O2 sensor waveform: how to tell a dead lambda from a live one by shape

An O2 sensor outputs voltage proportional to oxygen content in the exhaust. But you don’t diagnose it correctly by a single voltage in a moment — you look at the oscillation shape (waveform). Below: how to interpret upstream and downstream waveforms, and what the AI does.

Upstream vs downstream O2

• Upstream (B1S1, B2S1) — before the catalyst. Actively oscillates 0.1-0.9V at 1-2 Hz because the ECU is in closed-loop constantly adjusting the mix.
• Downstream (B1S2, B2S2) — after the catalyst. Should be stable at 0.6-0.8V if the catalyst is working (buffers oscillations).

Normal upstream O2 waveform

Healthy B1S1 reading at 2000 RPM idle:

V
0.9 |     __      __      __      __
0.8 |    /  \    /  \    /  \    /  \
0.7 |   /    \  /    \  /    \  /    \
0.6 |  /      \/      \/      \/      \
0.5 | /                                 \
0.4 |
0.3 |
0.2 |__________________________________
    time →  ~500ms per cycle (2 Hz)

• Amplitude: 0.1-0.9V (minimum 0.2 - maximum 0.85)
• Frequency: 1-2 Hz (0.5-1 sec per cycle)
• Symmetry: rise and fall roughly equal in time

O2 aging symptoms

1. Reduced amplitude

V
0.9 |
0.8 |
0.7 |     __      __      __
0.6 |    /  \    /  \    /  \    <-- 0.4-0.7V instead of 0.1-0.9
0.5 |   /    \__/    \__/    \
0.4 |
0.3 |

Cause: platinum coating ages, sensitivity drops. Still works but correction is imprecise.

2. Slow response

V
0.9 |     _________________
0.8 |    /                 \
0.7 |   /                   \
0.6 |  /                     \
0.5 | /                       \
0.4 |/                         \_
0.3 |
    time →  3-5 sec per cycle instead of 0.5-1 sec

Cause: sensor heater weak (for heated zonds) or surface fouled. Often precedes P0133/P0135.

3. Stuck lean

V
0.9 |
0.8 |
...
0.4 |__________________________
0.3 |                          <-- stuck at 0.3-0.4V
0.2 |
0.1 |

Cause: sensor “poisoned” (lead, silicone, oil). Won’t switch. P0134 — “Circuit no activity detected”.

Normal downstream O2 waveform (B1S2)

Healthy downstream reading — a stable line at 0.6-0.8V:

V
0.9 |
0.8 |________________________________  <-- flat 0.7V
0.7 |
0.6 |
0.5 |
0.4 |________________________________
    time →

Sign of a dead catalyst

If downstream O2 starts oscillating in sync with upstream — the catalyst isn’t holding, downstream mirrors input:

B1S1 (upstream):    /\/\/\/\/\/\/\/\/  <-- actively oscillates
B1S2 (downstream):  /\/\/\/\/\/\/\/\/  <-- copies!

That’s the sign of P0420 / P0430 (see P0420 post and P0430).

How to view it in Car Diagnostics

1. Live Monitor → add PID O2_B1S1 and O2_B1S2 simultaneously
2. Chart shows both in one window
3. Warm car to 90°C, hold 2000 RPM idle
4. Watch 30-60 sec

Healthy signs:

• B1S1 oscillates 0.1-0.9V at 1-2 Hz
• B1S2 stable at 0.6-0.8V

Symptoms:

• B1S1 amplitude < 0.4V → sensor tired
• B1S1 frequency < 0.5 Hz → slow response
• B1S2 oscillates → catalyst or downstream sensor

What AI agent adds

Our DTC Analyzer automatically analyses the waveform:

• Computes switch count (how many times B1S1 crossed 0.45V in 10 sec) — spec 5-10
• Compares with a “healthy” curve database for your make/model
• Returns a ranked verdict: “sensor healthy 92%”, “sensor tired 78%”, “sensor dead 95%”

That’s what a pro does with an oscilloscope in 5 min. AI — in 5 sec.

Practical tips

• O2 sensor replacement — 60-90k mi (preventive, don’t wait for P0133)
• OEM Bosch/Denso/NGK — don’t buy AliExpress clones, will kill the catalyst
• Only one side (bank 1 or bank 2) at a time, both sides rarely fail together
• After replacement — see post-repair verification

Try it on the free tier — pair an ELM327 adapter. Standard PIDs O2_B1S1, O2_B1S2 on every car. Automated AI waveform analysis — on paid tier.