Intermediate Equipment Checks

Procedure for verifying equipment performance between calibration cycles using pairwise En comparisons, per ISO 17025 clause 6.4.10.

About this task

This procedure defines how the laboratory verifies that measurement equipment continues to perform within specification between scheduled calibration cycles. It fulfils the requirement of ISO 17025 clause 6.4.10: "When intermediate checks are necessary to maintain confidence in the performance of the equipment, these checks shall be carried out according to a procedure."

This procedure covers three internal cross-check types plus an annual interlaboratory comparison:

  • Molbox pressure transducer cross-check (4 transducers across 2 molboxes)
  • Molbox resistance channel cross-check (4 channels across 2 molboxes)
  • Electrical instrument cross-check (ADT220 ↔ BK5491A)
  • Molbloc flow — yearly ILC with a supplier (commitment only; ILC protocol agreed bilaterally)

Schedule:

  • Quarterly: all internal cross-checks (pressure, resistance, electrical)
  • Event-triggered: after return from service or transport, suspected damage, environmental excursion, or any event that may affect equipment performance
  • Yearly: molbloc flow interlaboratory comparison

Pairwise En Evaluation

All cross-checks use the pairwise comparison method per EA-4/02. For n instruments measuring the same quantity, every unique pair is evaluated independently. This avoids cross-contamination (a biased instrument cannot influence the evaluation of others) and provides richer diagnostics than leave-one-out methods.

The pairwise En number is calculated as:

En = xi xj Ui2 + Uj2

where xi and xj are the measurement results from instruments i and j, and Ui and Uj are their expanded uncertainties (k=2).

Important:
Acceptance criterion: |En| ≤ 1 for all pairs.

The number of unique pairs grows combinatorially with the number of instruments:

Instruments Pairs Application
2 1 Electrical cross-check (ADT220 ↔ BK5491A)
4 6 Pressure transducers / resistance channels

The intermediate check results complement the calibration drift analysis documented in the drift analysis doctrine. Together with the validity assurance techniques described under clause 7.7, they form the laboratory's equipment monitoring strategy.

For molbloc flow standards, the laboratory participates in an annual interlaboratory comparison with a supplier. Since the laboratory has only one set of molblocs, internal cross-checks are not possible for flow measurements. The ILC protocol is agreed bilaterally with the partner laboratory and follows the pairwise En methodology. ILC results are recorded separately.

Procedure

  1. Gather the required equipment and materials for the checks to be performed.
    • For pressure cross-check: 5-bar pressure controller, tubing to connect both molboxes to a common pressure source
    • For resistance cross-check: resistance standards at 100 Ω, 107 Ω, 110 Ω, and 115 Ω (stability better than ±0.02 Ω/year; calibration not required)
    • For electrical cross-check: test leads for parallel voltage connection and series current connection
Molbox pressure transducer cross-check — 4 transducers, 6 pairwise comparisons. All transducers read the same applied pressure simultaneously.
  1. Connect both molboxes to the 5-bar pressure controller so that all four reference pressure transducers (upstream and downstream on each molbox) are exposed to the same applied pressure.
  2. Apply a stable pressure at each defined setpoint and record the reading from all four transducers.
    1. Set the pressure controller to the first setpoint and wait for stabilization.
    2. Record the pressure reading from each of the four transducers (molbox 1 upstream, molbox 1 downstream, molbox 2 upstream, molbox 2 downstream) along with the expanded uncertainty from the current calibration certificate.
    3. Repeat for each remaining setpoint.
    Note:
    Pressure setpoints to be defined based on the transducer calibrated range. Record readings and uncertainties in the intermediate check log.
  3. Vent the system, disconnect the pressure controller, and restore normal molbox connections.
Molbox resistance channel cross-check — 4 channels, 6 pairwise comparisons per resistance value. Each resistance standard is measured sequentially by all four channels. The standards do not need to be calibrated; this check verifies mutual consistency between channels, not absolute accuracy.
  1. Connect the first resistance standard (100 Ω) to each molbox channel in turn and record the reported temperature.
    1. Connect the 100 Ω standard to molbox 1, channel 1 using the molbloc connector pinout. Record the temperature displayed on the molbox.
    2. Move the standard to molbox 1, channel 2. Record the temperature.
    3. Move the standard to molbox 2, channel 1. Record the temperature.
    4. Move the standard to molbox 2, channel 2. Record the temperature.

    The four temperature readings for the same physical resistor should agree. Differences reflect channel-to-channel variation in the resistance measurement function.

  2. Repeat the previous step for each remaining resistance standard (107 Ω, 110 Ω, 115 Ω).
    You now have 4 sets of 4 temperature readings — one set per resistance value, each with readings from all 4 channels.
Electrical instrument cross-check — bilateral comparison between the ADT220 and BK5491A. With only two instruments, it is impossible to isolate which one has drifted if En > 1.
  1. Configure the ADT220 as a source and connect it simultaneously to both the ADT220's own measurement input and the BK5491A.
    1. For voltage checks: connect the ADT220 voltage output in parallel to both the ADT220 measurement input and the BK5491A voltage input.
    2. For current checks: connect the ADT220 current output in series through both the ADT220 measurement input and the BK5491A current input.
  2. Apply each defined setpoint and record the readings from both instruments.
    1. Set the ADT220 to the first voltage setpoint. Record the voltage reading from both the ADT220 and the BK5491A, along with the expanded uncertainty for each instrument at that range.
    2. Repeat for all voltage setpoints.
    3. Reconfigure for current measurement. Set the ADT220 to the first current setpoint. Record the current reading from both instruments with their uncertainties.
    4. Repeat for all current setpoints.
    Note:
    Setpoints should span the overlapping measurement ranges. Voltage: within 0.01–10 V (ADT220 source range overlapping BK5491A measurement range). Current: within 4–29 mA (ADT220 range overlapping BK5491A range).
Evaluate all intermediate check results.
  1. For each check type, calculate the pairwise En number for every unique pair of instruments at each setpoint.

    Present results as a symmetric matrix with |En| values and pass (≤ 1) / fail (> 1) indication for each pair.

  2. Interpret the En matrices using the diagnostic patterns in the result section and determine whether corrective action is required.

Results

The intermediate check passes when all pairwise En values ≤ 1 across all check types and all setpoints.

Use the following diagnostic patterns to interpret En matrices:

Pattern Interpretation Action
All En ≤ 1 Full mutual compatibility No action required
One instrument fails against all others That instrument is suspect Recalibrate suspect equipment; review impacted results per clause 7.10
Two compatible groups Systematic disagreement External reference needed; potentially recalibrate all; review impacted results
Single isolated failure Ambiguous, borderline Repeat the check; monitor at next quarter
Scattered failures Systemic or protocol issue Review the check procedure; potentially recalibrate all
Important:
For the electrical cross-check (ADT220 ↔ BK5491A), only one pair exists. If En > 1, it is impossible to determine which instrument has drifted. Treat both as suspect — send both for recalibration and review all calibration results issued with either instrument since the last passing intermediate check.

What to do next

When an intermediate check triggers recalibration, handle the suspect equipment and review of impacted results according to the nonconforming work procedure (clause 7.10). If recalibration confirms drift beyond specification, notify affected customers.

Record all intermediate check results, En matrices, and any corrective actions taken. These records provide evidence of ongoing equipment monitoring per clause 7.7.