← Back to: Battery Charging / Gauging / Protection / BMS
Scope & Anti-Overlap for Contactor & Pre-Charge Driver
This page is limited to the driver, protection, and sequencing for the high-voltage / medium-voltage contactor on the charging side, plus the pre-charge path right in front of it. It explains why the BMS must pre-charge the bus before closing the main contactor, so we do not get arcing, contact wear or large EMI events.
It belongs to the Charging domain execution layer, not to vehicle-level PDU design, not to Insulation / Leakage Monitor (HV), not to Pack-FET / eFuse, and not to multi-cell balancing. Those topics are handled in their own pages to avoid SEO overlap and to keep intent clean.
Many small-batch BMS projects are later “simplified” by purchasing: an original intelligent coil driver with surge clamp and pull-in/hold profile is replaced by a generic low-side switch. The real purpose of this page is to anchor BOM remarks so that substitutions without surge suppression, without coil feedback or without sequencing are rejected.
Only the “pre-charge voltage reached” check is mentioned here. How to measure insulation, how to localize leakage and how to run HV IR tests is out of scope and must be read from the dedicated HV leakage / insulation page. Likewise, CC/CV behavior, JEITA thermal control and charging-power decisions belong to the charger / gauging pages.
System Scenarios and Typical Topologies
In the charging domain, the contactor & pre-charge driver normally sits between the charger/DC source and the pack/battery bus. Depending on where the main isolation is implemented, we commonly see three patterns: charger-side contactor, pack-side contactor, and a dual-contactor arrangement with the pre-charge stage in the middle.
For each pattern, the page must tell: who starts the pre-charge (MCU, charger controller, vehicle host), whether the driver is high-side or low-side, and where the status/fault is reported back to (BMS AFE or charger control). The logic stays the same even if the hardware is a smart relay driver IC or a small BMS daughterboard using MOSFETs and TVS.
Charging-side contactors must be closed more gently than traction/discharge contactors because chargers and DC/DC modules often expose large output capacitors. Direct closure causes inrush, arcing and EMI. That is why this driver page belongs to the charging branch and not to the traction/PDU pages.
Do not use this page to describe vehicle-level coordination with dual inverters, dual mains or PDU master logic. Keep it to “what is needed to safely close a charging path”.
Coil Drive Architectures for Charging-Side Contactors
Charging-side contactors often see high inrush and large output capacitors from chargers/OBCs. Because of that, the driver cannot be a generic low-side MOSFET with a diode. It must drive, absorb surge, and report coil status. This chapter separates acceptable drivers from “prohibited substitutes”.
We describe four practical driver types: (1) low-side MOSFET + diode snubber (lowest cost, slow release), (2) smart high-side drivers (ST VND/VNQ, onsemi NCV84xx, TI TPS27xxx-Q1), (3) pull-in/hold contactor drivers (e.g. TI DRV110/DRV120) and (4) drivers for dual-winding contactors with programmable current profiles. Only drivers that detect coil fault, clamp the surge and reduce holding current can enter this page’s “qualified list”.
Drivers that only turn on/off and do not provide surge absorption or diagnostics must be flagged in the BOM as “not approved for charging-domain contactors”.
BOM remark (copy/paste): “Do not replace automotive high-side or pull-in/hold contactor drivers with generic low-side MOSFET switches that lack surge suppression and coil diagnostics.”
Pre-Charge Path Engineering in the BMS Domain
A safe charging-side closure needs at least three elements: a pre-charge resistor, a controlled switch (small contactor or MOSFET), and a voltage sense feeding back to the contactor & pre-charge driver IC. Only when the bus has risen to about 0.8–0.9 × target do we allow the main contactor to close.
The pre-charge resistor is selected from the DC link capacitance, target voltage and allowed pre-charge time. The rule of thumb is to limit inrush below the source limit while still reaching the ready voltage within the charging window. Once purchasing changes the resistor (to reduce cost or reuse stock), the whole pre-charge timing changes and the main contactor may close too early.
Even if the charger or OBC has a soft-start of its own, the BMS should still perform a pre-charge on its side, because line length, harness inductance, or charger firmware changes can all introduce extra inrush. This page documents the BMS-side pre-charge so it does not get dropped in small-batch builds.
The pre-charge switch can be a relay/contact, or an FET+driver. FET versions must be combined with surge/clamp devices; otherwise when the main contactor finally closes, cable and inductor energy will still create a spike.
