Bypass circuit for "Check Engine" Carbon build up..
02-02-2003, 03:51 AM
#1
Addict
Rennlist Member
Rennlist Member
Thread Starter
Join Date: Apr 2002
Location: Newcastle, WA
Posts: 195
Likes: 0
Received 0 Likes
on
0 Posts
Bypass circuit for "Check Engine" Carbon build up..
After reading many of posts about Carbon build up, I've decided
to post my discovery of eight months ago of a way to manipulate the
data sent to the OBDII system for the "Check Engine" light caused
by carbon build up. I want to spend more time testing and confirming
the results. Although, my preliminary tests suggest the results are
conclusive. My car is a '97 C4S with 61K miles.
I have carbon build up blockage of air injection ports of cylinders 1-3.
From what I have determined, the diagnostic is accomplished via the O2
(Oxygen) sensors. The OBDII monitors the output voltage of the Left / Right
banks of O2 sensors during the air pump cycle at cold start up.
If the catalytic converter reaches temperature during the air pump cycle,
the four oxygen sensor voltages drop to almost zero volts between the
O2 sensor output terminals, about two minutes after start up.
If one side of O2 sensors fail to drop in voltage after sufficient time
during the air pump cycle, the non voltage drop is stored
as a fault -giving the resultant "Check Engine" light. This fault is
recorded as blocked air passages, indicating banks 1-3 or 4-6.
Monitoring with an oscilloscope, during the air pump cycle,
I discovered the 1-3 side of O2 sensors were barely responding to
the air pump activation due to carbon build up. The 4-6 side of
O2 sensors were responding correctly. All four O2 sensors were
functioning correctly once running at full engine temperature after
the air pump cycle.
I resolved to design and install a circuit that compensated the two
O2 sensors of the plugged air port side by means of a transistor,
a few resistors, blocking diodes and a time delaying capacitor.
The circuit enabled the O2 sensor output voltage to drop down to proper
voltage when the air pump was activated. The OBDII Hammer scan tool
showed no errors, and I never had a "Check Engine" light.
After a few months, I decided to re-test my theory and disconnected the
mod circuit. The dreaded "Check Engine" light returned.
(Plugged air ports 1-3 again) I reconnected the mod circuit and added
all four O2 sensors to the control circuit. I've driven 6K more miles.
The scan tool has a clean sweep and there's no "Check Engine" light.
At the onset, this indicates the mod circuit is satisfying the
OBDII system.
The circuit does not disrupt the normal O2 sensor operation as it
automatically disconnects from operation once the air pump turns off.
Theoretically, there is a brief air/fuel mix mismatch lasting approximately
10-20 seconds, near the end of the air pump cycle. Although, the air/fuel
mix is always rich at a cold start regardless. There's no noticeable
engine performance difference with or without the control circuit.
I can post the schematic diagram for "educational" purposes.
Although, I have concerns with attempts to copy the circuit that may
result in faulty assembly or installation -thus possibly causing damage!
The circuit briefly lowers the voltage across the O2 sensors during the
air pump cycle. Thus causing the OBDII system to receive O2 sensor
voltages that emulate a fully functioning air injection system.
More time is needed to determine if this is a "permanent" repair.
In addition, this circuit, or similar circuit , should enable removal
of the secondary air pump and hardware for enhancing "performance"
engines. Currently, I'm happy with the results and I have no plans an
tearing down on an engine that uses 1 quart per 2.5K miles and
currently passes very strict emissions standards with flying colors!
Also, plugged air ports may indicate valve guide wear. Usually valve guide
wear is discovered by excessive oil consumption, valve guide noise,
or out of tolerance guide end play. I recommend "Motor Vac" if you have
carbon build up, as it can clean the carbon in many engines.
I invite everyone's feedback.
All the best,
to post my discovery of eight months ago of a way to manipulate the
data sent to the OBDII system for the "Check Engine" light caused
by carbon build up. I want to spend more time testing and confirming
the results. Although, my preliminary tests suggest the results are
conclusive. My car is a '97 C4S with 61K miles.
I have carbon build up blockage of air injection ports of cylinders 1-3.
From what I have determined, the diagnostic is accomplished via the O2
(Oxygen) sensors. The OBDII monitors the output voltage of the Left / Right
banks of O2 sensors during the air pump cycle at cold start up.
If the catalytic converter reaches temperature during the air pump cycle,
the four oxygen sensor voltages drop to almost zero volts between the
O2 sensor output terminals, about two minutes after start up.
