[Gmecm] Failing Calif SMOG
Aaron
aaron
Tue Sep 13 19:47:05 UTC 2011
I was so with you until you got to the pull 12 volts from the distributor.
Based on my antidotal experience with mid to late 70?s big block Olds with HEI Distributors, I consider that a EVERY poor choice.
The Olds 455 spark plug were originally gapped at 0.080? . Please trust me when I say, that 14 gauge wire running 10 feet away to an ignition switch, needed all the amperage it could get to fire those plugs.
Then there is my experiences with vehicle owners that use those @#$* %(@#^$ 3M Scotchlocks. OH HOW I HATE THOSE #*&$^ *#&$^ @ THINGS !!!!
I stalled a 26 foot motorhome in the middle lane of a five lane freeway at rush hour while test driving new wheel bearings and brakes. Imagine my surprise to find not one but three of those little blue monsters on the battery wire to the HEI.
But I digress into my hatred of Scotchlocks.
On a test vehicle, 78 GMC Motorhome with and OLDS 455 and long tri-y headers, we ran two HO2s. One at each collector. The both sensor wires were then run to a switch on the dash. When one bank was running the 7747, the other would run an LED gauge. Fun for testing and driving and comparing HO2 signals.
Please keep the stories coming Mr. Johnson.
I?m learning a lot from your hands on lessons.
From: gmecm-bounces at diy-efi.org [mailto:gmecm-bounces at diy-efi.org] On Behalf Of Command Performance Engineering
Sent: Monday, September 12, 2011 11:56 AM
To: 'Roger and Carol'; 'PERFormance DYNamics'; gmecm at diy-efi.org
Subject: Re: [Gmecm] Failing Calif SMOG
Just a quick comment on the location and type of o2 sensors that you mentioned in the end of your email. The 4L60E transmission was introduced in 1993 in the ? and light duty ? ton trucks, and with it the higher speed CPU and baud rate output of the PCM, but they continued to use a single wire o2 sensor located in the exhaust manifold on one side (typically drivers side). (The 4L80E was introduced in 1991, again with a faster processor and baud rate output, with a single wire o2 Sensor, but that?s another story.) Depending on the models and engines, GM switched over to the 3-wire Heated o2 Sensor (HO2S) in 1994 or 1995 for most of the TBI truck engines when they moved the o2 sensor back to the crossover area. This was prompted by tighter air quality standards beginning in 1994. By allowing the PCM to sample both exhaust branches, the HO2S could handle the total exhaust picture far better, resulting in cleaner air and lower emissions.
When converting an older ECM (back to 1987) to use a newer HO2S located in the crossover or when headers are used, I use GM P/N 19178938 (A/C Delco # AFS76) with its matching pigtail connector GM# 12117025 (ACD# PT420). (This HO2S is for the 1995 TBI trucks, but can be used successfully in all o2 Sensor retrofit applications.) Pins A & B have black colored wires which go to the heating element. They are not polarity sensitive and therefore one wire connects to an engine block or transmission bolt location for a good ground and the other lead should run through a 5-Amp fuse to a +12 volt ignition source, most easily found at the distributor. Pin-C is going to have a colored wire, usually either purple or blue, and that connects to the original purple wire for the o2 Signal going back to the ECM/PCM.
When we have needed to capture the exhaust stream output from both banks on older TBI and port injected engines, we have retrofitted these HO2S?s into older trucks, Caprices, Camaros, and Corvettes when relocating the o2 rearward, in some cases up to an additional three feet, with excellent results. The ECM snaps into Closed Loop quickly and stays there, because the heating element keeps the o2 sensor hot from the moment of ?key on.? The exhaust stream cools down somewhat after the exhaust manifold, thereby necessitating a HO2S when the exhaust is sampled farther back.
It might help this vehicle pass emissions better only by virtue of the computer?s ability to sense both banks of exhaust rather than only one, and thereby try to compensate for an errant BLM/INT situation that could be present in the other bank.
