NiMH battery charging and monitoring board

A board replacing centralised power supplies inside luminaires. Four-switch synchronous buck-boost charging, four hardware variants, one PCB architecture.

Released Gessler GmbH 2025–2026 8 min read Confidential

Role

Charger design and layout

Ownership

Led the power stage and magnetics sizing

Scope

Buck-boost stage · input EMI and ESD · controller selection study

Constraints

0402 maximum package where space allowed · ten-year service life

Standards

DIN EN 61347-2-7 · IEC 61000-4-2 · IEC 61000-4-4

Variants

Four · two charge controllers · two cell types

Representative model. Not the client's design.

What it is

The board replaces centralised power supplies inside luminaires. A three-cell NiMH pack is charged through a four-switch synchronous buck-boost stage giving a programmable 3–24 V output, with 1-Wire UART monitoring back to the host controller.

Four hardware variants cover two charge controllers and two cell types across a single PCB architecture. I designed the buck-boost power stage, the input EMI and ESD network, and the magnetics sizing for both cell variants, and produced the charge controller selection study that set the choice for all four.

Decisions

Kind

Design decision
Problem found
Verification

Problem found U3 · input filter

A common mode choke that would not have survived

Found: The part specified for the input filter was rated 300 mA with 1.6 Ω DCR, against a load of up to 1.5 A. It would have dropped roughly 2.4 V and run hot.
Did: Sized the replacement against the real load — a 2 A part at 0.19 Ω.
Result: Rail drop about 1.2 % of the 24 V input.

Design decision U3 · interface

Protecting a single-wire interface

Found: The 1-Wire UART shares a connector with the 24 V supply and its return passes through the input choke, so switching transients were going to move the local ground reference against the host and corrupt edges at 115200 baud.
Did: Added a series damping resistor of 33 Ω to control reflections on the cable, and a low-capacitance ESD diode — chosen for low capacitance specifically so the edge rate is preserved rather than rounded off.
Result: Clean 115200 baud signalling through the input choke.

Verification

Checked

Inrush current
Fast role-swap transition

Tool

LTspice

Building the model the vendor did not supply

No vendor SPICE model existed for the charge controller, so I built my own subcircuit and a test bench substituting a capacitor for the battery, to study inrush and the fast role-swap transition before committing to hardware. Building the model was also the fastest way to understand what the part actually does.

This is the kind of thing that passes schematic review and fails in the field.

What is not shown here, and why

The schematic, the layout and the measured data are Gessler GmbH intellectual property. The architecture, the reasoning and the decisions are mine to publish, and they are above.

What I cannot do is hand over the designs themselves, and I would not want to work somewhere that wanted me to.

Other boards