Tune for E85/emissions issues
11-11-2010, 01:19 PM
#22
Le Mans Master
Thread Starter
11-11-2010, 03:18 PM
#23
Le Mans Master
Here's a post from blue86iroc on thirdgen.org. I did this on my 84 Xfire when I installed a single pattern roller cam with 53d overlap. Manipulating the O2 sensor swing points along with tuning the proportional gains makes big cams manageable.
OTOH, you could try putting it into open loop and tuning by nose because a WB will see the same "false lean" the NB does.
From blue86iroc on www.thirdgen.org THE ILLUSTRATIONS HE REFERS TO ARE MISSING DUE TO MISSING LINKS.
"If you don't feel like reading everything, I'll briefly explain the gist of the situation.
In my tuning, I've experienced all of the above symptoms associated with idle since my motor rebuild. I'm running a ZZ4 cam with fifty degrees of overlap and ported 305 "601" heads which bumps the compression ratio to about 10.2:1 on my L98. Through experimentation and lots of datalogging, I determined that many of the unusual idling experiences were due to a heavily rich situation -- in other words, the injector pulse width was just too long for the operational condition of the engine at that point in time. Having a cam with more overlap than factory is a key contributor to this, and is probably one of the major reasons that tuning idle can be such a pain in the butt.
Cam overlap is the period of time, measured in crankshaft degrees, in which both the intake and exhaust valves are open. At low RPM, this allows some of the fresh intake charge (i.e. unburned fuel and air) to leave the combustion chamber via the exhaust. It can cause a car to smell like fuel when idling, depending on how much overlap there is.
Now, keep in mind that there is an oxygen sensor that measures the amount of oxygen present in the exhaust stream. When it detects the unburned intake charge from the high-overlap cam, it tells the ECM that there is a lean condition (assuming closed loop is enabled). The Integrator, INT, then changes to compensate for this "phantom lean condition." As the INT moves, it updates the BLM, essentially dumping extra, unneeded fuel into the engine because the ECM thinks it's needed. Run like this for a little while, let the learn feature kick in, and pretty soon your car will smell like a gas station AND it'll cause all sorts of unusual characteristics at idle. Observe the following datalogs:
RPM drops suddenly after falling from a rev at idle
In this example, the throttle is given a quick jab and released immediately. As the RPM begins to settle, note the BLM cell changes back to 0 (idle cell). Since it's already too rich (even though the ECM thinks it's lean), additional fuel is added and the engine literally begins to drown. The plot shows it very well -- RPMs decreasing and the injector PW increasing.
If the RPM gets low enough, the IAC opens to prevent the motor from stalling; this causes the extra fuel in the combustion chamber to react quickly with fresh oxygen, dramatically raising the RPM. You can see how this can lead to a surging condition, as well.
Immediate stall upon engine startup
This case is very similar to the previous one, in that the trouble starts once a "falsely rich" BLM cell is activated, though it occurs at engine startup. After the initial RPM spike, the engine tries to settle to the IAC-commanded idle speed... but that darn BLM cell 0 rears its ugly head again, sending the engine into a richness-induced stall. The injector pulse width reaches amusing values as RPM decreases. Part of this is because the load variable, LV8, is inversely proportional to RPM; changes in LV8 affect the open loop AFR ("Open Loop AFR % Change vs. LV8" in $6E), in this case reducing the AFR, further dumping fuel into the engine. Eisenhower's "Domino Theory" as applied to GM ECMs, haha.
So, how can you improve the situation? Well, there obviously has to be a reduction of fuel in BLM cell 0... but if the cam overlap confuses the oxygen sensor, how do you decide upon the right amount of fuel without melting the engine? From what I've read and done, the four main ways of doing this are by exhaust smell, header/manifold temperature, plug reading, and "engine happiness." A WB O2 wouldn't do much good in this situation, because it would read the extra oxygen from the overlap just like the factory NB O2.
