A late OK from the ELM327: one answer behind, phantom modules
On a 2015 Prius a slow adapter's late OK was read as the car's answer to 01 00. The scan reset the adapter and counted modules that never replied.
Field Notes
What real cars answered
Each note comes from a scan of a real vehicle in a working bay: the request, the exact bytes that came back, and what they turned out to mean. Where a car refused a request, or a reading was wrong, that is written down too. VINs are removed.
Hyundai odometer over UDS: DID B002 and its two record layouts
A 2019 Santa Fe answered 22 B002 with 12 bytes, and the last three matched the dash. Two 7-byte replies from other Hyundais gave no mileage to trust.
Kia UDS $19 02 FF: a code catalogue, not a fault list
A 2018 Optima with one stored code came back with 25, one flagged MIL-on. The reply was a code catalogue, and 13 lost CAN frames shifted the rest.
Mode 09 with several modules: whose calibration ID is whose
A 2014 Suburban listed its CalIDs and CVNs in different module orders, and on a 2009 BMW the parser turned clean CalIDs into garbage. How to pair them.
Only 4 supported PIDs? When several modules answer one request
A 2019 Buick Encore reported 4 supported PIDs because the fastest of three modules won. The same race skews readiness and flipped a GV70's distance since clear.
Toyota odometer over OBD: DID 0103 and where it answers
A 2018 Camry answered 22 0103 with 02 00 E2 36 and the dash read 57,910 miles. What each byte means, and which older Toyotas refuse the request.
0 codes across 0 modules is not proof of a clean car
A 2015 Odyssey's deep scan showed a green tick over zero modules read, after its ECM and TCM had answered. How a scan tool should tell unread from clean.