Transmitter Span, Turndown, and 4-20 mA Accuracy

The scaling itself is a straight line -- value = LRV + (mA - 4)/16 x span -- and the honest difficulty is entirely in the accuracy statement. READ THE BASIS: "0.1% of span" and "0.1% of URL" are the same statement ONLY at a turndown of 1:1, and a specification that does not state a basis has not stated an accuracy at all. Two transmitters with identical published figures can differ by an order of magnitude in the application, because one is referenced to the device's sensor limit and the other to the range it is actually calibrated to. TURNDOWN IS WHERE THE DIFFERENCE BECOMES CONCRETE. The absolute error on the URL basis is fixed by the sensor element, so as a percentage of the calibrated span it grows in DIRECT PROPORTION to turndown: a device ranged to a tenth of its limit performs exactly to a 0.1% specification and delivers 1.0% of its span. The span never got worse; it got smaller underneath a constant error. Selecting a wide-range device and turning it down for flexibility is a reasonable engineering choice that costs accuracy in a way the datasheet does not make obvious, and this computes both bases side by side so the trade is visible before the order is placed. THE BOTTOM OF THE RANGE IS WORSE AGAIN, AND IT IS WHERE CONTROL LIVES. A fixed absolute error is a large percentage of a small reading, so a loop controlling at twenty percent of span works with a proportionally much poorer measurement than the same loop at full scale -- and low-flow, low-pressure and low-load are exactly the conditions a modern control sequence spends most of its time in. The transmitter is accurate; the measurement is not, and the difference is entirely in where the range was set. This is a scaling and specification-reading aid, not a calibration. The actual error is established by calibration against a traceable standard at the range in use, and the published figure is a REFERENCE accuracy under stated conditions. It excludes several effects specified separately and frequently larger -- ambient temperature effect, static pressure effect on a differential-pressure device, long-term drift, mounting position, vibration -- and a total probable error must combine them. It does not address the sensing element's own installation error, which commonly dominates everything the instrument contributes. It does not cover square-root extraction for differential-pressure flow, which is a steeper problem at the bottom of its range, smart transmitters performing internal characterisation, digital protocols carrying the value alongside the analogue signal, NAMUR NE43 fault bands, or loop powering, burden and wiring resistance. The transmitter manufacturer's specification with its stated accuracy basis, a calibration against a traceable standard, and the controls engineer govern.

value = LRV + (mA - 4) / 16 x span, where span = URV - LRV; turndown = URL / span; the same quoted accuracy is applied on three bases -- percent of calibrated span, percent of upper range limit, and percent of reading.

The linear 4-20 mA scaling relation and the three accuracy bases a datasheet may use. The accuracy figure and its basis are ENTERED from the transmitter's specification.

A linear map and three percentage bases.

Estimate. AHJ and licensed professional govern.

Field names used by the API: lower_range_value, upper_range_value, upper_range_limit, loop_ma, accuracy_pct, low_reading_value, alt_upper_range_value, value_eng, span_eng, turndown, err_span_eng, err_url_eng, err_url_as_pct_of_span, basis_ratio, err_url_at_low_reading_pct, alt_span_eng, alt_turndown, alt_err_url_as_pct_of_span

• Linear transfer function no square-root extraction or internal characterisationthe transmitter's configuration
• Accuracy is a reference figure operating-condition effects are specified separatelythe manufacturer's full specification
• Installation error is not included it commonly dominates everything the instrument contributesa field verification