| MembersPage/Sascha/JPDiaz |
From Aug 2014
Issues with lambda getting "stuck" at 0.83
Issue with lambda getting stuck when doing a power pull, issue started out of no where and consistently getting worse so the point that I can no longer read lambda under large loads.
Wiring has been checked, and re-checked... Sensor is new.
Sensor had a hard time to calibrate at the beginning (seemed like heater wasn't activating correctly all the time under calibration) I had to start/stop the calibration process a few times in order to get the O2 heater activated for free air cal. Seemed to read fine otherwise and is accurate under light loads @ idle.
Screen shot of log:
Lambda sensor eventually recovers in approx 2-3 seconds of overrun:
Update - Aug 2016
Late last year the owner replaced the WBO2 w/ a new sensor and let me know that it calibrated ok. I wrote the lambda issue off to the sensor after all.
2 Days ago, pre-test race w/ swapped to E98 (from E85) and I noticed the WBO2 was not working at all again. After numerous attempts of free air calibration with the new sensor, the original sensor and one brand new never used O2 sensor. All 3 sensors had the same issue. During calibration the heater would never stop heating.
Checked wiring again, everything checks out, 12V source is good. I did not have my resistors with me to check the WB output from the ECU so I swapped to a known good ECU and the issue was solved.
It's clear the customers ECU has a defect from HU. Unfortunately it is outside of the warranty period now, but I will be bench testing the ECU and hopefully can be guided to do the repair on the ECU. I'm hoping it is maybe just a bad solder point or driver/IGBT for the WB heater.
UPDATE FEB 2017
I have bench tested the ECU with the following results:
Following test steps listed here: http://megasquirtavr.sourceforge.net/manual/Detailed.Sensor.Wide.html#Operate
Voltage test / ECU ON / WBO2 ON (Heater):
Pin 5 (pump -) + GND = 3.98V (no change to this when adjusting pw_zero up or down - w/ burn + restart of ECU)
Pin 1 (nernst) + GND = 4.41-42V
Pin 1 + Pin 5 = 0.445V
GenBoard/Manual/WBSensorControllerTesting steps:
Voltage Test / ECU ON / WBO2 OFF (Heater):
pump- = 0.0V
GND = -3.97V (again no change when adj. pw_zero)
Voltage Test / ECU ON / WBO2 ON (Heater):
pump- = 0.0V
nernst+ = 0.445V
pump+ = 0.910V
All tests done with correct 100 ohm resistors as stated in the links.
- tests seem OK
- pump- value = constant 3.97V is perfect ( 4 +- 0.1V so well within... even more deviation would not cause any problem)
- it is constant, pump- should NOT change lower then recommended 4.1V and will not change up/down when adjusting pw_zero setting at all.
- Sensor is not the problem
- how do we know ?
- So issue must be related somehow to this ECU
- HW seems good (please measure that WBO2 heater output is not continuously pulling to GND, eg. measure between +12V and EC18/18 before and after ECU powerup)
- so perhaps some (wideband related) configuration. Perhaps RI target is too low ? (normally 165 .. 180, try 180)
- Include config (as in any report), or better: vemslog captured during free air calibration
ECU SN: 8501
sometimes it happens that exhaust is blocked and then lambda fails
Dave reviewed config, all defaults applied no config issues found
- for any further diagnosys a vemslog of wb calibration (warmup from cold to operational running 1-2min after operational, engine NOT running, knock realtime data set to: individual power)
- IMPORTANT: use recent firmware (1.2.31 or higher)
- the config shows current firmware is 1.2.28 which does NOT log enough information for this problem
- Updated FW last night, set knock realtime data correctly, started WB calibration to ensure warm up... saved vemslog here: http://www.vems.hu/files/MembersPage/Sascha/JP/v3.3_u008501-2017.06.15-22.25.52.vemslog
- the config shows current firmware is 1.2.28 which does NOT log enough information for this problem
Review of v3.3_u008501-2017.06.15-22.25.52.vemslog:
- correct upload of config was bypassed on this firmware upgrade (see config validation)
- wbo2 configuration is sensible (matching defaults), LSU 4.2 selected
- wbo2 Ri does not come down to target range during the warmup period, this points to a problem with either:
- wbo2 wiring between ecu and sensor, please measure with DVM in resistance mode for proper continuity.
- bad connection from ec18 to wires or wires to LSU 4.2 plug, please measure with DVM in resistance mode for proper continuity.
- vbatt missing on wbo2 plug, please measure with DVM in voltage mode if vbatt is present on LSU 4.2 plug pin #3.
- failure of the wbo2 heater control output stage on the ecu, please measure with DVM in voltage mode the voltage between LSU 4.2 pin #4 and GND with wideband sensor connected (DVM probing from back of plug) with wbo2 off, wbo2 calibration just started and wbo2 at end of warmup.
If you provide measurements to each of the measurements Q's above we should be able to locate the cause of the issue. Thanks!
Best regards, Dave