kevm14

I figured this day would come eventually. I'm surprised it took this long. But I finally failed an emissions test with my 93 Caprice.

The car has 265k on it, but the manifolds and cats are from a 96 Impala SS. I put them on in early 2005 I believe. They have far fewer miles than the car does, maybe 150k or so. The AIR system is connected but I don't think it functions except during the first few minutes of a cold start. The car doesn't smoke on startup or any other time.

2 years ago, with 254k on the clock, it passed the RI dyno test (I think it is a modified IM240) with the following results:
HC GPM 1.71 / 2.25
CO GPM 13.91 / 23.74
NOx GPM 1.39 / 3.25
CO2 GPM 577.86

At the time I had disabled my EGR and had a dynomax dual catback with no H pipe. The sheet says the car was coded as having a single exhaust.

Fast forward to 1 month ago. I put a new cat-back on, but this time with an X-pipe. With 265k, the machine refused to run the car on the dyno, citing exhaust gas dilution. Brand new system, no leaks. Still coded as single exhaust. They eventually gave up and ran a 2-speed idle test with the following results:
HC PPM idle 237 / 220
HC PPM 2500rpm 369 / 220
CO % idle 0.64 / 1.2
CO % 2500rpm 0.27 / 1.2

The only thing that changed was I pulled out about 2° timing globally from the 2 year prior. My mitigating actions to try getting it to pass on the re-test were the following:
Plugs (AC43TS), wires, cap, rotor, PCV
Extensive re-tune of low rpm VE. Was off a bit but not more than about 8%.
Added timing back that I took away.

I went back for the re-test today and got the guy to plug the driver's side pipe and put the sniffer in the pass side (to simulate single exhaust). The machine automatically forced an idle test remembering the previous attempt, and failed with the following results:
HC PPM idle 827 / 220 (wow, terrible)
HC PPM 2500rpm 166 / 220 (ah, a pass!)
CO % idle 0.72 / 1.2 (a bit richer than before)
CO % 2500rpm 0.65 / 1.2 (a lot richer yet a cleaner result)

I have a couple thoughts then I'd be curious about opinions.

The car was running quite warm as my fan controller is set a certain way. Might have been over 230°. I drove for 30 minutes prior to arriving, though it ended up sitting for 20 before going in.
Could my cats be tired?
I ordered an AFS-74 heated O2 and will be installing that this weekend, to both freshen my old single wire O2 and rule out cooling issues, even though it passed 2 years ago with the single wire in the LT1 manifolds (bung is on the cat side of the flange, not in the manifold like OEM TBI).
Denatured alcohol?
WTF is up with the exhaust gas dilution issue? Is the X-pipe to blame? Maybe it passed with the non-H dual system coded as a single exhaust since the banks are independent but with the X the single exhaust coding caused a problem? Maybe I should just retry with them re-coding it as a dual system (they can do this, I confirmed).
Should I pull some spark advance before I go back?

I don't know exactly what I should do. I'd like to keep the car on the road.

kevm14

I am going to put my spark advance back to stock. There are about 3° of base advance unaccounted for in the tune, another ~3° at light load (1200-2800rpm) and ~3° some higher load 2000-3600rpm. In addition, the cool comp spark table added another 1-2° under heavy load at operating temp. So yeah, there was extra spark in a bunch of places over stock. But all of this was present 2 years and 10k miles ago. However, it seems extra spark advance can cause excessive HC. More specifically, less advance creates more exhaust heat which burns up more HC.

Anyway, I'll remove it all and install my AFS-74 this weekend. I suppose I should re-tune my VE if trims are way off after the AFS-74 and spark reduction. Should I reduce to levels below stock or is stock a good goal for my re-test?