Tuning via smell is fairly straightforward, but I'd imagine that with larger cams, there will be a certain amount of odor that you won't be able to get rid of. Just don't go so lean that the headers start glowing... Now, from my understanding, headers can get super-hot by either being lean OR rich. A very lean mixture will simply cause higher combustion temperatures, raising the temperature of the headers. Being overly rich can actually cause the unburned intake charge (from cam overlap, etc.) to ignite within the header tubes. I haven't personally experienced either of those situations, but I could see how it would make for a contradictory tune -- but take into account how much overlap you're actually running, I suppose, to figure out if excessive header temperatures are lean- or rich-based.
Member 87_TA comments on header temperature from the "cam tuning question" thread above:
Quote:
A near stock type cam you will see 400 + degree temps at idle. A performance cam with healthy O/L [overlap] will see
300 and below because of the Unburned air cooling the charge as it gets pulled from intake tract.
Plug reading is simple, too. Let it idle for a bit, shut it down, and pull a few plugs. If they're really black, you know you're too rich. White is lean; tan is good. See this page if you need some visual references. Lastly, "engine happiness". Grumpy and numerous others have used this term, and it's quite true -- tune for what the engine likes best based upon its response. For example, before I really understood that I had a fueling issue, I reset the ECM (BLMs to 128), disabled closed loop, and pulled out a bit of fuel at idle. The engine could be revved without stalling or stumbling, so I acknowledged that as an improvement and one more step towards making the engine happy. When you can't reliably depend on an oxygen sensor it helps to think this way. I ended up with about 1.90-2.15 ms pulse widths at 700 RPM on a warm engine, but YMMV.
Alright, so you've gotten the fuel the way you want it by changing the MAF tables, VE entries, etc. We now come to the most important thing for maintaining a stable idle in closed loop: preventing the O2 sensor from thinking the engine is lean when it's really not. If you ignore this, you might see your BLMs doing something like the following:
We can do this by manipulating the "ideal" AFR that the oxygen sensor expects, in millivolts, to compensate for cam overlap. Again, I'm going to provide a basic explanation of how this works, but for the big story, read the aforementioned threads.
The typical narrow-band oxygen sensor is set to read a certain AFR at a certain median voltage (usually 0.450 V). The ECM maintains this stoichiometric AFR by constantly "switching" between rich and lean states, equalizing the amount of time spent in each state. Each transition from one state to another is called a "cross count."
At idle, with lots of cam overlap, a value less than the median voltage will be reported by the oxygen sensor. The ECM will then increase the fuel until the median voltage is reached, and it will switch states to maintain this voltage. By lowering the median voltage one can account for the extra oxygen and reduce the amount of correction that the ECM employs. This is the same way the ECM operates the AIR pump and diverters; when air is being injected into the exhaust stream, the oxygen sensor threshold is reduced by 0.100 V. Note: If you have removed your AIR pump and diverter solenoids, ensure that you disable its operation the EPROM. Read Removing Smog pump, AIR: $8D 730 modifications for more information.
The nice thing about the oxygen sensor threshold in $6E is that there are different settings for different MAF flow rates... in other words, you can tweak the idle and let part throttle alone. I'm not sure of the tables on a '730, but I know they're there.
I determined the values for the table by looking at what the oxygen sensor was reading as soon as the engine entered closed loop, before the INT/BLM entries had a chance to change. I was seeing about 0.230-0.250 V at 8.xx g/sec, so I set the two lowest table entries to 0.260 V. Only do this if you're SURE you're not running too lean with the fuel settings you've specified earlier.
I'm now in the mid-130s in terms of BLM, and the car idles much better than it ever did. The surging is gone, it doesn't stumble when revved, it starts on the first try, and it the smell of fuel is greatly minimized. The plugs look fairly good, too.
I'm greatly thankful for all of the helpful people on these forums and the wealth of knowledge we've amassed as a group. I hope that this helps some people tune the idle on their cammed engines... it's a world of difference. If I've goofed in my explanation or turned anything upside-down, I'd appreciate any criticism.
Even with all of the fueling "logic" explained, I've still experienced a few anomalies that I don't understand -- maybe someone has something to offer. Before I got my BLMs to settle down by adjusting the O2 threshold, I let the engine idle as the BLMs ascended to 160 (closed loop, of course). Unusually, the injector pulse width did not change to reflect the added fueling... only when I placed the transmission in Drive did the PW increase by about 0.30 ms. Shifting back into Park or Neutral brought the PW back down, but the BLM stayed the same. I thought that the fuel correction was accounted for at all times in closed loop. Can anyone shed some light on this?