David Johnson
From: gmecm-bounces at diy-efi.org [mailto:gmecm-bounces at diy-efi.org] On Behalf Of Roger and Carol
Sent: Sunday, September 04, 2011 12:41 PM
To: PERFormance DYNamics; gmecm at diy-efi.org
Subject: Re: [Gmecm] Failing Calif SMOG
Thanks for the comeback;
I?ve completed your table?hope that helps. I' am aware that the more data supplied, the better the discourse. Should have included every thing my first post.
idle readings : Std's ~2500 RPM Std's
RPM 649 2635
CO . 03 % 1.00% 2.06 % 1.00 %
CO2 14.5% 13.3%
HC 21 ppm 120 ppm 45 ppm 180 ppm
O2 .3% .3%
NOx na na (no load tests)
I agree with all you?ve said, but would add that the BLM?s are the result of ECM base-line fuel mapping, plus INT adjustments for variances from optimum fuel control caused by either base-line fuel mapping or current changes in operating parameters. If they (BLM?s) are steady, it indicates the INT?s are averaging 128; mine were varying betwee 125 to 129, idle to 3000 RPM (2008 log). Latest log (Sept 2011) shows INT?s varying between 126.5 and 129.8, over the same RPM range. The conditions for both loggings were the same, and similar to the SMOG test runs---no load, two-idle speeds, sitting in the garage, same altitudes, etc. The BLM?s indicate current state of the fuel delivery system; meaning, if the BLM?s are steady between ~120 and ~136 (by some people?s standards) action could be taken to improve the base-line fuel delivery system, but the ECM does have control as it is, and the resulting fuel-ratio is averaging 14.7, so it could be left alone. Since those parameters have been met with my system, it would indicate ?failed SMOG?, for whatever cause ?high HC, or high CO, the problem has to be with something specific to an individual cylinder; not the MAP, TPS, or CST; however, your point is well taken that a fuel distribution issue (injectors), or possibly one or more spark plugs failing at the higher RPMs on the monitored bank could cause the ECM to correct the problem on that bank, causing the un-monitored bank to run rich.
I know I can't connect two O2 sensors in parallel. I?d like to have the option of selecting one, then the other (via switch), to compare affects on INT readings. If I can find room, I may install a second O2 sensor on the right bank, along with a switch.
Your comment regarding ECM response-time due to the distance the exhaust charge would have to travel before the ECM would see it, if the sensor was placed after the cross-over pipe is interesting. I know the auto-trans equipped ?92-?95 Chevy trucks used a PCM, and placed the O2 sensor after the cross-over. As you know, the PCM operates at a much higher CPU speed, and has much more memory than the 160 BAUD ECM, that is used with the manual transmission trucks of the same years (my application is manual trans). As far as I know, all of the ECM equipped Chevy trucks had a single-wire O2 sensor mounted directly in the driver?s side manifold. I?ve considered it may be an issue to move my O2 sensor from the manifold to the cross-over collector; I don?t know. GM sure seem to consider that.
From: PERFormance DYNamics <mailto:perfdyn at att.net>
Sent: Saturday, September 03, 2011 5:07 PM
To: Roger and Carol <mailto:rogerancarol at cox.net> ; gmecm at diy-efi.org
Subject: Re: [Gmecm] Failing Calif SMOG
It appears to me that one of the quirks of Gmecm is not everyone seems to receive every post. You refer to info "from previous posts", yet in reading back through every post I've received, some of that info is missing.
What about the CO2 and O2 readings ?? Combined they all tell the whole story. Individually they can be misleading. Still working half blind. A table form makes it a lot easier to piece together the picture.
Idle readings : Std's ~2500 RPM Std's
RPM 649 2635
CO .03 % ??? 2.06 % 1.0 %
CO2 ??? ???
HC 21 ppm 120 ppm 45 ppm 180 ppm
O2 ??? ???