DemonRS

If you have dual exhaust and they only test it as single exhaust (one side), won't that cut the emissions in half?
The high engine and cat temperature is good for the smog test. Always warm up the car before taking it to be tested.
With that many miles, it would not hurt to sea-foam the engine, change the oil, add a fuel cleaner, & fill up with premium for the test. It worked for me with 242K.
Is the place you are taking it the same as the one that passed it last time?

pound

iTrader: (7)

Is the motor stock? If it is, then I'd try setting the timing to stock. It must have passed at one time with the stock timing. Also don't worry too much about the higher load areas of the spark map as they will never be reached during the test.

FWIW my prom has about 20 degrees at idle in the main SA table and then by my guess about 27+ if you were to just hold it at 2500. I didn't data log the test so that second number is a guestimate on my part, but I do know where it idles so the first number is accurate. Maybe my timing in those areas of the map is too high and not optimal. I am not really sure, but I passed NJ emissions with no AIR and no EGR thusly:

Idle Test:
HC Actual: 201 Standard: 220
CO Actual: 0.04 Standard: 1.2
CO2 Actual: 14.5 Standard: N/A
O2 Actual: 0.2 Standard: N/A

High Idle Test (2000 RPM):
HC Actual: 145 Standard: 220
CO Actual: 0.34 Standard: 1.2
CO2 Actual: 14.5 Standard: N/A
O2 Actual: 0.1 Standard: N/A

This is with a TBI 350, AFR 180 heads, an LT4 Hot Cam (.525/.525 218/228), and a 3 inch cat.

kevm14

With my old straight dual system, yes, half, as in half of the engine (1 bank). With the X-pipe, things seem more complicated. That's why I got them to plug one side and test with the single sniffer on this past test. 2 years ago they presumably put both sniffers in but coded as a single exhaust, I assume the machine uses only one of them for the actual measurements.

Oil was changed pretty recently. Seafoam...well it's been a while. I was considering it. Though I'm not shy with the throttle.

Also, yes, same place as last time. They at least seem willing to work with me. The one guy who ran it this last time suggested running it above idle for the whole test, since it passed there. No tach hookup, so it is possible...

kevm14

Yup, stock L05 internally. Just a replacement timing chain and the exhaust mods. I've had the injectors professionally cleaned and flow tested a few years ago, and they were clean even then, so they should be fine. 2 years ago it passed on the dyno with all this extra timing. Some of it I added to help breathe a little life into this high miler. But yeah stock timing may be a very good place to head next. I also think I will have them re-code it as a dual exhaust so both sniffers are going. This may allow me to run on the dyno, where I think it will pass as it did 2 years ago. Somehow I think the single sniffer in an X-pipe system made it kick out for exhaust dilution. That's about the only theory I have on that.

kevm14

I got the AFS-74 in and all wired up. I significantly shortened the cold closed loop timer and reduced the temp threshold. It seems to be working as I have O2 activity basically off a cold start, something the single wire would never have been capable of. My BLMs don't seem different really when warm, so I'm not sure I really needed it, or a new O2 at all.

I also put all the spark advance stuff back to stock. Took some snap and power out of the engine, but if it passes this way, I can certainly put it back...

kevm14

I went to another shop at the suggestion of a RI emissions tech/representative. They had the same issue with exhaust dilution. I got them to configure the car for dual exhaust and initially, he used only 1 probe. I convinced him to try using the 2nd probe and it had the same problem, though it did complete the preconditioning ramp up/down. The preliminary results also looked pretty good. But at the start of the test where it idles for a few seconds, it just would NOT even get to the part where you ramp up. Ridiculous. Has to be the X-pipe, somehow.

Anyway, he kicked it back to the 2-speed idle test (this seems familiar...). He did the whole test at rpm, including the idle portions. He was at around 3000rpm according to my scantool. HC came way down (~56/220 now) probably from a combination of the high rpms and my spark advance reduction since last time. CO, however, came way up, almost to the limit. I guess the cats are good, and I kind of expected CO to come up with later ignition, but I wonder if the engine is just getting tired (exhaust valve burned, compression, etc).

In any event, I have 2 years until this BS all over again.

