DIAG GOLDNetwork Diagnostics
The CAN Bus Playbook
Every meter reading and scope pattern a network complaint can throw at you, what actually causes it, and what to check next. Ground rules: resistance is measured at DLC pins 6↔14 with the key OFF and the bus asleep (a live bus skews ohms); voltages are key ON. Measure at the DLC or backprobe with proper leads — piercing the twisted pair plants next year's corrosion no-comm.
✅ Healthy bus
DLC 6↔14, key off:60 Ω
CAN-H idles 2.5V, pulses to ~3.5V · CAN-L idles 2.5V, dips to ~1.5V — perfect mirror images
Likely causes (in order):
- Network is fine. A no-comm complaint here means one dead module (or the scan tool), not the bus
The mirror test: H and L should sum to ~5V at every instant. If the bus checks good but one module is missing from the scan list, go to THAT module and check its power, ground, and bus pins — most "CAN problems" are a dead module, not a dead network.
Tap to enlarge ⤢
🔌 Meter reads 120 Ω
DLC 6↔14, key off:120 Ω
Waveform still switching, but with overshoot/ringing on the edges
Likely causes (in order):
- One terminating resistor missing — a terminator module unplugged or dead
- Bus severed between the two terminators (broken wire, corroded connector)
The two 120Ω terminators live inside modules at opposite ends of the bus — commonly the ECM and the cluster or gateway. A one-terminator bus often still RUNS, just marginal: intermittent U-codes, worse hot or over bumps. Find the break between the terminators; wiggle-test the harness while watching the meter.
Tap to enlarge ⤢
💥 Meter reads ~0 Ω
DLC 6↔14, key off:0–10 Ω
CAN-H and CAN-L shorted together — both lines ride at ~2.5V, total no-comm, warning lights everywhere
Likely causes (in order):
- Chafed twisted pair — engine-to-body pass-through grommets, under-dash kick panels, seat harnesses
- Crushed or backed-out connector terminal
- Pinched harness after a collision or previous repair
Split-half search: open the bus at an accessible mid-point connector and see which side keeps the short. Repeat until you corner it. Check anywhere the harness was recently disturbed first — the last repair is the prime suspect.
Tap to enlarge ⤢
🪫 A line stuck at 0V
Pin 6 (or 14) ↔ ground:0 Ω
One CAN line flat on the floor — bus dead or in fault-tolerant limp, scan tool refused
Likely causes (in order):
- Wire chafed to body ground
- Water in a connector (check the DLC itself!)
- Failed module transceiver clamping the line down
Unplug modules ONE at a time with the scope connected — the moment the line pops back to 2.5V, the last thing you unplugged (or its branch) is the perp. No luck at the modules? It's the harness: split-half the wiring the same way.
Tap to enlarge ⤢
⚡ A line at battery voltage
Pin 6 (or 14) ↔ ground:≈12 V
CAN line pulled to ~12V — off the top of the scope, total no-comm
Likely causes (in order):
- Harness chafe against a 12V feed (shared loom with power wires)
- Failed transceiver leaking supply voltage onto the bus
Same unplug-one-at-a-time hunt as a ground short. Bonus clue: check whether the 12V on the line is switched or constant — it tells you which power feed it's touching, which tells you WHERE the looms run together.
Tap to enlarge ⤢
😴 Flat 2.5V — asleep or dead?
DLC 6↔14, key off:60 Ω
Both lines flat at 2.5V, zero activity
Likely causes (in order):
- Bus is legitimately ASLEEP (networks power down minutes after key-off) — not a fault
- Key on and still silent: no module able to transmit, or bus-off from a fault that just cleared its short
Wake it before you condemn it: key on, open a door, tap the brake pedal — traffic should erupt instantly. Still flat with key on? Check power and ground at the main talkers (ECM, BCM) before touching the wiring. Flat lines with GOOD resistance = a power problem, not a bus problem.