NOx na na (no load tests)
Yes, a missfire results in unburned fuel(elevated HC readings) and OXYGEN. The ECM interprets this oxygen as a too lean A/F mixture and adds more fuel, resulting in an over-richened A/F mixture(elevated CO readings) in all cylinders. The Oxygen Sensor doesn't react at all to HC content, only O2. CO is typically formed when there is insufficient oxygen(rich A/F mixture) to complete the formation of CO2.
"It would seem that if BLM?s are steady, and near 128, that the resulting exhaust gas (post CAT) should read ?normal? emissions.". BLM's don't have to be near 128 to be at stoichiometric or to achieve low emissions. The better indicator is the INT numbers. If they are near 128 and varying then the ECM is in control and the BLM's haven't hit an upper or a lower limit. The BLM's near 128 only indicate that the fueling calculations are in need of little adjustment. This of course is a good thing.
Almost anything that effects fuel or air distribution in either side could be at fault. The fact that only one side is monitored will increase the error in the unmonitored side. If the average A/F mixture starts out correct but a distribution issue causes one side to be slightly lean, the other side will be slightly rich. With the monitored side lean, the ECM will richen the fueling to all cylinders, getting the slightly lean side correct and driving the slightly rich side richer yet.
The new information of past emission test results and past datalog BLM numbers would suggest the ECM operation is unchanged yet the emissions have. That lends credence to the right bank rich injector theory. This assumes that the BLM readings are at similar conditions and similar to those when emissions testing. Otherwise it might be comparing apples to oranges. You could try swapping injectors side to side and see how it effects BLM numbers and/or CO readings. While you're at it, check the injector resistance in all injectors. Low resistance can cause somewhat unpredictable driver operation.
As for the dual O2 sensor idea, this would be great to have if you could incorporate both readings in the datalog to observe side to side differences. To be able to use both O2 sensors for ECM control you need an ECM that accepts two O2 inputs. The sensor signals can't be combined into one signal input.
Placing the O2 sensor where it sees both banks minimizes the maximum error in either bank. With everything being perfect this should not be neccesary. The down side is a slower correction reaction time due to the increased exhaust length between the cylinder and the O2 sensor.
Carl
----- Original Message -----
From: Roger and Carol <mailto:rogerancarol at cox.net>
To: PERFormance DYNamics <mailto:perfdyn at att.net> ; gmecm at diy-efi.org
Sent: Saturday, September 03, 2011 10:08 AM
Subject: Re: [Gmecm] Failing Calif SMOG
Thanks for the info?
Original readings From previous posts, this subject, for SMOG measurements:
?---HC at idle measured 21 ppm, well within the limit of 120 ppm. HC at "high
idle" (2635 RPM) measured 45 ppm, also well within the limit of 180 ppm.
In past years, CO has averaged (past SMOG tests) .00% at both idle and high idle., and HC
has averaged 16 ppm at both idle and high idle. Looking at old logs I've made
of BLM's show an average of 130, depending on RPM/MAP. Again, confirming
that the engine has been biased slightly lean since the install, but corrected
by the ECM.---?
AND:
?----?94 Chevy 5.7L V8 with Edelbrock MPFI. ECM 16171199, with EPROM tuned by
Edelbrock, based on broad cast code BDUY.
The engine failed recent SMOG test with a CO reading on the ?high idle?
test of 2.06% at 2635 RPM; max CO allowed is 1.00%.
CO at ?idle? measured .03%, at 649 RPM, which was well within the limit.-----?
Wouldn?t a ?lean misfire?, or any ?misfire? imply no fuel burnt? No burnt fuel would imply no CO generated?which only occurs when the the fuel ratio results from a combination of too much fuel vs oxygen (air), resulting in an incomplete burn?not a misfire. A misfire results in an unburned charge of fuel, and air (O2), being expelled into the exhaust manifold; I?m not sure how the O2 sensor reacts to this combination of raw fuel VS additional oxygen simultaneously. It would seem that if BLM?s are steady, and near 128, that the resulting exhaust gas (post CAT) should read ?normal? emissions.
