A series string charges all cells with the same current but cannot guarantee they reach full charge together, and in that gap lies most pack-level NiMH trouble. Cells differ slightly in capacity, self-discharge and resistance from manufacture and diverge further with age; during charge those differences mean some cells reach full and enter damaging recombination while others still need current, and the pack-level voltage that a simple charger watches becomes a blurred average that hides the abuse. This paper explains how mismatch develops through a charge, why the first-full cell effectively decides the string's fate, how mismatch erodes termination and life, and what matching, topology and monitoring can do about it - setting up the balancing paper that follows.
Even matched production cells carry small spreads in capacity and internal resistance; storage at different self-discharge rates and varied temperature history widen initial SOC differences, and in service non-uniform temperature and ageing amplify them - hotter cells age faster and lose capacity, which makes them reach full sooner, which makes them run hotter, a divergence loop. A pack that begins well matched does not stay matched.
Quantifying incoming spread (capacity binning, resistance sorting) bounds the initial divergence; controlling the thermal field (Paper 28) bounds the in-service divergence; neither can be eliminated, so the charge strategy must tolerate a residual spread.
Because series current is identical, the cell with the smallest usable capacity or highest starting SOC reaches 100 percent first; continued string current drives it into oxygen recombination and heating while lagging cells continue storing charge. The first-full cell's voltage peaks and begins to fall even as laggards' voltages still rise, so the summed pack voltage - the only signal a string-level charger sees - smears the individual peaks into a broad, weak feature (the effect introduced in Paper 6).
The pack cannot finish cleanly: stopping at the first cell's peak undercharges the laggards, leaving capacity unused and worsening perceived imbalance; continuing to fill the laggards overcharges the first-full cell, spending its life on every cycle. Mismatch converts a single decision into a trade-off.
Higher-resistance cells show larger voltage drops and generate more I-squared-R heat, and the first-full recombination adds heat on top; these become the pack's hot cells (Paper 28), and their local warmth further lowers their oxygen-evolution potential so they overcharge still sooner. Electrical mismatch and thermal mismatch thus reinforce each other, localising wear in a subset of cells that age ahead of the rest and eventually drag the whole pack's capacity down to their level.
Pack capacity is set by its weakest cell in both directions: the low-capacity cell empties first on discharge and fills first on charge, so reducing mismatch is equivalent to recovering usable pack energy and extending pack life.
A clean -delta-V requires all cells to peak together; with mismatch, individual peaks occur at different times and partly cancel in the sum, reducing pack-level dip magnitude and delaying it until the majority are overcharged. The greater the spread, the more the charger must lean on temperature and timers - and the more abuse the leading cells absorb before those trigger. This is a primary reason matched cells and, for larger packs, per-cell supervision matter more than the charger's termination algorithm.
Reflex negative pulses (Paper 14) can marginally relax divergence by letting leading cells back off, but they do not substitute for matching or active balancing and risk reversing lagging cells if over-used.
Incoming quality control - capacity grading and resistance sorting into tight bins - is the first and cheapest defence; topology is the second (independent charge channels for small consumer packs, per-cell supervision for larger strings); in-life detection logs per-cell voltage and temperature to flag divergence before it becomes failure. The first figure illustrates how cell SOCs fan out through a mismatched series charge; the second contrasts termination clarity for tight and loose matching.
The decision scales with pack value: a two-cell consumer pack relies on binning; a mobility or traction pack warrants per-cell monitoring and the active balancing examined next.
Specify capacity and resistance match bins for pack cells, maximum allowable divergence, the supervision topology, and a charge strategy that terminates on leading-cell criteria while limiting their overcharge; validate by deliberately mismatched strings to confirm worst-case leading-cell temperature and overcharge. Weijiang supplies graded, matched cell sets with documented spread and divergence guidance, giving pack builders a controlled starting point. The companion technique that actively narrows divergence during charge - cell balancing - is the subject of the next paper.
Weijiang Power designs and manufactures nickel-metal hydride cells, matched packs and charging-ready configurations for consumer, industrial, medical and mobility customers, and supports partners with charge-protocol guidance, IEC 61951-2 performance files, IEC 62133-1 safety evidence and charger co-validation. Share your cell format, charge rate, thermal envelope and cycle target and our engineers will specify a cell-and-charge combination that protects both runtime and service life. Review the range on the products page.