__________________
1986 Camaro IROC-Z
350 TPI V8
14.10 seconds @ 96.96 mph
--------------------------------------------------------------------------------
Last edited by RBob; 04-29-2008 at 08:42 AM. Reason: edited for blue86iroc, update picture links
OTOH, you could try putting it into open loop and tuning by nose because a WB will see the same "false lean" the NB does.
From blue86iroc on www.thirdgen.org THE ILLUSTRATIONS HE REFERS TO ARE MISSING DUE TO MISSING LINKS.
"If you don't feel like reading everything, I'll briefly explain the gist of the situation.
In my tuning, I've experienced all of the above symptoms associated with idle since my motor rebuild. I'm running a ZZ4 cam with fifty degrees of overlap and ported 305 "601" heads which bumps the compression ratio to about 10.2:1 on my L98. Through experimentation and lots of datalogging, I determined that many of the unusual idling experiences were due to a heavily rich situation -- in other words, the injector pulse width was just too long for the operational condition of the engine at that point in time. Having a cam with more overlap than factory is a key contributor to this, and is probably one of the major reasons that tuning idle can be such a pain in the butt.
Cam overlap is the period of time, measured in crankshaft degrees, in which both the intake and exhaust valves are open. At low RPM, this allows some of the fresh intake charge (i.e. unburned fuel and air) to leave the combustion chamber via the exhaust. It can cause a car to smell like fuel when idling, depending on how much overlap there is.
Now, keep in mind that there is an oxygen sensor that measures the amount of oxygen present in the exhaust stream. When it detects the unburned intake charge from the high-overlap cam, it tells the ECM that there is a lean condition (assuming closed loop is enabled). The Integrator, INT, then changes to compensate for this "phantom lean condition." As the INT moves, it updates the BLM, essentially dumping extra, unneeded fuel into the engine because the ECM thinks it's needed. Run like this for a little while, let the learn feature kick in, and pretty soon your car will smell like a gas station AND it'll cause all sorts of unusual characteristics at idle. Observe the following datalogs:
RPM drops suddenly after falling from a rev at idle
In this example, the throttle is given a quick jab and released immediately. As the RPM begins to settle, note the BLM cell changes back to 0 (idle cell). Since it's already too rich (even though the ECM thinks it's lean), additional fuel is added and the engine literally begins to drown. The plot shows it very well -- RPMs decreasing and the injector PW increasing.
If the RPM gets low enough, the IAC opens to prevent the motor from stalling; this causes the extra fuel in the combustion chamber to react quickly with fresh oxygen, dramatically raising the RPM. You can see how this can lead to a surging condition, as well.
Immediate stall upon engine startup
This case is very similar to the previous one, in that the trouble starts once a "falsely rich" BLM cell is activated, though it occurs at engine startup. After the initial RPM spike, the engine tries to settle to the IAC-commanded idle speed... but that darn BLM cell 0 rears its ugly head again, sending the engine into a richness-induced stall. The injector pulse width reaches amusing values as RPM decreases. Part of this is because the load variable, LV8, is inversely proportional to RPM; changes in LV8 affect the open loop AFR ("Open Loop AFR % Change vs. LV8" in $6E), in this case reducing the AFR, further dumping fuel into the engine. Eisenhower's "Domino Theory" as applied to GM ECMs, haha.
So, how can you improve the situation? Well, there obviously has to be a reduction of fuel in BLM cell 0... but if the cam overlap confuses the oxygen sensor, how do you decide upon the right amount of fuel without melting the engine? From what I've read and done, the four main ways of doing this are by exhaust smell, header/manifold temperature, plug reading, and "engine happiness." A WB O2 wouldn't do much good in this situation, because it would read the extra oxygen from the overlap just like the factory NB O2.