The only way I can see high levels of emissions in the ?post CAT? exhaust, while at the same time the BLM?s are ?normal?, is either a ?false lean?, or a faulty O2 sensor that is heat sesitive, i.e.: reads OK at idle-speed exhaust temps, but reads a rich mixture as a lean mixture at higher temps resulting from higher engine RPMs. By the way, my O2 sensor is heated. Also, the EGR valve is inhibited from opening when the truck is not moving (needs VSS input).
I think a stock ?87-?95 Chevy V8 would react differently to a bad injector, than my Edelbrock MPFI equipped engine. The stock TBI injectors affect both banks if one is dripping or has a diminished flow. While they are alternately fired by the ECM, they?re mounted above the manifolds plenum, allowing each injector to affect all cylinders. Obviously, the MPFI injectors are isolated to one cylinder. So I keep coming back to reasoning that if the ECM, which is only monitoring the left bank, is showing a BLM that is stoichiometric (14.7:1 fuel ratio), but the post CAT exhaust is saying there is elevated CO, and to some degree, elevated HC as well, then the problem lies (lays?) in the right bank, and has to be a problem specific to the right bank, that could effect high CO at high idle only, with also causing elevated (but not excessive) high HC at the same time. It?s my understanding that elevated CO is generally accompanied by elevated HC; this makes sense, since CO is caused by an incomplete burn, which also results in some unburned fuel being expelled from the cylinder post fire as well.
As stated in another post, I?d love to put another O2 sensor in the right bank to make comparisons with the left bank, but space is at a premium, meaning I?m not sure I have the room. I?ll probably install the sensor in the collector after the cross-over pipe, and feed both banks to the ECM, and see what happens to BLM?s.
From: PERFormance DYNamics <mailto:perfdyn at att.net>
Sent: Friday, September 02, 2011 11:23 PM
To: Roger and Carol <mailto:rogerancarol at cox.net> ; gmecm at diy-efi.org
Subject: Re: [Gmecm] Failing Calif SMOG
A right bank Injector problem that would cause high CO (leaking ?, stuck open?) would probably cause high CO at both idle and 2500 no load. It is likely it would effect the idle readings most. The same would hold true for a leaking fuel pressure regulator. Though not impossible, this seems improbable.
A missfire in the left bank would cause excess oxygen at the O2 sensor driving the fueling richer(high CO). This of course also elevates the HC levels. The Catalytic Convertor then does its job combining the excess O2 with elevated HC and CO, lowering both. In this case the HC levels could be low enough to pass with the CO levels still at a level that fails. Of course this is just conjecture. What are the HC levels? The O2 levels? The CO2 levels?
With the INT is varying around 128, then the ECM is controlling the A/F mixture at what it believes to be 14.7:1 in the left side. The high CO only at ~ 2500 RPM suggests either:
1] the O2 sensor is lying to the ECM
a] a faulty O2 sensor, contaminated sensor (higher temperature at RPM could trigger the issue)
b] exhaust leak allowing oyxgen into the left side
c] missfire causing elevated oxygen content in the left side, drives the fueling richer
2] there is a difference in operation between the left and right banks
a] fuel distribution issue (port FI with healthy injectors - this seems unlikely but worth mention)
b] air distribution issue, restriction to flow in the intake or exhaust of the right bank
(less air flow with the same amount of fuel = rich mixture)
c] worn cam lobe (see b])
d] the EGR gasses could be effecting one side more than the other
(check CO readings with ERG disabled)
3] the engine operation changes between idle and ~ 2500 RPM
a] weak/broken valve spring
b] a restriction to flow in the intake or exhaust of the right bank
c] EGR operation
An exhaust leak that would effect O2 sensor operation would most likely be audible . . . using the right listening device. I use a length of 3/4" heater hose, 3 1/2 feet long. This is the best listening device I've used, and I believe I've tried them all. The best time to listen is just after a cold start when the engine is inefficient (lots of exhaust pulse).
