Calibration sequence selection in DKD-R 6-1 is driven by two variables: the gauge's specified accuracy expressed as a percentage of full scale (% FS), and whether the device uses a Bourdon tube sensing element, which forces a fixed 5-minute waiting time at each measurement point [S1][S2].
The guideline is published by the Physikalisch-Technische Bundesanstalt (PTB) for the Deutscher Kalibrierdienst (DKD) and is cited in EURAMET CMC declarations such as EURAMET-M-AL-00000MLU-1, approved 14 August 2021 for the Albanian DPM in the 1.00E5 Pa to 7.00E7 Pa range with oil medium [S5].
Point count split at the 0.6% FS accuracy threshold
DKD-R 6-1 distributes calibration points into 5 nominal set points (0%, 25%, 50%, 75%, 100% of range) for gauges whose accuracy is greater than 0.6% of full scale, and into 9 set points (0%, 10%, 30%, 40%, 50%, 60%, 70%, 90%, 100%) for gauges whose accuracy is 0.6% FS or better [S3]. The five-point routine is the cost-efficient default used for industrial process gauges, while the nine-point routine is required for test-gauge-class instruments and reference standards because the higher density reveals non-linearity and hysteresis that the five-point scheme can mask. For compound gauges covering vacuum and positive pressure, DKD-R 6-1 requires at least two points in the negative range, for example -1 bar and -0.5 bar, with the remaining points distributed in the positive range [S3].
The 0.6% FS split is not arbitrary: it tracks where hysteresis, linearity, and repeatability errors tend to fall below the application tolerance band, so a five-point sweep with computed deviation is statistically defensible above the threshold and under-samples below it. Operators applying pressure gauge calibration in the field should map the manufacturer's spec (typically referenced to the measuring span, not the indicated value) against Table 1 of the guideline to pick the correct sequence before the pump is connected.
Calibration sequences A, M1, M2, M3: which one to run
Items with a measurement range greater than 2500 bar in principle require sequence A, because at that pressure clamping and seal effects dominate the error budget and the simplified up-only sweep is inadequate [S1]. For lower ranges, DKD-R 6-1 defines sequence M1 (a single up-cycle with the 5- or 9-point set) and the M2/M3/M4 pair that perform up-down-up-down cycling to expose hysteresis.
The M2 first down-cycle carries the same 9 measurement points and 2-minute dwell used on the up-cycle in M1, with M3 being the second up-cycle and M4 the second down-cycle [S3]. Hysteresis in DKD-R 6-1 is defined as the maximum difference between readings taken at the same nominal pressure on an up-cycle and a down-cycle, and the M-series sequences exist specifically to quantify it. The default test pressure is read against the standard, not the unit under calibration: set points are adjusted on the standard pressure gauge and the corresponding reading is then taken on the unit under calibration (UUC) [S3].
For Bourdon tube instruments, DKD-R 6-1 specifies a 5-minute waiting time at each measurement point, and this dwell can only be reduced for quasi-static applications [S1][S2]. The waiting time ensures the spring element and the pressure medium reach thermal and mechanical equilibrium; cutting it on a Bourdon gauge is the single most common reason a clean repeatability check fails the first run.
Reference standards, working standards, and the 1/3rd accuracy rule
DKD-R 6-1 requires the reference standard to be a long-time-stable pressure device such as a pressure balance or a liquid-column manometer, calibrated at PTB at regular intervals and traceable through the CIPM MRA [S4]. The reference certificate states expanded uncertainty at reference conditions of standard gravity, 20°C, and 1 bar ambient; if calibration is performed outside those conditions, corrections for influence quantities must be applied and the uncertainty contribution carried into the budget [S4].
For the working standard used at bench level, DKD-R 6-1 inherits the conventional 1:3 or better accuracy ratio, i.e. the working standard shall have an accuracy of at least one-third (and preferably one-fourth) of the unit under calibration [S3]. This is the same ratio mandated by ISO/IEC 17025 risk-based thinking, and it is the reason a 0.1% FS laboratory reference cannot economically back a 0.5% FS industrial gauge (the test uncertainty ratio collapses toward 1:5, which inflates the combined uncertainty). The two comparison methods allowed under DKD-R 6-1 (adjust to the UUC, or adjust to the standard) yield mathematically identical results only when both instruments are linear, and the standard-adjusted method is preferred for traceable reporting because the standard's reading is what carries the metrological chain.
CMC declarations citing DKD-R 6-1: what changed in 2021
On 14 August 2021 the BIPM KCDB approved CMC EURAMET-M-AL-00000MLU-1 for the Albanian DPM, covering digital pressure gauges from 1.00E5 Pa to 7.00E7 Pa in oil medium with absolute expanded uncertainty 2.6E3 Pa to 7.0E3 Pa, using the formula 2.6E+03 + 6.3E-05 x p and citing DKD-R 6-1 Edition 03/2014 as the basis [S5]. The CMC uncertainty statement uses the EURAMET Q[a,b] = sqrt(a² + b²) convention and reports at coverage factor k = 2 (95% level of confidence) unless otherwise stated [S5].
For calibration laboratories looking to register or extend a pressure CMC, the 2014 edition is the current baseline cited in the KCDB, with Revision 3 (DOI 10.7795/550.20210422) being the latest published text on the PTB OAR server as of 22 April 2021 [S1]. DAkkS, the legal successor to the DKD since 3 May 2011, supervises the accredited laboratories that now perform the routine work previously done under DKD accreditation [S1]. The commercial use of the guideline content is explicitly allowed in calibration laboratories even though the document is published under the CC by-nc-nd 3.0 license [S1].
Practical limits and when to stop calibrating
DKD-R 6-1 is a guideline, not a regulation, and explicitly permits deviations from its sequences in agreement with the accreditation body if they are justified by technical aspects [S2]. Two cases that should trigger an immediate stop rather than a calibration attempt: (1) a gauge whose leak rate on the test bench exceeds 1% FS per minute at any set point, because the reading will not stabilise within the 5-minute dwell, and (2) a gauge whose zero shift between pre- and post-calibration exceeds the manufacturer tolerance, which usually indicates mechanical damage rather than drift.
The guideline's main operational limits, beyond the 2500 bar threshold that forces sequence A, are: Bourdon tubes always carry 5-minute dwell, M-series sequences add 2-minute dwell per point on top of the stabilization check, and the comparison-method setup must include a torque-controlled thread seal with the PTFE tape kept out of the pressure path to avoid acting as an unintended flow restrictor [S1][S3]. Anomalies caught during a DKD-R 6-1 calibration should be cross-checked against multivariate drift baselines; readers running continuous sensor health programs can extend the same logic to a multivariate anomaly detection layer over the calibration data. The next verifiable signals to track are any new PTB revisions to DKD-R 6-1 published after Revision 3 (April 2021) and the next KCDB CMC approvals citing the 03/2014 edition.
For component-level specifications, see process calibration, and temperature calibration bath.