RBob

iTrader: (1)

Isn't dilution too much air? If so maybe the AIR system isn't working correctly, or there is an exhaust leak that is pulling in air.

OK, found this:

http://www.ctemissions.com/expect-dilution.html

According to the info in that link it may be the exhaust system. But could also be what I mentioned above.

RBob.

kevm14

I do have the AIR system plumbed up to the LT1 manifolds. I'm not exactly sure how to check if the divert solenoid is working, though. From the factory, it had 4 port injection on the passenger side manifold, and nothing on the driver's side, which is where the O2 is. The LT1 manifolds each had an AIR port for the whole manifold. I plumbed in my AIR pump to both of those ports. I don't think it is supposed to operate past initial warm-up or during closed loop operation.

2 years ago, I ran on the dyno fine with a dual 2.5" system. But it was straight true duals, no crossover pipes or connections. This system is also a 2.5", which is admittedly bigger than an L05 needs. But I noticed when I put it on, the X-pipe somehow seemed to reduce the volume of gas coming out each pipe, particularly at idle. Prior to the exhaust, the car had a pronounced high velocity chuff. It would actually suck and blow a piece of paper against the tip. At the time I had begun to wonder if I had a burned exhaust valve.

After the X-pipe, it worked like a normal car, very low velocity at idle, and fairly hot. It seems that the X-pipe changed the impedance of the flow, possibly causing the machine to sense dilution at idle. Like I said, it does physically put out a lot less apparent flow at idle than before, which I thought was an improvement at the time I installed the exhaust. Little did I know...

According to your link, dilution is lack of CO2.

One thing I should have remembered to suggest was to plug one pipe and use a single probe. That would make the system act more like a Y pipe (but not 100% as the length of pipe from the X to the plugged tail pipe would work like a Helmholtz resonator, with an unknown response frequency).

Since the car is now in the system at these two shops as 2-speed idle required, I would probably have to try a third shop in 2 years, if I want to try my plugged pipe experiment. I DID fail the 2-speed idle on the first re-test with that exact setup, but I am curious how it would change the dilution error at the start of the dyno test.

RBob

iTrader: (1)

To check the AIR system, just pull the hoses off and see/feel which way the air is going. Put that hose back on and check the other(s). Can run the engine RPM up and down along with CTS for the checking.

RBob.

kevm14

I'll have to check it out just to be sure. For now, I passed, at least.

JP86SS

iTrader: (1)

What were the final passing numbers?

kevm14

Well at 3000rpm for both the idle and 2500rpm portions, it was something like:

HC PPM: 56 / 220
CO %: 1.1 / 1.2

He screwed up the gas cap test so the final sheet doesn't have the results, but I did look at them before he threw the sheet away.

Does retarding the spark typically trade HC for CO?

kevm14

I'm back! 2 years has passed and I'm now at 280k. I just got back from the inspection.

Knowing the history, I took the following measures before bringing the car to the shop:

- I selected a different shop so they could try the dyno (previous machines would remember the car and insist on a 2-speed idle test straight off so a new shop would at least let me attempt a dyno run)

- I seafoamed the engine (it had been quite some time)

- AutoRX treatment

- Cleaned out the PCV hoses (the breather side had quite a bit of oil junk in the hose, but not plugged at all)

- I tested the AIR system to make sure it wasn't injecting during closed loop (I don't think it was) but capped it internally just in case, and disabled it in the cal for good measure

- When I upgraded to the AFS-74 HO2S, I set the min temp for closed loop about 30°F lower than stock, and significantly shortened the cold closed loop timer. I set that all back to stock, thinking it would help the cats light off by running richer/OL during warmup. I don't see how that could have had an effect on this test but I am mentioning it anyway.

