Slower-Than-Spec DC Bus Discharge After Service Disconnect
A 2019 battery-electric sedan (350 V nominal, 96-series Li-ion pack) is in for high-voltage cable replacement at the charge-port end. Following the OEM procedure, the technician powers down, removes the HV service disconnect under the rear seat, waits the specified 5 minutes, then measures DC bus voltage at the inverter terminals with a CAT III 1000 V DMM whose calibration sticker is current. Readings: 78 V at 5 minutes, 62 V at 10 minutes, 55 V at 20 minutes, still falling slowly. Shop temperature is 9 °C. The OEM service information states the bus must be below 5 V DC before any HV connector is opened. A second technician says 55 V is fine because "anything under 60 volts isn't high voltage." The vehicle is on a two-post lift with the pack installed and the disconnect lying on the bench.
1. How should the technician act on the 55 V reading, and what does the decay pattern suggest?
Read the worked answer
Stop the job - the release criterion has not been met. 60 V DC is the threshold at which a system is classified as high voltage (SAE J1673 / ISO 6469-3). It is not a touch-safe line and it is not the release criterion. The release criterion here is the OEM's: < 5 V DC. At 20 minutes the bus is 11× that value, so no HV connector gets opened.
The shape of the curve is the important finding. A healthy active-discharge/bleeder circuit takes a ~350 V DC link to a few volts in seconds to a couple of minutes. Losing only ~23 V in 15 minutes looks like a capacitor with almost no load on it - or like a bus that is being fed.
Two hypotheses, both of which keep the vehicle quarantined:
- Failed bleeder / open active-discharge circuit - common after an inverter fault or a prior repair.
- A parallel path is still connected - a welded or held HV contactor, a DC-DC converter or charger back-feeding, or a second HV source. If a contactor is welded, "key off, disconnect out" did not isolate the pack, and the car is energized right now.
Actions: tag and lock out, keep the disconnect out and in the technician's control, bar anyone from opening HV connectors, re-verify the meter and leads on a known live source, confirm the disconnect is fully unseated rather than merely unfastened, then run the OEM discharge-circuit and contactor-weld diagnostics. Coolant temperature and ambient temperature do not explain a 20-minute discharge.
2. What PPE and measurement practice apply while this is being diagnosed, and why?
Read the worked answer
Treat the bus as energized throughout, because that is exactly what the evidence says it is.
Gloves: Class 0 rubber insulating gloves (1000 V AC / 1500 V DC) with leather protectors, current on the 6-month dielectric test, visually inspected and air-tested (roll the cuff, watch for a balloon-and-listen leak) before each use. No jewelry, no watch.
Body: Arc-rated face shield and long sleeves if the task risk assessment (opening a connector, probing a live terminal) calls for it. Insulated tools rated for the working voltage. Nonconductive floor mat or an insulating barrier over the pack.
Meter: CAT III 1000 V, shrouded probes, leads of matching rating, no cracked insulation, no exposed metal at the tips. Verify the meter on a known source first - a meter that reads 0.0 V because the lead is broken has killed people.
Technique: single-hand probing with the other hand off the vehicle, meter in a hands-free mount rather than lying on the pack lid, a second person present with the service-disconnect key in their pocket. The vehicle stays on the lift only if nobody is working beneath a suspended pack.
3. The cable replacement is finished. The shop wants to release the car and schedule the discharge issue for a later visit. What is your position?
Read the worked answer
No release. A vehicle whose DC bus is still at 55 V twenty minutes after isolation - or that shows evidence of a welded contactor - has an active HV safety defect. Released to the customer, that car can present an energized bus with the key off and the disconnect out, to a customer, a tow driver, or the next shop.
What to do instead:
- Write it up as an unresolved HV safety condition, keep the vehicle in a controlled area, and document the voltage-versus-time measurements.
- Continue to the contactor weld and discharge-circuit diagnosis rather than deferring it; the two are usually the same repair.
- Get the OEM technical line involved if the discharge circuit tests good but the bus still will not fall below spec - that combination points at something in the pack or a non-obvious parallel path.
- Never 'solve' it by holding a resistor across the bus or by shorting the terminals to bleed it down. Both defeat the fault detection you are relying on and one of them starts fires.