Tap to enlarge ⤢
🗯️ One module flooding the bus
DLC 6↔14, key off:60 Ω
Wall-to-wall corrupt frames, scan tool connects and drops, modules randomly offline
Likely causes (in order):
- A failing module transmitting garbage (the classic "babbling idiot")
- Water-damaged module doing the same
- Aftermarket devices: insurance dongles in the DLC, trailer-light converters, remote starts, cheap LED controllers backfeeding
Unplug the DLC dongle FIRST — thirty seconds, fixes more of these than anyone admits. Then aftermarket add-ons, then factory modules one at a time until the bus cleans up. The scope tells you instantly when the noise stops; a scan tool takes a full re-poll each time.
Tap to enlarge ⤢
🕳️ Scan tool sees some modules, not others
DLC 6↔14, key off:60 Ω
Diagnostic bus tests perfect — the missing modules live on a DIFFERENT bus
Likely causes (in order):
- Most vehicles run 2–5 separate networks; only some reach the DLC directly
- A gateway module bridges the rest — if the gateway is down, whole bus segments vanish at once
- Budget scan tool that only speaks the OBD2 protocols
Modules vanishing in GROUPS is the tell — one whole bus is down or the gateway dropped it. Pins 6/14 are the high-speed diagnostic bus almost universally; body/comfort buses ride behind the gateway (or on other pins — see the pin map below). U-codes on the modules you CAN reach will name the ones you can't.
Tap to enlarge ⤢
The DLC pin map
The 16-pin connector under the dash carries more than one network. Knowing which pins carry which bus is half of "scan tool sees some modules, not others." Tap any pin for the full story — what it carries, exact probe values, and which makes we've verified it on.
HS-CAN High — pin 6HS-CAN Low — pin 14Ford MS-CAN — pins 3 / 11GM single-wire CAN — pin 1J1850 pre-CAN — pins 2 / 10K-line / L-line — pins 7 / 15Grounds — pins 4 / 5Battery + — pin 16
| Pin(s) | Bus | What it is |
|---|---|---|
| 6 / 14 | HS-CAN (500 kb/s) | The main diagnostic + powertrain bus. Near-universal on 2008+, common back to ~2004. |
| 3 / 11 | MS-CAN (125 kb/s) | Ford medium-speed body bus — audio, HVAC, cluster on many 2005–2019 Fords. |
| 1 | Single-wire CAN (33.3 kb/s) | GM GMLAN low-speed body bus — one wire, no mirror trace. |
| 2 / 10 | J1850 (pre-CAN) | GM Class 2 (VPW, pin 2 only) and Ford SCP (PWM, both pins) — the 1996–2007 domestics. |
| 7 / 15 | K-line / L-line (ISO 9141) | Pre-CAN serial diagnostics — most 1996–2007 European and Asian vehicles. |
| 4 / 5 | Grounds | Chassis ground / signal ground. Verify these FIRST when the scan tool itself won't power up. |
| 16 | Battery + | Always-hot feed for the scan tool. Blown DLC fuse = "my scanner won't even turn on." |
| 8, 9, 12, 13 | Manufacturer discretionary | Anything the maker wants — extra buses, immobilizer lines, nothing. Check the wiring diagram before probing. |
Reading U-codes like a map
Every module on the bus logs "lost communication with X" (U0100 = ECM, U0121 = ABS, U0155 = cluster…), so the PATTERN is the answer, not any single code. Every module pointing at ONE module → that module or its branch is dead — check its power, ground, and bus pins. One module logging lost-comm with EVERYONE → that module's own bus connection is the problem. Whole groups gone at once → a bus segment or the gateway is down. Pull codes from ALL modules before you touch a meter — the network draws you its own fault map.
The two rules that prevent 80% of network comebacks
(1) Check the simple stuff first — DLC dongles, aftermarket add-ons, battery voltage, module powers and grounds. Low system voltage alone will litter every module with U-codes; don't diagnose a network on a weak battery. (2) Never condemn a module while it's wearing a bad bus — fix the wiring fault, clear the codes, THEN re-evaluate. Modules get replaced every day for shorts that lived in the harness.
Network maps by make
The values above are universal. WHERE the terminators live, which buses a make runs, and what breaks on them is not — pick your make:
Chased everything and it still isn't talking? That's what the Diag Desk is for — a vetted master tech looks at YOUR readings on YOUR vehicle. $25, free if unanswered in 24h. Looking up a specific U-code? The DTC library has the full network family.