- I bought a rubber plug/stopper and inserted it into the driver's side exhaust. This sealed the system well, and since I have an X-pipe, it would essentially become a post-cat Y-pipe. Given the issues with the dyno in the past, I was hopeful this would solve the "exhaust gas dilution" issues which previously kicked me back to the 2-speed idle test (which the car also had trouble with)

- I tried to warm it up a bit before bringing it in but the "state regulations" mean they do safety first so you cannot bring it in hot off the road no matter what

The results? They told me it could not run on the dyno due to "a locking rear end" which sounds like a bunch of BS to me. It does have a Truetrac but it had that for the last test and the problem was always the dilution, NOT the rear end...this makes no sense. Needless to say, the car ended up back on the 2-speed idle test, which it failed. Here are the results:

----------------RESULT LIMIT
HC PPM IDLE: 161 220 << PASS
HC PPM 2500rpm: 243 220 << FAIL
CO % IDLE: 0.13 1.2 << PASS
CO % 2500rpm: 0.41 1.2 << PASS

So this is surprising. It did better than I thought.

When they pulled the car around, the exhaust plug was still installed in the driver's side tip. Assuming they tested it with a single probe for the 2-speed idle test, is it possible that the extra backpressure from turning my dual 2.5" system into a single 2.5" system made the cats work more effectively? It seems possible...or it was the seafoam...not sure.

In any event, my proposed solution is simply to drop some spark advance for the 2500rpm portion of the test (maybe 40kPa and lower?). I do have more than stock in this area, about 2°. But how much? I guess it was close to the limit so I could see pulling 3°. Or maybe more?

Thoughts?

kevm14

I did a little poking around with WinALDL. Here's a rough chart of no-load RPM vs MAP:

NOTES:

• Injectors are async the whole time
• OL in the notes means BLM not enabled (per the WinALDL flag)
• INT stays active regardless (and the "CL" flag stays checked)

So even if the BLM is not enabled, does the INT still affect fueling? I would assume so. The min kPa for BLM enable is 18 so not sure what the code is doing here. Maybe it's just out of bounds for the BLM RPM/MAP cell table, but the INT still works.

Anyway, in my spark map, the entire 30kPa column is stock (I remember now that I didn't want anything funky in overrun so I left it stock). How does the ECM scale spark for loads under 30kPa? Does it use anything above 30kPa to scale (which have about 2° greater advance than stock, up to about 45kPa (and 1200-2400rpm)?

So my question is, should I try to pull spark in the 2000-2500rpm range, which is also about 25kPa? If so, how should I go about pulling spark at 25kPa when the spark table only goes down to 30kPa?

kevm14

I actually e-mailed Tunercat support on this question. According to him, at least for my '6965/$62 mask ECM, any load under 30kPa just uses the SA for 30kPa. Does this sound right?

My 30kPa SA is stock. That doesn't mean I shouldn't lower it, but I should also check my VE. It may be higher than stock.

So is the INT controlling fuel during these conditions (i.e. 2500rpm and 25kPa)? If so, I probably shouldn't be too worried about my VE as my AFS-74 can probably be trusted to correct it.

pound

iTrader: (7)

Yes.

As long as you are in closed loop then INT should be active.

You are so close to passing I'd probably just pull a few degrees out of the 2500rpm @ 25 kPa and adjacent cells and I bet it will pass. My car is failing now too. At first HC at idle and 2500rpm was pretty high (over 300 ppm). I dropped my HC at 2500rpm over 100 ppm by pulling approx 4 degrees out of those cells. It passes the high idle test easily now, but idle still is an issue. I've finally given up on tuning to get it to pass. I am going to be replacing the cat next.

kevm14

I wasn't 100% sure what was going on because the "BLM enabled" checkbox in WinALDL went unchecked around 1400rpm free rev. But closed loop was still checked, and the INT was still responding. So maybe it just stopped updating the BLM. As long as the INT isn't at a limit, I should still be under full closed loop control, then. In which case, modifying my VE would provide no utility, unless I go to open loop (no plans to do that). If I recall, INT come up to around 136 but I could re-log and confirm.


