew. If product is pitching on entry, your upstream timing belt is running 5 to 8% faster than the checkweigher belt — synchronize them to within ±1.5%.

Root Cause #5: Belt Condition and Drive Geometry

The weigh belt itself is part of the weighing system. A belt that has stretched, glazed, or picked up contamination will false-reject. The CWS-001 uses a 420 × 40 mm PU belt with a blue fabric top layer — the color is so you can see product dust. Belt tension pulls directly on the load cell through the drive pulley; a stretched belt adds preload that drifts the zero.

Diagnosis: Stop the belt and run a finger along the surface. If it feels slick or you see a glossy film, the belt is glazed. Weigh the belt on a bench scale — a new CWS-BELT-420 weighs 92 ±2 g; if yours is over 100 g, it has absorbed product moisture or oil and is no longer balanced. Check tracking — the belt should run centered on the pulley within ±1 mm over 60 seconds at full speed.

Fix: Replace belts that are over weight, glazed, or have visible cracks. After replacement, set tension to the 4 to 6 mm spec above, then run the belt for 15 minutes at full speed before zero calibration — new belts stretch 1 to 2% in the first hour, and if you calibrate before that stretch, your zero will be off within 30 minutes of production. Tracking is adjusted via the idler pulley set-screws (3 mm hex, 1/4 turn at a time, give it 30 seconds to settle between adjustments).

Root Cause #6: Environmental Factors and Air Currents

The CWS-001 can resolve 0.1 g. That is less than the force of a ceiling fan on a flat pouch. I have seen checkweigher false reject rates spike 4% because someone installed a fan-coil unit blowing across the line. HVAC supply air within 2 m of the checkweigher will ruin your day.

Diagnosis: Take an anemometer (a cheap Kestrel 1000 is fine) and measure air velocity at the weigh table height, 200 mm above the belt, with the line running. Anything above 0.5 m/s will cause false reject. Also check ambient temperature swing — if the plant goes from 18°C at dawn to 32°C by noon, the load cell flexure thermally expands and your zero drifts 0.4 to 0.8 g.

Fix: Install the CWS-DRAFT-001 acrylic draft shield kit — it mounts to the four threaded M6 holes on top of the frame uprights. For thermal drift, schedule an automatic zero re-calibration every two hours via HMI > Schedule > Auto-Zero. The CWS-001 will pause line infeed for 25 seconds, run the zero routine, and resume — this single change took one frozen-potato line I worked on from 3.2% false reject to 0.4%.

Calibration and Component Specification Table

Frequently Asked Questions

How often should I calibrate the CWS-001?

Daily zero check at the start of every shift (30 seconds, no tools), span calibration weekly with the 500 g calibration weight, and full traceable calibration quarterly. If your plant temperature swings more than 8°C in a shift, enable auto-zero every two hours — the traceable quarterly interval does not change.

My reject rate jumped overnight with no settings change. What is the first thing to check?

Walk to the FRL unit and look at the gauge. Compressed-air systems fail silently overnight when the dryer regenerates or a hose elsewhere on the line cracks. Below 5.5 bar, the rejecter cylinder slows and starts clipping good product. Fix the air first; nine times out of ten, this resolves the overnight spike.

Can I run the CWS-001 without the OEM calibration weight?

You can, but you should not. A generic 500 g lab mass is often within tolerance, but the CWS-CAL-500 is certified to OIML R111 Class M1 with a traceable certificate, which is what your HACCP and customer audit trail require. If your weight is missing, order a replacement before the next production run — do not substitute.

The load cell graph shows a steady sine wave even when nothing is touching the belt. Is the cell bad?

Almost certainly not. A clean sine wave is frame resonance coupling in from another machine. The cell only fails open (no reading at all) or with a permanent offset that re-zero will not hold. Before replacing a $1,200 load cell, isolate the frame and check the Live Graph again.

Rejects only happen on one product, not the others. Why?

That product has a centering or transfer geometry problem — usually the side-guide gap was set for an old SKU and the new one is narrower or wider. Re-measure and reset the side-guide gap to product width + 10 mm per side. Also check that the product is not bouncing on landing; lightweight pouches (under 80 g) sometimes need the infeed belt speed dropped 5 to 8%.

Can I clean the CWS-001 with high-pressure water?

No. The CWS-001 is IP65 rated against splash, not pressure wash. Direct high-pressure water will ingress through the load cell cable boot and the rejecter solenoid vent, and will absolutely cause checkweigher false reject within one shift. Wipe down with a damp cloth and food-grade sanitizer only. If you have been pressure-washing it, plan to replace the load cell boot and re-grease the cylinder.

Need Help With Your CWS-001?

If you have walked through every section above and your false reject rate is still above 1%, you likely have a combination issue — vibration plus a worn belt, or thermal drift plus a slow cylinder. Esper Foodtech ships OEM parts and offers a remote diagnostic service that reads your CWS-001 HMI logs over a secure connection; in most cases we can pinpoint the cause within an hour. Email [email protected] with your machine serial number, current production rate, and a screenshot of the Load Cell Live Graph. We will get you back to spec.