[Diy_efi] Making a fake O2
phil at injec.com
Sat May 28 12:52:51 CDT 2005
Without access to all the ECUs software it would be impossible
to "fool" an ECU with a constant or switching voltage source.
There are many conditions that must be met for closed loop
operation and one is the response time of the O2 sensor
to a fuel or spark change, all ECUs "test" the O2 sensor for
deterioration that comes with age, fouling etc.
As this response time varies with rpm as well as spark advance
you would need to construct some serious piece of equipment
that would be nearly as complex as the ECU itself.
As Adam pointed out a disconnect is something that is part of
the ECUs std operating conditions.
The resultant fuel and spark is always safe. Dead O2 sensors are
a simple fact of day to day opps. There are lots of oil and fuel
additives that are sold in speed shops that can block the "pores"
of an O2 sensor in minutes.
The manufacturers generally won't flag a check engine warning as
many of the sensors would have to be replaced under warranty. It
would be very hard to prove that the sensor was killed by the
customer. Many are killed by the use of the wrong jointing
compounds after exhaust manifold removal.
Closed loop is only used at rpm and load sites that result in clean
combustion suitable for the rich lean cycling that the cat wants to
see for good reduction/oxidation of the feed gas.
Forcing open loop mode will in most cases not cause the ECU to
drop the lean cruise mode or do any thing that cost any power or
performance. Keep in mind that if there were no cat fitted there would
never be a reason to invoke closed loop with a narrow band sensor.
The rub is that the long term block learn modes (of some ECUs) rely
on the O2 feedback for table correction in the fuel area. Simply
the ECU can determine that the injector flow is down by X % and
correct for dirty injectors. A reconnect and a very gentle cruise
at below 80 kM/hr for 20 minutes will bring you back to a good
More information about the Diy_efi