Tuning via smell is fairly straightforward, but I'd imagine that with larger cams, there will be a certain amount of odor that you won't be able to get rid of. Just don't go so lean that the headers start glowing... Now, from my understanding, headers can get super-hot by either being lean OR rich. A very lean mixture will simply cause higher combustion temperatures, raising the temperature of the headers. Being overly rich can actually cause the unburned intake charge (from cam overlap, etc.) to ignite within the header tubes. I haven't personally experienced either of those situations, but I could see how it would make for a contradictory tune -- but take into account how much overlap you're actually running, I suppose, to figure out if excessive header temperatures are lean- or rich-based.
Member 87_TA comments on header temperature from the "cam tuning question" thread above:
Quote:
A near stock type cam you will see 400 + degree temps at idle. A performance cam with healthy O/L [overlap] will see
300 and below because of the Unburned air cooling the charge as it gets pulled from intake tract.
Plug reading is simple, too. Let it idle for a bit, shut it down, and pull a few plugs. If they're really black, you know you're too rich. White is lean; tan is good. See this page if you need some visual references. Lastly, "engine happiness". Grumpy and numerous others have used this term, and it's quite true -- tune for what the engine likes best based upon its response. For example, before I really understood that I had a fueling issue, I reset the ECM (BLMs to 128), disabled closed loop, and pulled out a bit of fuel at idle. The engine could be revved without stalling or stumbling, so I acknowledged that as an improvement and one more step towards making the engine happy. When you can't reliably depend on an oxygen sensor it helps to think this way. I ended up with about 1.90-2.15 ms pulse widths at 700 RPM on a warm engine, but YMMV.
Alright, so you've gotten the fuel the way you want it by changing the MAF tables, VE entries, etc. We now come to the most important thing for maintaining a stable idle in closed loop: preventing the O2 sensor from thinking the engine is lean when it's really not. If you ignore this, you might see your BLMs doing something like the following:
We can do this by manipulating the "ideal" AFR that the oxygen sensor expects, in millivolts, to compensate for cam overlap. Again, I'm going to provide a basic explanation of how this works, but for the big story, read the aforementioned threads.
The typical narrow-band oxygen sensor is set to read a certain AFR at a certain median voltage (usually 0.450 V). The ECM maintains this stoichiometric AFR by constantly "switching" between rich and lean states, equalizing the amount of time spent in each state. Each transition from one state to another is called a "cross count."
At idle, with lots of cam overlap, a value less than the median voltage will be reported by the oxygen sensor. The ECM will then increase the fuel until the median voltage is reached, and it will switch states to maintain this voltage. By lowering the median voltage one can account for the extra oxygen and reduce the amount of correction that the ECM employs. This is the same way the ECM operates the AIR pump and diverters; when air is being injected into the exhaust stream, the oxygen sensor threshold is reduced by 0.100 V. Note: If you have removed your AIR pump and diverter solenoids, ensure that you disable its operation the EPROM. Read Removing Smog pump, AIR: $8D 730 modifications for more information.
The nice thing about the oxygen sensor threshold in $6E is that there are different settings for different MAF flow rates... in other words, you can tweak the idle and let part throttle alone. I'm not sure of the tables on a '730, but I know they're there.
I determined the values for the table by looking at what the oxygen sensor was reading as soon as the engine entered closed loop, before the INT/BLM entries had a chance to change. I was seeing about 0.230-0.250 V at 8.xx g/sec, so I set the two lowest table entries to 0.260 V. Only do this if you're SURE you're not running too lean with the fuel settings you've specified earlier.
I'm now in the mid-130s in terms of BLM, and the car idles much better than it ever did. The surging is gone, it doesn't stumble when revved, it starts on the first try, and it the smell of fuel is greatly minimized. The plugs look fairly good, too.
I'm greatly thankful for all of the helpful people on these forums and the wealth of knowledge we've amassed as a group. I hope that this helps some people tune the idle on their cammed engines... it's a world of difference. If I've goofed in my explanation or turned anything upside-down, I'd appreciate any criticism.
Even with all of the fueling "logic" explained, I've still experienced a few anomalies that I don't understand -- maybe someone has something to offer. Before I got my BLMs to settle down by adjusting the O2 threshold, I let the engine idle as the BLMs ascended to 160 (closed loop, of course). Unusually, the injector pulse width did not change to reflect the added fueling... only when I placed the transmission in Drive did the PW increase by about 0.30 ms. Shifting back into Park or Neutral brought the PW back down, but the BLM stayed the same. I thought that the fuel correction was accounted for at all times in closed loop. Can anyone shed some light on this?
