SAE J1850 is the American legislated diagnostic bus that sat beside ISO 9141 on cars built before CAN became the emissions path. It comes in two electrical forms, and a car uses one of them. Variable pulse width is a single wire at 10.4 kbit/s, used by General Motors and by Chrysler. Pulse-width modulation is a pair at 41.6 kbit/s, used by Ford. Both are half-duplex, both let more than one module talk, and both carry an 8-bit cyclic redundancy check. The 16-pin socket that presents them is on the diagnostic ports page. The order a scan tool tries them against the K-line and CAN is on the K-line page.
Variable pulse width
VPW uses one wire, pin 2 of the 16-pin connector, against vehicle ground. The idle state is low. A transmitter drives the wire up to about 7 V for an active bit and lets it fall for a passive bit. The width of the pulse is the data. A short cell is 64 microseconds and a long cell is 128 microseconds, so the same voltage means a different bit depending on how long it lasts. The start of a frame is a longer active pulse, about 200 microseconds, which is how a listener finds the beginning without a separate clock wire. The nominal rate quoted for the bus is 10.4 kbit/s. That is the design rate of the short and long cells, not a UART baud rate. A tool set up for the K-line's 10.4 kbit/s start-and-stop bytes will not decode these pulses.
In-frame response is part of the same transmission. After the checksum, a module that was addressed can append a short reply in the same frame, instead of waiting to win the bus for a separate message. A trace that shows extra bits after the check is often that reply, and the next start-of-frame is the next message.
Pulse-width modulation
PWM uses two wires, pin 2 and pin 10. The bit time is 24 microseconds, which is 41.6 kbit/s. Each bit is a pulse whose width, inside that 24 microsecond cell, is the value. The pair is driven so that the two wires move as a couple, which is why the loom for this Ford form is a twisted pair and why a single-wire VPW transceiver on pin 2 alone does not speak it. The frame still ends in an 8-bit cyclic redundancy check, and in-frame response is used in the same role as on VPW.
A module that sees a dominant condition while it is trying to send a passive bit gives the bus up. The higher-priority frame continues without starting again. That is the same idea as CAN arbitration, applied to pulse widths. More than one module can be a talker. There is no single commander handing out turns.
Finding it on a car
SAE J1978 includes both forms in the protocol test a legislated scan tool performs. One worked order tries PWM, then VPW, then the K-line protocols, then CAN. The wires are distinct, so a tool can look at pin 2 and pin 10 without having completed a K-line wake-up. Activity on pin 2 with pin 10 idle points at VPW. A differential signal across pin 2 and pin 10 points at PWM. Silence on both, with the K-line still high and the CAN pair recessive, means this car's emissions conversation is one of the other protocols.
J1850 stayed in service on vehicles whose emissions electronics were designed around it, including later years of those platforms. It is a bus of its own. A later module that offers CAN diagnostics and still has a J1850 pin in the socket is following that car's wiring, and the socket pinout is the document that says which of the two forms is present.
Sources
- SAE J1850. Class B data communication: variable pulse width at 10.4 kbit/s on a single wire, and pulse-width modulation at 41.6 kbit/s on a pair, with an 8-bit cyclic redundancy check and in-frame response.
- SAE J1962. The 16-pin diagnostic connector. Pin 2 is the positive J1850 line. Pin 10 is the negative line used by the pulse-width-modulated form.
- SAE J1978. The scan-tool protocol test that includes both J1850 forms.