If one side of O2 sensors fail to drop in voltage after sufficient time
during the air pump cycle, the non voltage drop is stored
as a fault -giving the resultant "Check Engine" light. This fault is
recorded as blocked air passages, indicating banks 1-3 or 4-6.
Monitoring with an oscilloscope, during the air pump cycle,
I discovered the 1-3 side of O2 sensors were barely responding to
the air pump activation due to carbon build up. The 4-6 side of
O2 sensors were responding correctly. All four O2 sensors were
functioning correctly once running at full engine temperature after
the air pump cycle.
I resolved to design and install a circuit that compensated the two
O2 sensors of the plugged air port side by means of a transistor,
a few resistors, blocking diodes and a time delaying capacitor.
The circuit enabled the O2 sensor output voltage to drop down to proper
voltage when the air pump was activated. The OBDII Hammer scan tool
showed no errors, and I never had a "Check Engine" light.
After a few months, I decided to re-test my theory and disconnected the
mod circuit. The dreaded "Check Engine" light returned.
(Plugged air ports 1-3 again) I reconnected the mod circuit and added
all four O2 sensors to the control circuit. I've driven 6K more miles.
The scan tool has a clean sweep and there's no "Check Engine" light.
At the onset, this indicates the mod circuit is satisfying the
OBDII system.
The circuit does not disrupt the normal O2 sensor operation as it
automatically disconnects from operation once the air pump turns off.
Theoretically, there is a brief air/fuel mix mismatch lasting approximately
10-20 seconds, near the end of the air pump cycle. Although, the air/fuel
mix is always rich at a cold start regardless. There's no noticeable
engine performance difference with or without the control circuit.
I can post the schematic diagram for "educational" purposes.
Although, I have concerns with attempts to copy the circuit that may
result in faulty assembly or installation -thus possibly causing damage!
The circuit briefly lowers the voltage across the O2 sensors during the
air pump cycle. Thus causing the OBDII system to receive O2 sensor
voltages that emulate a fully functioning air injection system.
More time is needed to determine if this is a "permanent" repair.
In addition, this circuit, or similar circuit , should enable removal
of the secondary air pump and hardware for enhancing "performance"
engines. Currently, I'm happy with the results and I have no plans an
tearing down on an engine that uses 1 quart per 2.5K miles and
currently passes very strict emissions standards with flying colors!
Also, plugged air ports may indicate valve guide wear. Usually valve guide
wear is discovered by excessive oil consumption, valve guide noise,
or out of tolerance guide end play. I recommend "Motor Vac" if you have
carbon build up, as it can clean the carbon in many engines.
I invite everyone's feedback.
All the best,
02-02-2003, 09:32 AM
#2
David, You Da Man! Yes Please post the Educational Schematic, I feel there are a number of people who would like to be educated on this one. Very impressive indeed. It appears this could be used for all 3.8 conversions also.
02-02-2003, 11:25 AM
#3
Addict
Lifetime Rennlist
Member
Lifetime Rennlist
Member
David,
Yes, nice work and your explanation makes 100% sense! I assume then that you splice into the O2 sensor circuits? Have you found a convenient spot on the wiring harnesses to do this?
If the sensors and air pump circuits come together on one spot this may make the insertion easier. It would be really neat if there was a connector or spot that you could add something like this without cutting into the wiring harnesses.
Sounds like a good step forward though!
Cheers,
Mike
Yes, nice work and your explanation makes 100% sense! I assume then that you splice into the O2 sensor circuits? Have you found a convenient spot on the wiring harnesses to do this?
If the sensors and air pump circuits come together on one spot this may make the insertion easier. It would be really neat if there was a connector or spot that you could add something like this without cutting into the wiring harnesses.
Sounds like a good step forward though!
Cheers,
Mike
02-02-2003, 12:02 PM
#4
Burning Brakes
David
In earlier thread several rennlisters had problem with cat bypass tubes. CE light came on post 96 cars (pre- and after cat lambdas). may-be You have cure for them also.
In earlier thread several rennlisters had problem with cat bypass tubes. CE light came on post 96 cars (pre- and after cat lambdas). may-be You have cure for them also.
02-02-2003, 03:29 PM
#6
Addict
Rennlist Member
Rennlist Member
Thread Starter
Join Date: Apr 2002
Location: Newcastle, WA
Posts: 195
Likes: 0
Received 0 Likes
on
0 Posts
I tried to post it, but the posting function is
only for registered Rennlisters. I'll try registering.
At present, I can be emailed privately if you need information.
All the best,
only for registered Rennlisters. I'll try registering.
At present, I can be emailed privately if you need information.
