Why Do Two Scales Show Different Readings?
Need a scale that stays consistent? Shop the A&D EW-1500i NTEP Class III compact balance on sale for $297.00 (regular $560.00), in stock — or call us to match two instruments that disagree.
Why do two scales show different readings? Usually one (or both) of the instruments is off-level, uncalibrated, still warming up, sitting in a draft or temperature swing, or simply reading to a coarser division than the other. Two honest scales can also disagree by a small amount that still falls inside each scale’s allowed tolerance. Below is a practical checklist you can cite, with a worked example and the metrology ideas behind it.
Worked example: the same 100 g weight on two scales
Suppose you place a 100 g reference weight on Scale A (readability 0.01 g) and Scale B (readability 0.1 g):
| Scale | Display | What it can mean |
|---|---|---|
| A (0.01 g) | 100.02 g | Within a tight lab tolerance after recent calibration — or a small bias from level, temperature or corner loading |
| B (0.1 g) | 100.1 g | Rounded to the nearest 0.1 g; may still be “in tolerance” for a Class III bench scale even if it does not match A’s hundredths |
Before you declare either scale “wrong,” treat the reference weight as the referee only if that weight’s class and uncertainty support the finer scale. A worn shop weight is not a lab standard.
Seven reasons two scales disagree
| Cause | What happens | Quick check |
|---|---|---|
| Not level | Off-level platforms introduce a cosine error; the vertical component of gravity on the load cell changes | Center the spirit-level bubble; weigh again |
| Calibration / span | One scale’s span drifted since the last calibration | Apply the same known test weight to both; see how to calibrate a scale |
| Warm-up | Electronics and mechanics settle after power-on; early readings wander | Follow the maker’s warm-up (often 15–30+ minutes) |
| Temperature | Load-cell sensitivity and air density change with temperature; Handbook 44 evaluations use a stated temperature range | Weigh in the approved range; avoid HVAC blasts |
| Drafts / air currents | Moving air pushes on the pan like a tiny extra force — worse at fine readability | Close windows; use a draft shield on analytical balances |
| Readability vs accuracy | More display digits are not the same as smaller error | Compare tolerances, not just the last digit |
| Different gravity / site | Local g varies slightly by latitude and altitude; some indicators store gravity compensation | Calibrate at the use site after a move |
Tolerance, not “perfect match”
NIST Handbook 44 sets acceptance and maintenance tolerances by accuracy class and the verification scale division e (often equal to the display division d on simple instruments). OIML R 76 uses the same class idea (I, II, III, IIII) with maximum permissible errors that grow with load. Two Class III scales can both be in tolerance and still differ by a fraction of e. “They don’t match to the last digit” is not automatically a failure.
For selling by weight in the U.S., the scale generally needs an NTEP Certificate of Conformance and place-in-service under state rules — see what NTEP / legal for trade means.
How to compare two scales fairly
- Level both; same solid bench; no draft.
- Warm both up per their manuals.
- Zero both empty (and tare the same container if you use one).
- Use one known test weight appropriate for the finer scale.
- Record both readings, the weight’s nominal value and class, room temperature and time.
- If they still disagree beyond either tolerance, calibrate or service the outlier — don’t “average” them for trade use.
When to replace or upgrade
If one scale fails repeatability, fails a corner test, or cannot hold calibration, repair or replace it. For counter and light industrial work, an NTEP Class III compact like the A&D EW-1500i (300/600/1500 g) is a clear, trade-capable choice at $297.00 on sale. For finer jewelry work, compare a Class II model such as the A&D EK-1200i (1200 g × 0.1 g, on sale from about $205.99 depending on listing).
Related explainers
- How temperature, level and drafts affect scale accuracy
- Why do two scales show different readings?
- Scale readability vs accuracy
Frequently asked questions
Why do two scales show different weights for the same item?
Most often: level, calibration, warm-up, temperature, drafts, or different readability. Both scales can also be within tolerance and still differ slightly.
Which scale should I trust?
Trust the one recently calibrated with a suitable test weight at the use site — or a reference weight whose uncertainty supports the decision. For trade, trust the sealed legal-for-trade scale after place-in-service.
Can calibration fix the difference?
If the span or zero is wrong, yes. If the cause is drafts, level or a damaged load cell, calibrate after you fix the environment or hardware. MNM Scales calibrates complete scales from $95 in Houston.