Small-Batch Procurement & Cross-Brand Alternatives (7 Vendors)
In small-lot BMS charging projects, this is one of the most frequently downgraded nodes. The original design uses an automotive smart driver with surge absorption and diagnostics. Purchasing later replaces it with: (1) a simple relay/low-side switch, (2) a non-automotive high-side, (3) a driver without TVS/active clamp. This chapter locks those three “bad replacements” in the BOM so the charging-domain sequence is not broken.
Cross-brand alternatives are allowed only if all of these stay the same or better: 1) energy/surge absorption, 2) diagnostics (coil open/short, pre-charge done/timeout), 3) AEC-Q grade, 4) timing/sequence support, 5) package/thermal capability.
Smart driver → relay switch
Original: current limit + clamp + diagnostics.
Downgraded: on/off only → not acceptable.
Automotive HS → consumer HS
Automotive has surge & temp envelope. Consumer part often misses it.
With clamp → no clamp
If original used TVS/active clamp, replacement must match or exceed surge energy.
BOM remark 1: Coil driver shall maintain pull-in/hold current profile. Simpler on/off drivers are not acceptable.
BOM remark 2: Pre-charge and main contactor sequencing must be preserved when cross-brand devices are used.
BOM remark 3: Select automotive-qualified high-side/relay drivers from TI / ST / NXP / Renesas / onsemi / Microchip / Melexis only.
TI
DRV110, DRV120 → pull-in/hold for contactors/valves
TPS27S100-Q1, TPS1H100-Q1, TPS1H200-Q1 → smart HS with diagnostics
TPS2H160-Q1 → 2-ch HS, good for “pre-charge + main”
ST
VND5T035AK, VN7020AJ, VNQ5E160K-E, VNQ5050K-E → automotive smart HS
VNLD5300TR-E → small pre-charge path
NXP
MC10XS3412 / MC10XS3435 → protected HS w/ SPI diag
MC33926 / MC33928 → actuator / high-current
MC33888 → multi HS/LB with diag
Renesas
RAJ2800044H12HPF → intelligent power device
ISL78434 / ISL78420 → HV HS / half-bridge for FET pre-charge
Select IPD with current feedback enabled
onsemi
NCV8401A, NCV8410, NCV8461 → classic smartFET
NCV7708 / NCV7718 → multi-channel automotive driver
Microchip
MIC5019 → HS MOSFET driver
MAQ4123 / MAQ4124 / MAQ4125 (AEC-Q) → low-side driver building block
Pair with external TVS/RC for pre-charge
Melexis
MLX81150, MLX81123, MLX81124 → monitored coil / LIN-capable
Use when BMS expects feedback lines directly from driver
Validation & Test Plan
Validation for this page is still in the charging domain: we verify that pre-charge, measurement and main-contactor closure work for the intended charger/bus capacitance, and we verify that replacements (second-source drivers from the seven vendors) do not break waveforms, timings or thermal margins.
The plan has four pillars: waveform capture → thermal / soak → boundary conditions → alternative re-test. Every time a new driver is sourced, rerun the same four steps and write differences directly into the BOM remark.
Waveform capture
Capture: Vbus (pre-charge ramp), Icoil (pull-in → hold), Driver output / gate (clamp action). Trigger on pre-charge start and main close.
Thermal / soak
Coil temp at hold, driver junction temp, TVS/clamp temp after one failed + one successful pre-charge.
Boundary conditions
12 V → 9~10 V cold crank, long cable, bigger charger capacitor, −40°C / +125°C.
Alternative re-test
Re-run all items with 2nd-brand driver; log timing differences; update BOM so purchasing knows it cannot silently downgrade.
BOM Remark Library (Charging-Domain Contactor & Pre-Charge Driver)
This section is specifically for the charging-domain contactor & pre-charge driver. It is used when the original design selected an automotive, surge-suppressed, diagnostic-enabled driver, but small-batch purchasing tries to replace it with a cheaper FET or consumer high-side switch. Keep these remarks in English to talk to EMS / distributors.
Use these remarks only on this page. Do not reuse them on pack-FET / eFuse / insulation-leakage pages to avoid cross-page mixing.
Surge / spark suppression
“Driver must provide built-in surge/spark suppression equivalent to original design.”