All the best,
02-02-2003, 11:54 PM
#7
Pro
Join Date: Oct 2002
Location: Tulsa, OK
Posts: 637
Likes: 0
Received 0 Likes
on
0 Posts
This is great work from David, and sounds like a permanent fix.
Reading his method, I have another idea that might accomplish the same thing without electronics - though I don't know how easy it would be to actually implement.
It looks like the DME expects the O2 sensors to go to low voltage during the air pump cycle. This is what the O2 sensor does when the mixture is lean, i.e. excess of oxygen - and what we expect to happen when the air pump is adding lots of air to the exhaust stream.
The alternative solution might be to route the air pump output to a point in the exhaust system ahead of the 02 sensors, and to a location more easily serviced in the event of clogging than in the heads. It might not work as well to actually light off the catalytic convertors, since mixing may not be as thorough, but it should be sufficient to make the O2 sensors think there is an excess of oxygen, which is what they are looking for during that air pump cycle.
I still think David's electronic solution is more elegant and the superior solution, but the less electrically inclined might consider rerouting air lines more to their taste.
Chip
Reading his method, I have another idea that might accomplish the same thing without electronics - though I don't know how easy it would be to actually implement.
It looks like the DME expects the O2 sensors to go to low voltage during the air pump cycle. This is what the O2 sensor does when the mixture is lean, i.e. excess of oxygen - and what we expect to happen when the air pump is adding lots of air to the exhaust stream.
The alternative solution might be to route the air pump output to a point in the exhaust system ahead of the 02 sensors, and to a location more easily serviced in the event of clogging than in the heads. It might not work as well to actually light off the catalytic convertors, since mixing may not be as thorough, but it should be sufficient to make the O2 sensors think there is an excess of oxygen, which is what they are looking for during that air pump cycle.
I still think David's electronic solution is more elegant and the superior solution, but the less electrically inclined might consider rerouting air lines more to their taste.
Chip
Trending Topics
02-03-2003, 12:08 AM
#8
Pro
Join Date: Oct 2002
Location: Tulsa, OK
Posts: 637
Likes: 0
Received 0 Likes
on
0 Posts
David also mentions that his circuit might cause an air/fuel mismatch near the end of the air pump cycle, but I am thinking that it wouldn't do so at all.
When the engine is running during the air pump cycle, it is almost certainly using a fixed air/fuel map and not using the oxygen sensor data at all to adjust air fuel mixture. That being the case, the circuit should have absolutely no effect on how the engine itself is managed.
If I'm missing something here, chime in.
Chip
When the engine is running during the air pump cycle, it is almost certainly using a fixed air/fuel map and not using the oxygen sensor data at all to adjust air fuel mixture. That being the case, the circuit should have absolutely no effect on how the engine itself is managed.
If I'm missing something here, chime in.
Chip
02-03-2003, 05:34 AM
#9
Addict
Rennlist Member
Rennlist Member
Join Date: May 2001
Location: Portland Oregon
Posts: 311
Likes: 0
Received 0 Likes
on
0 Posts
David,
Email it to me and I will post it on the <a href="http://www.993tt.com" target="_blank">www.993tt.com</a> site, if you would like.
For educational purposes only!
Rick
Rick_NW@msn.com
Email it to me and I will post it on the <a href="http://www.993tt.com" target="_blank">www.993tt.com</a> site, if you would like.
For educational purposes only!
Rick
Rick_NW@msn.com
02-03-2003, 01:42 PM
#10
David,
This is excellent information. It sure blows my guess that there was a sensor detecting air flow out of the water! I’m thinking that this may explain why the ’95 seem immune to this problem. My theory is that OBD1 does not look at the output of the O2 sensors during air pump operation. It’s well known that OBD1 does not monitor as much as OBD2. The air ports and passages in the ‘95’s are just as clogged as the other years, but it just doesn’t get detected. What are your thoughts?
Tom
’95 993
This is excellent information. It sure blows my guess that there was a sensor detecting air flow out of the water! I’m thinking that this may explain why the ’95 seem immune to this problem. My theory is that OBD1 does not look at the output of the O2 sensors during air pump operation. It’s well known that OBD1 does not monitor as much as OBD2. The air ports and passages in the ‘95’s are just as clogged as the other years, but it just doesn’t get detected. What are your thoughts?
Tom
’95 993
02-03-2003, 04:30 PM
#11
Pro
Join Date: Jan 2002
Location: Manhattan Beach, CA
Posts: 653
Likes: 0
Received 0 Likes
on
0 Posts
David,
This is great news. I will look at this procedure closely as I would much prefer this approach, than pulling the engine apart <img border="0" alt="[jumper]" title="" src="graemlins/jumper.gif" />
It does not sound like there should be any consequences to this procedure as this air pump is an added device in the US.
