A weak or disturbed powerline link

SLAC needs several attempts or matches the wrong charger, SDP repeats, TCP retransmits and answers come late: the modems can barely hear each other. How to tell a poor link from a poor capture, and what usually helps.

What you see

  • SLAC takes several attempts, each with a new RunID, or reports high attenuation (SL-06, SL-07, SL-08).
  • Several chargers answer CM_SLAC_PARM.REQ, and the car picks one that was not the loudest (SL-09).
  • SDP requests repeat before an answer comes (SD-02), TCP segments are sent again (TC-03), and response times spread widely with an occasional very slow answer (TIM-06).
  • In the worst case the session ends mid charge with Connection lost or Timeout, although neither side reported a fault.

Link or capture?

A sniffer is a third modem on the line, and its link can be worse than the one between the two modems it listens to. Before you blame the cable, check whether car and charger heard each other:

SignPoints at
The car moved on to its next request although the answer is missing from the fileThe capture: the charger answered and the car got it. The Capture tool lists these as answers missing from the capture
Only the car is heard during SLAC, yet V2G messages followThe capture: the sniffer does not hear the charger modem
The same TCP segment twice, the original and the resendUsually the link, since a sender only resends what was not acknowledged; a sniffer that records one frame twice looks similar, so compare with the attenuation
The car repeats a request after its timeoutThe link or the charger: the car did not get an answer in time
High attenuation in CM_ATTEN_CHAR.INDThe link between car and charger, as the charger modem measured it, whatever the sniffer heard

The Capture tool grades the file good, fair or poor and says for each problem what it means for the checks. Read it before the findings.

Likely causes

CauseHow to confirmWho
Coupling loss: a long or damaged pilot wire, a missing or wrong coupling component, a poor contact in the connectorHigh attenuation in every carrier group, with every car at this chargerEVSE or cable
Noise from power electronics coupling into the pilot circuitThe link is fine during negotiation and degrades once current flows: retransmissions and slow answers cluster in the charge loopEither
Narrowband interferenceHigh attenuation in a few carrier groups while the rest look normalSite
Crosstalk from neighbouring charge pointsSeveral responders with close values in the SLAC tab; the problem goes away with only one car plugged inSite
A modem configured for the wrong role or with a wrong parameter blockA modem that answers late or not at all, or attenuation far from other units of the same model; compare with a known good modemEV or EVSE
The car matched the wrong chargerSL-09; the SDP answer comes from a charger other than the one plugged in, or none comes at allEV or site

What to check in this tool

  • SLAC tab: the attempts with their RunIDs, every responder with its average attenuation, and the attenuation per carrier group.
  • Capture tool: answers or requests missing from the capture, TCP bytes that never arrived, retransmissions, a side never heard.
  • Timeline: the SDP and TCP/TLS lanes for repeated requests and resends, and where they sit relative to the start of current.
  • Timing tab and Timers tool: TIM-06 for the spread of response times, and how close each answer came to its limit.

Fixes that usually work

  • Test with a short, known good cable, then with the real one, and compare the attenuation.
  • Inspect the pilot contacts of connector and inlet, and the coupling circuit on both sides.
  • Separate the cables of neighbouring charge points and check the site's filtering; test with one car plugged in.
  • Put the sniffer close to the connector, on the cable in use, and check it hears both modems before you trust a capture.
  • Set slac.attenuationMaxDb to what your coupling normally shows, so SL-06 flags real outliers rather than every session.

slac.attenuationMaxDb (40 dB) is lab practice, not a value from the standard, and the SLAC timers are flagged for verification against ISO 15118-3.

Sources

  • ISO 15118-3:2015, SLAC, attenuation characterization
  • HomePlug Green PHY specification, HomePlug Powerline Alliance
  • SPEC.md of this tool, FR-TOOL-04, the capture quality tool