Automotive vs consumer
“Do not replace automotive-grade high-side contactor drivers with consumer-grade low-side switches.”
Pre-charge feedback required
“Pre-charge completion signal is required; alternatives without feedback shall not be used.”
Pull-in / hold profile
“Driver shall support pull-in and hold current profile to limit coil heating.”
Sequence validation
“Validate sequencing under cold/nominal/hot supply before approving cross-brand alternatives.”
Seven-vendor rule
“For HV charging domain, use only parts from TI, ST, NXP, Renesas, onsemi, Microchip, or Melexis.”
Diagnostics must stay
“Diagnostics (coil open/short and pre-charge timeout) shall remain available at the driver pins or via MCU sense; parts without diagnostics are not approved.”
Clamp energy rating
“Clamp energy rating shall be equal or higher than the original part; do not approve replacements with unspecified clamp behavior.”
FET-only fallback
“Do not approve FET-only low-side solutions unless external TVS/RC network is explicitly added to match original clamp energy.”
Tip: paste 2–3 of the above into the actual BOM line of the contactor driver IC so that EMS/purchasing cannot drop surge, diagnostics or automotive grade.
Request a Quote
Frequently Asked Questions
Why do we need pre-charge before closing the main contactor in the charging domain?
Because the charger/DC link often has large input/output capacitors. If the main contactor is closed directly, inrush causes spark, contact wear and EMI. Pre-charge raises the bus to 0.8–0.9× target first, then it is safe to close.
What happens if purchasing replaces the smart driver with a simple low-side switch?
You lose current limiting, surge absorption and diagnostics. The sequence is no longer enforced, so a not-fully-charged bus can still close the main contactor. This shortens contactor life and increases spark risk. Such replacements must be rejected.
Can I use a driver without built-in surge suppression if I add an external diode?
A plain diode is the slowest way to release coil energy and is only acceptable as a minimum fix. If the original design used TVS or active clamp, the replacement must match or exceed that energy absorption. Otherwise the contactor and PCB will see higher stress.
How long should the pre-charge phase be before reporting PRECHARGE_DONE?
Long enough for the bus to reach about 80–90% of the final voltage, based on bus capacitance and chosen pre-charge resistor. The driver should not signal done until this level is detected. Timeouts must stop the main contactor from closing.
Can I run the contactor coil directly from the BMS MCU output?
No. MCU pins are not designed to pull in high-energy HV contactors, nor do they handle surge. Always use an automotive driver or smart high-side that can manage pull-in/hold, surge and diagnosis, then report status back to the BMS.
What diagnostic lines must the driver expose for a replacement to be accepted?
At minimum: COIL_OK (pulled in), PRECHARGE_DONE (bus reached target), and FAULT (timeout, overcurrent, coil issue). If available, WELD_DETECTED is also useful. Drivers without these signals should not be approved.
Why do we require automotive-grade drivers here?
Because this driver is directly tied to the HV charging path. Failure modes have higher energy and must be controlled. AEC-Q qualified parts specify temperature, surge and diagnostic behavior, while consumer parts usually do not.
How do we validate cross-brand replacements under cold and hot supply?
Re-run the full test plan: waveform capture (pre-charge, main close), thermal/soak, boundary supply (cold crank, high temp), and alternative re-test. Record timing differences and add them as BOM notes so purchasing cannot downgrade silently.
What if pre-charge times out or the bus never reaches 0.9 × target?
The main contactor must not close. The driver should assert FAULT or keep PRECHARGE_DONE low and let the charging controller decide the next action. Do not bypass this by forcing the main contactor on.
Can we disable the pull-in/hold current profile to simplify the BOM?
Not recommended. Pull-in/hold is what keeps the coil cool and reliable in low-airflow battery packs. Disabling it often makes purchasing think a cheaper driver is OK, but it shortens coil life and increases thermal stress.
Why is this handled in the charging domain instead of the pack-FET domain?
Because the decision to close the main contactor depends on pre-charge success, which is a charging-only sequence. Pack-FET/eFuse pages handle pack-level disconnects and leakage events, not charger-side inrush issues.
Which vendors are acceptable for this driver?
Limit to the seven automotive lines used in this page: TI, ST, NXP, Renesas, onsemi, Microchip, and Melexis. Other vendors may be used only if they match surge, diagnostics, sequence support and AEC-Q grade.
