Minimum Sample Weight: The Smallest Mass Your Balance Can Weigh to Spec — and How to Test It

Why readability is the wrong number

Readability (the scale interval, d) is the smallest change the display can show. Repeatability is the standard deviation of repeated weighings of the same load — the true resolving power of the balance on your bench. On a well-installed 0.1 mg balance the two are close; add an air-conditioning draught, a vibrating bench or a warm-up skipped, and the repeatability grows while the readability stays exactly where the datasheet put it. That is why two balances with the same readability can have minimum weights that differ by a factor of three, and why the same balance can have a different minimum weight on the second floor than in the basement.

A weighing error is roughly constant in absolute terms — a few tenths of a milligram on an analytical balance regardless of the load. Expressed as a percentage of the sample, that same error is negligible on 10 g and ruinous on 10 mg. Minimum weight is simply the sample size at which the constant absolute scatter becomes an acceptable relative one.

How USP ⟨41⟩ defines it

USP General Chapter ⟨41⟩ Balances applies wherever a pharmacopoeial procedure says a material is to be "accurately weighed", and most GMP and ISO 17025 laboratories adopt its logic whether or not they are pharmaceutical. It sets two tests. Repeatability: weigh one test weight not fewer than ten times, and the requirement is met when twice the standard deviation of those readings, divided by the smallest net weight you intend to use, is not more than 0.10%. Accuracy: the reading for a test weight of between 5% and 100% of capacity must be within 0.10% of the weight's certified value.

Rearranging the repeatability requirement gives the working rule: minimum weight = 2,000 × the repeatability standard deviation. One more provision matters at fine readabilities — if the measured standard deviation comes out below 0.41 d, USP tells you to use 0.41 d instead, because a display that only moves in steps of d cannot demonstrate a smaller scatter than that. The floor this creates is 820 d: 82 mg on a 0.1 mg balance, 8.2 mg on a 0.01 mg balance, 0.82 g on a 1 mg balance, 8.2 g on a 0.01 g balance. No amount of care takes a balance below its floor; only a finer instrument does.

These figures are the acceptance limit, not a comfort zone. Many quality systems apply a safety factor of two or three on top, so that a balance qualified at 140 mg is used for samples of 300–400 mg and above. USP does not mandate the factor; your own procedure sets it.

Testing it on your own bench

Let the balance warm up as its manual specifies, level it, close the draught shield, and choose a test weight near the smallest sample you weigh. Weigh it ten times, removing it from the pan and re-zeroing between readings so that each weighing is a genuine repeat and not a repeated read of the same settled value. Calculate the standard deviation of the ten results; compare it with 0.41 d and take the larger; multiply by 2,000. That is the minimum weight of that balance, on that bench, on that day. Record the readings, the weight used, the date and the room conditions — the record is what an auditor asks for.

Repeat the test after the balance is moved, serviced or recalibrated, and periodically as your quality system requires. When our calibration service calibrates a balance on your bench, ask for the ten-weighing repeatability test to be run and recorded on the certificate alongside the reference standards and the test points, so the minimum weight can be taken straight from the certificate. A&D Gulf is not an ISO/IEC 17025-accredited laboratory; where your quality system requires an accredited certificate, tell us and we will say so plainly.

What the A&D range gives you

A&D publishes a typical minimum weight for each class, calculated per USP ⟨41⟩ from measured repeatability: about 140 mg for its 0.1 mg analytical balances, about 1.4 g for the 1 mg Fortis models (1.8 g for the IP65 versions) and about 14 g at 0.01 g. The datasheet repeatability is a guaranteed maximum — 0.1 mg on the GH-120, GH-200, GR-120, GR-200, GX-124A and FZ-104 — which would give 200 mg; the lower published typical figure reflects what the instruments actually achieve.

Where your smallest sample sits below that, the semi-micro tier is the answer rather than better technique. The dual-range GH-202 (220 g / 51 g) and GH-252 (250 g / 101 g) read 0.01 mg in their fine range with a specified repeatability of 0.02–0.03 mg; the GR-202 (210 g / 42 g) is the same idea in the GR housing; the BA-225TE reads 0.01 mg across 220 g with 0.03 mg repeatability. At 2,000 × 0.02–0.03 mg those figures place the minimum weight in the 40–60 mg region under datasheet conditions. For filter weighing and trace work the BM-20 micro balance (22 g × 0.001 mg, repeatability 0.0025 mg at 1 g) brings it to a few milligrams.

Two features are worth asking for. The GX-A/GF-A Apollo balances carry a minimum-weight function that stores the qualified figure and flashes a Min-S alert until the sample on the pan reaches it — the rule enforced at the point of use rather than in a procedure nobody re-reads. And the GH, GR and GX-A families self-calibrate when the ambient temperature drifts, which is the disturbance most likely to widen repeatability in a Gulf laboratory between one test and the next.

When the sample is below the minimum

Three options, in order of cost. Weigh more: many methods allow a larger aliquot with a proportionally larger dilution, and a 300 mg sample on a 0.1 mg balance is a comfortable weighing that a 30 mg sample is not. Weigh by difference: tare the vessel, add the sample, weigh again — this removes the container from the net weight but does not change the repeatability floor, so it helps only when the container was the problem. Move up a class: a 0.01 mg semi-micro balance for samples in the tens of milligrams, a micro balance below that. Buying the finer class for a sample that clears the coarser one buys nothing but a slower, fussier instrument.