__________________
1986 Camaro IROC-Z
350 TPI V8
14.10 seconds @ 96.96 mph
--------------------------------------------------------------------------------
Last edited by RBob; 04-29-2008 at 08:42 AM. Reason: edited for blue86iroc, update picture links
11-12-2010, 05:53 PM
#24
Le Mans Master
Thread Starter
great info, I think in my situation getting an open loop chip to keep the o2s from adding fuel at idle would be great
I need to find somewhere that has a gas analyzer and can setup the tune to get it to pass closed loop
not many of those around and all the good tuners don't seem to have much focus on emissions.
passing with this combo will be tough, fresh cats are a start, but I don't want to spend a mint doing this either
anyone can tune in VA?? I can likely line up an emissions station to help as I know a few gas station guys that would like to help
thanks!
I need to find somewhere that has a gas analyzer and can setup the tune to get it to pass closed loop
not many of those around and all the good tuners don't seem to have much focus on emissions.
passing with this combo will be tough, fresh cats are a start, but I don't want to spend a mint doing this either
anyone can tune in VA?? I can likely line up an emissions station to help as I know a few gas station guys that would like to help
thanks!
11-12-2010, 06:15 PM
#25
Le Mans Master
great info, I think in my situation getting an open loop chip to keep the o2s from adding fuel at idle would be great
I need to find somewhere that has a gas analyzer and can setup the tune to get it to pass closed loop
not many of those around and all the good tuners don't seem to have much focus on emissions.
passing with this combo will be tough, fresh cats are a start, but I don't want to spend a mint doing this either
anyone can tune in VA?? I can likely line up an emissions station to help as I know a few gas station guys that would like to help
thanks!
I need to find somewhere that has a gas analyzer and can setup the tune to get it to pass closed loop
not many of those around and all the good tuners don't seem to have much focus on emissions.
passing with this combo will be tough, fresh cats are a start, but I don't want to spend a mint doing this either
anyone can tune in VA?? I can likely line up an emissions station to help as I know a few gas station guys that would like to help
thanks!
Actually, if you read the post closely you'll note that tuning in open loop using a WB won't help because the WB will read the same "false lean" as the Narrow Band is reading. The trick is to change the O2 swing points which tell the ECM where the "sweet spot" is. This can be done separately for idle vs cruise vs WOT conditions. In most cases, modifying idle is enough.
11-12-2010, 07:24 PM
#26
Le Mans Master
Thread Starter
Ron,
Actually, if you read the post closely you'll note that tuning in open loop using a WB won't help because the WB will read the same "false lean" as the Narrow Band is reading. The trick is to change the O2 swing points which tell the ECM where the "sweet spot" is. This can be done separately for idle vs cruise vs WOT conditions. In most cases, modifying idle is enough.
Actually, if you read the post closely you'll note that tuning in open loop using a WB won't help because the WB will read the same "false lean" as the Narrow Band is reading. The trick is to change the O2 swing points which tell the ECM where the "sweet spot" is. This can be done separately for idle vs cruise vs WOT conditions. In most cases, modifying idle is enough.
but it seems simpler to just tune it to meet the emissions standards closed loop with the gas analyzer to eliminate the 02 altogether
that way I could just tune for best emissions ignoring the 02 altogether
maybe the other way is easier though.
either way, I don't have the knowledge or equipment to tune, so my earlier question to get someone in VA that can help is next steps
,makes sense to me to "tune on the dyno" with the gas analyzer running and "tune it to pass" either way so I can validate the test will pass as part of the emissions test
I think I have a station lined up that might allow me to do that
any takers?
11-12-2010, 09:06 PM
#27
Le Mans Master
Ron,
I don't think you can tune in Closed Loop while ignoring the O2 sensor. Closed Loop is all about the feedback loop with the O2 sensor.
Do you have a scantool that can datalog?
I don't think you can tune in Closed Loop while ignoring the O2 sensor. Closed Loop is all about the feedback loop with the O2 sensor.
Do you have a scantool that can datalog?