Carl
----- Original Message -----
From: Roger and Carol <mailto:rogerancarol at cox.net>
To: PERFormance DYNamics <mailto:perfdyn at att.net> ; gmecm at diy-efi.org
Sent: Thursday, September 01, 2011 10:01 PM
Subject: Re: [Gmecm] Failing Calif SMOG
Thanks for the reply;
INT is varying slightly around 128 (127-129), which indicates the ECM is in control. The BLM?s are steady, and within tolerance.
My understanding of what causes CO would imply a misfire would result in high HC, not CO. Co is caused by incomplete combustion, due to a rich mixture.
A ?false lean? could certainly cause a rich mixture, resulting in high CO. Not sure what level of leak would be required, to cause the O2 sensor to read ?lean?. Would such a leak be audible? I?ve searched for reports of verified examples of such a leak without success. At any rate, I replaced the exhaust manifolds (but not to solve this issue), and gaskets without affect. I certainly do not here an exhaust leak.
No AIR used, so that?s not a cause. Remember, the high CO is only at RPMs above ~2300; idle HC is well within limits. If the INT and BLM?s are close to 128, that implies stoichiometric to me; HC?s and CO?s should both be ?good?, unless the O2 is sensing a false lean condition.
I?m inclined to believe the problem is with the right bank, which is not monitored by the ECM, but is ?analyzed? by the SMOG machine.
From: PERFormance DYNamics <mailto:perfdyn at att.net>
Sent: Thursday, September 01, 2011 2:56 PM
To: Roger and Carol <mailto:rogerancarol at cox.net> ; gmecm at diy-efi.org
Subject: Re: [Gmecm] Failing Calif SMOG
Look at both the INT and BLM readings at 2600 RPM, warm, no load. Is the ECM in control under those conditions? At minimum the INT should vary some, even if the BLM seems stable.
If the ECM is in control and the CO readings are that high, check for anything that can cause the introduction of oxygen into the O2 side exhaust. Any additional oxygen will be interpreted as a lean condition and the ECM will drive the mixture richer in all cylinders. Exhaust leak? Upstream AIR air? Could you have any missfire for any reason on the side with the O2 sensor ? (a missfire contains not only the unburned fuel but also the unused oxygen)
If the INT is at 128 and not varying under those conditions, something may be disabling learn.
Happy hunting.
Carl
----- Original Message -----
From: Roger and Carol <mailto:rogerancarol at cox.net>
To: gmecm at diy-efi.org
Sent: Tuesday, August 30, 2011 3:27 PM
Subject: [Gmecm] Failing Calif SMOG
?94 Chevy 5.7L V8 with Edelbrock MPFI. ECM 16171199, with EPROM tuned by Edelbrock, based on broad cast code BDUY.
The engine failed recent SMOG test with a CO reading on the ?high idle? test of 2.06% at 2635 RPM; max CO allowed is 1.00%.
CO at ?idle? measured .03%, at 649 RPM, which was well within the limit.
My BLM?s at 2500 RPM read 130 on my scanner, which I believe means the PROM?s fuel map is biased ?lean?, and the ECM is compensating as it should. I?m confused as to how the BLM?s can indicate that the ECM has control over the fuel ratio but the exhaust gas (SMOG test machine) indicates a ?rich? condition, which results in high CO? Note that CO at idle was .03%, which is well within the limit, thus at idle fuel mixture is correct.
This engine has only one O2 sensor, which is located in the driver?s side exhaust manifold (manual trans applications only). Does it make since that the cause for the rich condition is probably associated with the passenger-side only (not any sensor that affects all cylinders, like the MAP), and therefore not ?sensed? by the O2 sensor? Since I have Edelbrock?s MPFI, I think it could be an injector problem on the right bank only. Edelbrock?s system is batch fire, by the way. Inputs from ?fresh? minds would be appreciated.
_____
_______________________________________________
Gmecm mailing list
Gmecm at diy-efi.org
http://lists.diy-efi.org/mailman/listinfo/gmecm
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.diy-efi.org/pipermail/gmecm/attachments/20110913/e13520b4/attachment-0001.html
More information about the Gmecm
mailing list