Skip to content

Measurement and Precision in UHPC

The working rule of high-precision coulometry: to measure the rate of a degradation mechanism reliably, the instrument's accuracy and precision should be roughly one order of magnitude better than that rate. The weaker the signal, the tighter the requirement.

Source vs. sense

Every cycler both sources current or voltage to the cell and senses what it records. The analysis uses the sensed values, not the intended ones. So while the source side sets the true conditions the cell experiences, it is the sense side — its accuracy and stability — that decides how much you can trust the data.

Four terms people mix up

  • Accuracy — how close a reading is to the true value; sets consistency between channels.
  • Precision — repeatability; sets the scatter between cycles.
  • Resolution — the smallest change that can be distinguished; should be well below accuracy and precision.
  • Noise — random fluctuation; it caps the useful resolution.

For a signal as small as coulombic efficiency, high precision with poor accuracy gives a clean curve that is offset; high accuracy with poor precision gives the right average but jumps every cycle. You need both.

Full scale vs. reading: the accuracy caveat

Current accuracy is normally quoted as a percent of full-scale range (FSR), not of the reading. On a 2 A range, "0.01% FSR" means a fixed tolerance of ±200 µA at any output — because fixed offsets and noise do not shrink as the set current shrinks. Dividing an absolute error by a small reading inflates an apparent error rate that the hardware never promised. Good instruments switch ranges (auto-ranging) so small currents are measured against a small full scale, tightening the absolute guardrail.

Temperature and cutoff

Current-sensing elements have a temperature coefficient, so uncontrolled temperature drift contaminates the measurement; the chamber must hold temperature steady. Cutoff conditions must be consistent, because a small change in temperature or condition shifts where a cycle really ends and moves CE with it.

Next