When a belt scale stops reading right, the first assumption is that the scale is failing. Usually it isn't. The load cells and the control box are doing exactly what they were built to do. Something around them changed.
That's because a belt scale isn't really one component. It's a small system built on top of your conveyor, and its accuracy depends on the conveyor as much as on the scale. Here's what matters, in the order it matters.
A single-idler belt scale weighs on one idler, but it measures across five: the scale idler, the two idlers before it and the two after. The belt drapes across all five, and the load cells read whatever share of the material weight that geometry gives them.
Change one of those four neighbors and you change the reading, with nothing wrong with the scale itself. The common culprits:
For a multi-idler scale, the field grows the same way: two idlers before, every scale idler, and two after.
Every idler in the scale field must be the same make, model, can size and troughing angle, and should be scale-duty idlers. On a new install, the strongest recommendation is five new matching idlers. Building a precision weighing system on top of worn, mixed idlers is a false economy.
Then align them:
The installation manual is blunt about this step: do not skip it. It's the most common reason a well-built scale reads poorly.
Location matters as much as alignment. The guidelines from the Weigh Shark 2 manual:
In practice, that means the scale needs a decent length of straight, unobstructed conveyor. If a short conveyor can't give it that, it's worth looking at a different conveyor in the same circuit.
A standard speed wheel rides the return belt. MCR's heavy-duty version, introduced in 2023, uses a 7.5-inch wheel that doubled bearing life over the earlier 4.5-inch design, and it retrofits onto older scales where there's clearance between the top and return belt. Mount it on an idler other than the scale idler.
For wet sand, corrosive or sticky material, a magnetic shaft encoder mounted on a pulley is the better choice. The head pulley is preferred because it stays cleaner than the tail. For belt speeds under 40 feet per minute, check alternative sensor options with the manufacturer.
Calibration happens in two steps:
Zero test. The scale learns what an empty belt weighs. Run it with the belt completely empty, at normal speed, for one full belt revolution. Do it at least weekly — daily is better. Belt wear, material buildup and temperature all shift what "empty" means.
Span calibration. This sets how the scale converts load cell readings to weight, using calibration weights or a known load. Run it two or three times; large swings from test to test point back to a mechanical problem in the scale field.
The most reliable verification is a material test: run a known amount of material across the scale, weigh it on a certified truck scale, and adjust. Calibration weight kits are available for troughing and wing scales when a material test isn't practical, but a material test is preferred.
Most belt scale complaints trace back to the scale field, the location or a skipped zero test. Get those right, and a good scale stays accurate for years.
If your scale has drifted, or you're planning an install and want it right the first time, Davis Industrial installs Weigh Shark® scales — including the matched idlers, alignment and calibration — and works directly with MCR on sizing.
Schedule a free site walk at conveyors247.com, or call 813-247-3620.