If this works, I think that alot of folks in the future with the clogged ports will be indebted to you...myself included <img border="0" alt="[thumbsup]" title="" src="graemlins/bigok.gif" />
This is great news. I will look at this procedure closely as I would much prefer this approach, than pulling the engine apart <img border="0" alt="[jumper]" title="" src="graemlins/jumper.gif" />
It does not sound like there should be any consequences to this procedure as this air pump is an added device in the US.
If this works, I think that alot of folks in the future with the clogged ports will be indebted to you...myself included <img border="0" alt="[thumbsup]" title="" src="graemlins/bigok.gif" />
02-04-2003, 02:42 PM
#12
Rennlist Member
David,
Bravo! Excellent, fascinating post.
I especially admire folks who can cobble analog circuits.
I have a comment on why the O-sensors are behaving in the way you describe.
I believe what you are seeing on the oscilloscope is an output signal of intermediate voltage that the O-sensors emit while they are warming up, followed a low voltage output indicating an oxygen rich environment. In other words, that voltage dip you see just before the 2 minutes expire, is the O-sensors waking up and saying "I see extra oxygen". With the airpump going and the injection ports clear, the O-sensors should see extra oxygen. (The O-sensor output voltage is inversely related to oxygen content, so: lower voltage => more oxygen; and higher voltage => less oxygen. Which explains why fuel consumption increases when O-sensors fail and emit lower or no voltage).
For those who are curious, here is a <a href="http://www.wps.com/LPG/o2sensor.html" target="_blank">link</a> to a nicely written FAQ on oxygen sensors.
The other subtlety is that at least one of the O-sensors on each bank, and maybe both, are located upstream of the catalytic converter and therefore cannot know whether the cat is hot or cold. At least one O-sensor must be located upstream of the cat, because an operating cat scrubs the exhaust clean, thereby removing information that the engine controller needs to adjust the fuel mixture. Since each bank has two O-sensors, I suppose one could be located upstream and one downstream of the cat.
By the way, anyone know a) why 993s have 4 O-sensors, and b) where they are located on each bank?
Again, thanks for the excellent post!
-Dave
'90 C2
[P.S. You could add "For Off-Road Use Only" to your disclaimer.]
Bravo! Excellent, fascinating post.
I especially admire folks who can cobble analog circuits.
I have a comment on why the O-sensors are behaving in the way you describe.
I believe what you are seeing on the oscilloscope is an output signal of intermediate voltage that the O-sensors emit while they are warming up, followed a low voltage output indicating an oxygen rich environment. In other words, that voltage dip you see just before the 2 minutes expire, is the O-sensors waking up and saying "I see extra oxygen". With the airpump going and the injection ports clear, the O-sensors should see extra oxygen. (The O-sensor output voltage is inversely related to oxygen content, so: lower voltage => more oxygen; and higher voltage => less oxygen. Which explains why fuel consumption increases when O-sensors fail and emit lower or no voltage).
For those who are curious, here is a <a href="http://www.wps.com/LPG/o2sensor.html" target="_blank">link</a> to a nicely written FAQ on oxygen sensors.
The other subtlety is that at least one of the O-sensors on each bank, and maybe both, are located upstream of the catalytic converter and therefore cannot know whether the cat is hot or cold. At least one O-sensor must be located upstream of the cat, because an operating cat scrubs the exhaust clean, thereby removing information that the engine controller needs to adjust the fuel mixture. Since each bank has two O-sensors, I suppose one could be located upstream and one downstream of the cat.
By the way, anyone know a) why 993s have 4 O-sensors, and b) where they are located on each bank?
Again, thanks for the excellent post!
-Dave
'90 C2
[P.S. You could add "For Off-Road Use Only" to your disclaimer.]
02-04-2003, 04:01 PM
#13
Rennlist Member
With the off-line help of another forum member I found answers to my questions in the previous post.
OBDII apparently uses an O-sensor downstream of the cat, to monitor health of the cat. So it appears that each bank has an O-sensor upstream of the cat, and another downstream of the cat.
OBDII apparently uses an O-sensor downstream of the cat, to monitor health of the cat. So it appears that each bank has an O-sensor upstream of the cat, and another downstream of the cat.
04-03-2003, 03:50 PM
#15
Instructor
96P993 I got the Ckt from him and some mods to it I've been very busy lately with work deadlines but I'll come up with some pages and pics once I get to it. The proximity of the sensor connectors and the air pump relay makes it look very promising. My mechanic alllowed me to see the electric diagrams for double checking the location of the connectors and stuff.