11-13-2010, 08:45 AM
#28
Le Mans Master
Thread Starter
sorry, I got that backwards, I meant tune with an open loop and just set it up for what you want.
I do have ease but I need to get the license renewed.
and I don't have any chip burning or tuning skills.
11-13-2010, 10:20 AM
#30
Le Mans Master
Thread Starter
I don't have any laptops with good working serial ports, everything is USB now
I think there are usb to serial adapters, and that would probably work.
I haven't used it in a really long time.
I bought it a while back to do traces and thought about getting more into tuning, but never pursued it.
thanks for all your help.
11-13-2010, 10:53 AM
#31
Le Mans Master
the laptop I had it running on went bad (really old) and the floppy disk with the license key can't be read anymore (I bought ease probably 9 years ago)
I don't have any laptops with good working serial ports, everything is USB now
I think there are usb to serial adapters, and that would probably work.
I haven't used it in a really long time.
I bought it a while back to do traces and thought about getting more into tuning, but never pursued it.
thanks for all your help.
I don't have any laptops with good working serial ports, everything is USB now
I think there are usb to serial adapters, and that would probably work.
I haven't used it in a really long time.
I bought it a while back to do traces and thought about getting more into tuning, but never pursued it.
thanks for all your help.
I could probably send you my cable w/serial2USB conector. Works fine.
Bought it at Tiger Direct or CompUSA. Also could send you my floppy with the key. =Jeff= also has copy of EASE.
11-15-2010, 08:09 PM
#32
Le Mans Master
Thread Starter
thanks dominic.
I'll see if I can get some traces going. I want to try some fresh cats on it and see how far off it is. definitely stinks up the joint and has the rich smell you are describing.
I have the 3" cats, and looking for the best cats to clean up the HC.
don't really want to mess around with stock stuff, my current bullet cats are slip jointed and I can just drop the exh
any recommendations for a decent "clean" aftermarket cat?
thinking of the magnaflows
might be interesting to "dual cat" the car too if the second set could get hot enough to still work. that might help
any thoughts?
I'll see if I can get some traces going. I want to try some fresh cats on it and see how far off it is. definitely stinks up the joint and has the rich smell you are describing.
I have the 3" cats, and looking for the best cats to clean up the HC.
don't really want to mess around with stock stuff, my current bullet cats are slip jointed and I can just drop the exh
any recommendations for a decent "clean" aftermarket cat?
thinking of the magnaflows
might be interesting to "dual cat" the car too if the second set could get hot enough to still work. that might help
any thoughts?
11-15-2010, 10:46 PM
#34
Team Owner
thanks dominic.
I'll see if I can get some traces going. I want to try some fresh cats on it and see how far off it is. definitely stinks up the joint and has the rich smell you are describing.
I have the 3" cats, and looking for the best cats to clean up the HC.
don't really want to mess around with stock stuff, my current bullet cats are slip jointed and I can just drop the exh
any recommendations for a decent "clean" aftermarket cat?
thinking of the magnaflows
might be interesting to "dual cat" the car too if the second set could get hot enough to still work. that might help
any thoughts?
I'll see if I can get some traces going. I want to try some fresh cats on it and see how far off it is. definitely stinks up the joint and has the rich smell you are describing.
I have the 3" cats, and looking for the best cats to clean up the HC.
don't really want to mess around with stock stuff, my current bullet cats are slip jointed and I can just drop the exh
any recommendations for a decent "clean" aftermarket cat?
thinking of the magnaflows
might be interesting to "dual cat" the car too if the second set could get hot enough to still work. that might help
any thoughts?
11-16-2010, 06:44 PM
#35
Le Mans Master
Thread Starter
Ron - if you want some of the best cats in the business, get yourself some stock C5 cats and put them in line - these are some of the finest cats put on any car. Think about where they are located on the C5 - way back away from the engine, yet still super efficient. There should be lots of them all over the place. Call LPE - I left them mine when I did the TT upgrade!
if I can get them to fit behind the headers, that would work well
used stock parts would be a good option, and likely give me a fighting chance
thanks for the suggestion
11-16-2010, 07:02 PM
#36
Drifting
Good stock cats designed for "downstream" use would seem like a good bet - either right off the header or inline using the bullets cats as pre-cats.
