Balancing is the active counterpart to cell matching: it narrows the state-of-charge divergence that series mismatch creates (Paper 29) so that cells approach full charge together and the charger can terminate without sacrificing laggards or abusing leaders. NiMH balancing differs from lithium balancing in a crucial way - overcharge current is consumed by internal recombination rather than stored, giving the chemistry a limited self-equalising tendency that a designer can exploit but must not over-trust. This paper compares passive and active balancing and a third, NiMH-specific approach - charge-profile shaping - and explains how to size and time balancing so it works with, rather than against, the recombination chemistry.
In lithium packs balancing is essential because any overcharge is unsafe; in NiMH the oversized negative and recombination loop tolerate a bounded overcharge, and the leading cell, once full, effectively 'absorbs' continued current as heat while laggards catch up - a crude, passive self-balancing mechanism. This is why simple NiMH packs can series-charge with no balancer at all: controlled overcharge of the leaders is precisely how laggards get filled.
The cost is exactly the overcharge wear of Papers 21 to 23, localised in the leading cells; engineered balancing aims to fill laggards while minimising that enforced leader overcharge, trading added hardware for longer pack life and a cleaner termination.
Passive balancing switches a resistor across the highest cells to shunt some current (or to discharge them during rest), holding leaders near full while current continues to laggards; it is simple and cheap but dissipates energy as heat - awkward in a pack already managing charge heat - and is slow, because practical bleed currents are small relative to charge current, so it works best at the low currents of the end band or during rests rather than against a 1C bulk charge.
For NiMH, passive balancing is most effective as a top-of-charge and rest equaliser: as current steps down near full, modest bleed currents become comparable to charge current and can genuinely hold the leaders while laggards complete, aligning naturally with a multi-stage profile.
Active balancers move energy from higher cells to lower cells (capacitor, inductor or transformer topologies) with less waste heat, equalising faster and at higher effective currents; they suit large, high-value traction and mobility packs where pack life and energy efficiency justify cost. During charge they continuously take excess from leaders and deliver it to laggards, keeping the string tight so all cells peak together and pack -delta-V regains its clarity (Paper 29).
Their control must be reconciled with NiMH's flat OCV (Paper 16): because voltage barely reveals SOC through mid-charge, active balancing decisions are best based on end-of-charge behaviour, coulomb counts or relative peak timing rather than mid-charge voltage differences, which mostly reflect resistance and temperature mismatch instead.
A third, NiMH-specific approach equalises through the charge profile itself: pulse charging with rests lets leading cells relax and partially self-discharge through recombination while laggards continue to absorb on subsequent pulses (Paper 13); Reflex negative pulses briefly discharge the whole string, with leaders giving up relatively more (Paper 14); and terminating bulk charge early then finishing at low current gives recombination-based self-equalisation time to pull cells together at minimal stress. These methods reduce divergence with no per-cell power hardware.
Charge shaping cannot correct large mismatch and is best combined with tight incoming matching; it is, however, essentially free and is why a well-designed multi-stage/pulse profile both manages gas and partially balances.
Required balancing current follows from the expected divergence per cycle: estimate the ampere-hour spread accumulated between equalisation opportunities and size the bleed/transfer current and equalisation window to clear it, concentrating action at top of charge and during rest where NiMH equalises most efficiently. Control should use the hottest-cell and first-full logic of Papers 28 to 29, avoid balancing a faulty cell, and log residual divergence as a health metric.
The first figure compares how passive, active and charge-shaping approaches pull diverging SOCs together; the second sequences a combined strategy of tight matching, shaped charge and top-band active equalisation.
Specify the match bin, balancing topology and current, the SOC basis for decisions (end-of-charge behaviour rather than flat mid-charge OCV), and the interaction with termination and thermal control; validate on deliberately aged, mismatched strings to show leaders stay within overcharge and temperature limits while laggards reach full. Weijiang supplies matched sets and the cell behaviour data needed to size balancing, supporting both simple self-equalising consumer packs and actively balanced mobility packs. The series next leaves the cell and pack to survey how charging differs across the major application worlds.
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.