Ah, thanks for the feedback. Did your CO come up when your HC came down? If the burn finishes outside of the combustion chamber, then maybe CO wouldn't come up. CO, to me, indicates an incomplete burn, which could be from a rich mixture (all the O2 is used up but there's still fuel left to burn) or from late ignition (unless it burns in the exhaust). Of course, I'm not taking into account the cats at all here...I should stop my speculation.

Sounds like even 1° would do it. 2.5° is a number...

What is your CO @ idle? Do you still have that LT4 hotcam setup? Could just be trouble caused by the overlap. Have you considered raising the idle? They don't check RPM here in RI, I've noticed (or timing, or anything other than exhaust emissions).

pound

iTrader: (7)

I don't have the numbers in front of me, but all the other readings were withing passing levels except the HC. HC is a measure of incomplete combustion. Good luck on your retest I'm sure you'll get it to pass eventually.

kevm14

I just got back from my re-test. The opposite happened; it got worse when I dropped ~2.5° from the free-rev 2500rpm area. I did nothing that would affect the idle area though both HC and CO dropped a bit there. One possible reason is they didn't have to do safety again so there was less cool-down time before they started it up and ran it on the dyno machine.

Specifically:

----------------RESULT LIMIT
HC PPM IDLE: 136 220 << PASS
HC PPM 2500rpm: 382 220 << FAIL (much worse)
CO % IDLE: 0.12 1.2 << PASS
CO % 2500rpm: 1.44 1.2 << FAIL (much worse)

Why would it go the other way? I thought my INT would have been keeping things at stoich but is it possible I also need to decrease VE now that I've pulled SA? The CO indicates too rich....right? I'm confused here.

Would the obvious course of action to be advance timing by a couple degrees at ~2500rpm and 25kPa? I don't know what else to do.

It looks like they tested again with my plug in the driver's side exhaust, for what that's worth.

kevm14

I am reversing the 2.5d that I originally removed from the 2500rpm free-rev area, and adding an additional 1.4d on top of it. I have all week to waffle back and forth on this (I will make another appt. for next Saturday), pending feedback here.

morgsie

iTrader: (1)

Just put new 3-way cats on it and be done with it.

kevm14

Good news! I passed.

The numbers:

----------------RESULT LIMIT
HC PPM IDLE: 96 220 << PASS
HC PPM 2500rpm: 58 220 << PASS
CO % IDLE: 0.19 1.2 << PASS
CO % 2500rpm: 0.59 1.2 << PASS

What changed:

As I mentioned before, added 1.4° of SA in the 2500rpm free-rev area. Contrary to emissions theory, this made a HUGE difference (vice removing 2.4° which completely trashed the 2500rpm HC and CO results). I also took it on a more elaborate warmup involving highway driving. That explains the idle HC further dropping (I changed nothing in the cal for idle). But adding the 1.4° of SA is the majority factor on the 2500rpm HC PPM drop. And it was a big drop.

So there you go.

kevm14

I thought I'd continue this thread. Despite having issues in the past, I did very little for my 2015 emissions check.

I did a cap/rotor recently due to another problem I was having. Other than warming the car up before the test, I just brought it straight in. They had done safety at another time (the entire computer system was down and nothing could be recorded), so the car went straight into the emissions test. Well, not exactly. The machine went through a lengthy calibration process, during which time the car sat idling.

Two fairly amazing things happened. First, the tech had no issues running the car on the dyno (same place and even same tech as I had 2 years ago). This in and of itself is big news. I was confident it would do well on the dyno, as I had been saying for all these years. And so it was. Check out these results with 292k on it now:

HC GPM 0.37 / 2.25
CO GPM 4.25 / 23.74
NOx GPM 0.42 / 3.25
CO2 GPM 215.80

Cleaner than it was 6 years ago when I have the last recorded dyno results.

Never did cats, by the way.

Not sure exactly how the math works, but the car will turn 25 in 2017 when it would next be due. This MAY be the last emissions test it is actually required to pass.