Stage: SLAC matching
How the car's powerline modem finds the charger it is plugged into: parameter exchange, sounding, the attenuation report, the match and the network key, and what makes the stage fail.
What the stage does
Powerline signals leak between neighbouring cables, so a car can hear several charger modems at once. SLAC (Signal Level Attenuation Characterization) lets the car measure how strongly each charger hears it and join the logical network of the one with the lowest attenuation, which should be the one at the end of its own cable. Nothing above it, SDP, TCP or any V2G message, can run before the match.
- It starts when the charger sets the 5 percent duty cycle on the Control Pilot and the car is in state B. CP-04 checks that the 5 percent came first.
- The car drives the sequence. Every attempt carries a RunID that stays the same from the first to the last message of the attempt (SL-02).
- It ends with CM_SLAC_MATCH.CNF, which hands the car the network ID and the network membership key of the charger. The car writes the key into its own modem with CM_SET_KEY.REQ, and from then on both modems share one private network.
The messages
- EV: CM_SLAC_PARM.REQ, broadcast, opens an attempt with a new RunID
- EVSE: CM_SLAC_PARM.CNF, from every charger that hears it; names the number of sounds and a timeout
- EV: CM_START_ATTEN_CHAR.IND, announces the sounding
- EV: CM_MNBC_SOUND.IND, the sounds themselves, as many as NUM_SOUNDS
- EVSE: CM_ATTEN_PROFILE.IND, from the charger modem to its host, one per sound
- EVSE: CM_ATTEN_CHAR.IND, the averaged attenuation per carrier group
- EV: CM_ATTEN_CHAR.RSP
- EV: CM_SLAC_MATCH.REQ, to the charger with the lowest attenuation
- EVSE: CM_SLAC_MATCH.CNF, network ID and key
- EV: CM_SET_KEY.REQ, car host to its own modem; the modem joins the network
An optional validation step (CM_VALIDATE) can sit between the attenuation report and the match, where the charger toggles the Control Pilot to prove which cable the car is on. Most sessions skip it.
Timers and retries
| Timer or limit | Limit profile key | Profile default |
|---|---|---|
| First CM_SLAC_PARM.REQ to CM_SLAC_MATCH.CNF | slac.matchTimeoutS | 10 s |
| Silence after one SLAC message before the next is overdue | slac.stepTimeoutS | 2 s |
| TT_EV_atten_results, first CM_START_ATTEN_CHAR.IND to the attenuation report | slac.ttEvAttenResultsMs | 1200 ms |
| Spacing of the sound messages | slac.batchIntervalMinMs to slac.batchIntervalMaxMs | 20 to 50 ms |
| TT_EVSE_SLAC_init, how long the charger waits for SLAC to start | cp.retry.ttEvseSlacInitMinS to cp.retry.ttEvseSlacInitMaxS | 20 to 50 s |
| T_step_EF, the E or F step before the charger offers SLAC again | cp.retry.tStepEfMinS to cp.retry.tStepEfMaxS | 3 to 5 s |
| Sounds the charger normally asks for | slac.numSoundsExpected | 10 |
| Attempts before SL-07 warns | slac.retriesWarnAbove | more than 1 |
| Average attenuation a link should stay below | slac.attenuationMaxDb | 40 dB |
A car that does not get a match starts a new attempt with a new RunID. A charger that hears no SLAC at all within TT_EVSE_SLAC_init drops the 5 percent, steps through state E or F and offers it again (CP-05), which resets a car that got stuck.
All SLAC timers in the profile, and the attenuation threshold, are flagged for verification. The threshold is lab practice rather than a value from ISO 15118-3; set it to what your lab accepts.
What usually breaks
- High attenuation: a long or damaged cable, a poorly coupled modem, a missing ground reference. The attenuation report comes back, but the car refuses to match or retries (SL-06, verdict SLAC attenuation too high).
- No answer to CM_SLAC_PARM.REQ: the charger modem is not in the right mode, or its host is not listening yet. The verdict SLAC timeout names the side that owed the next message.
- Missing or late sounds: fewer CM_MNBC_SOUND.IND than announced (SL-03), gaps outside the window (SL-04), or the attenuation report after TT_EV_atten_results (SL-05).
- Crosstalk: the car matched a neighbouring charger. SL-09 compares the charger the car picked with the lowest attenuation it was offered.
- The match completes but the key never reaches the modem. SLAC looks fine and the failure shows one stage later as an unanswered SDP request.
- Many attempts before a match (SL-07) or a long total duration (SL-08): the session works, but a marginal link will fail on another day.
Symptom pages: Car plugged in, no SLAC match and A weak or disturbed powerline link. A full walk through a failing capture: Worked example: SLAC that never matches.
In the analyzer
- Stage strip: SLAC, the second stage of every high level session, after Plug-in. It passes at the first CM_SLAC_MATCH.CNF of an attempt that matched.
- Checks: SL-01 to SL-09, CP-04 and CP-05 in the Control Pilot and SLAC groups of the Checks tab.
- SLAC tab: every attempt with its RunID, the responders and the attenuation per carrier group.
- Verdicts: SLAC timeout, SLAC attenuation too high, or No digital communication when the car never started.
- A capture taken on the Ethernet side of the modem has no SLAC at all. The stage then shows as not in the capture rather than failed, as long as SDP and the rest are there.
Sources
- ISO 15118-3:2015, SLAC procedure, its timers and constants, V2G3-M06-07 for the retry through E or F
- IEC 61851-1:2017, Table A.8, the 5 percent duty cycle
- HomePlug Green PHY specification, management message types
- ChargeX Consortium: ev-charge-seq-state, Argonne National Laboratory, Apache-2.0: PlantUML diagrams of SAE J1772 PWM charging and the Control Pilot state machine, and ISO 15118-2 and -20 sessions with OCPP 1.6J, 2.0.1 and 2.1; the diagrams start after the link is set up and do not draw the SLAC messages