CA Smog Test only, passed pre-test failed the one that matters

[emcon5]

Time for the semi-annual pain in the arse that is the California emissions testing.

1982 911 SC.

Changed oil, new plugs, timing set, plugged a dwell meter into the lamda test port, and it swung from 40-50 degrees, averaging at 45º which meant the frequency valve is running at 50% duty cycle, right where it should be.

Took it to a regular station for a pre test, like I always do, and it passed.

15MPH: 13.9%CO2, 0.4%O2, HC 59, CO .61% NO 387 Pass 1545 RPM in 2nd gear
25MPH: 14%CO2, 0.2%O2, HC 35, CO .68% NO 295 Pass 1726 RPM in 3rd gear

The only thing close was the CO at 25 MPH, which was .68% with a max of .79%.

Took it across the street to the test only station for the official test, and it failed.

15MPH: 13.5%CO2, 1.14%O2, HC 51, CO .69% NO 352 Pass 1540 RPM in 2nd gear
25MPH: 13.5%CO2, 0.9%O2, HC 9, CO .92% NO 278 Fail 2384 RPM in 2nd gear

The only portion that failed was the CO% at 25MPH, and the primary difference between tests is the Test Only guy did it in 2nd gear instead of 3rd. NO and HC are not even close.

Any guidance on what to check to get it lower? I only need it to come down about .2%

Is there any significance to the difference in the %O2 readings between tests? It was about 10% higher at the test only.

The pre-test guy said his timing light showed the ignition a little over 6º advanced, in theory I could retard the timing 2 degrees and still be in spec, would this help?

Thanks,

[Eagledriver]

Are you required to use the test only station? You could try asking the operator to use 3rd gear for the 25mph test.

-Andy

[emcon5]

Yes, I have to go to a test only. I have always done a pre-test to make sure it will pass before the official test.

I did go ahead and replacedthe O2 sensor last night, and for what it is worth, the variation in the duty cycle of the frequency valve seemed to flucuate a little less than on the old one. I also retarded the timing one degree, so it is still in spec. It used to hunt very slightly at idle before, now it doesn't seem to, but I haven't driven it enough to be sure.

I asked him yesterday, about the gear, he said that the state wants them to do it in 2nd gear, unless the RPMs are out of the range.

Retests are free, so I will take it back today and plead with the guy to do the run in third gear.

I am still having a hard time understanding the differeing O2 values. Why did it jump from 0.4% and 0.2% O2 to 1.4% and 0.9% with no changes to the car.

[Peter Zimmermann]

What was your pre-cat exhaust sample CO% on a hot (front cooler just opened) engine? It should be 0.4-0.8% with the oxs sensor disconnected and the idle at 900 rpm. 1981-83 cars are happier at 0.4-0.6%, which can "steady" the idle a bit.

[love911]

It's the law ,smog tech's should start the test in 2nd gear and proceed to the 3rd gear if the vehicle is equipped with manual transmission unless the rpm exceeds the limit ,which in that case they usually start with 1st & finish in2nd, which in your case as it shows and all the porsche 911's the tech in the testonly station should have used the 3rd gear for the 25mph portion of the test,and yes the correct gear makes a difference on the CO reading as you see in the pretest.Maybe the testonly guy doesn't know or is not confident enough to change the gear in fear of not recovering the required speed in time during the 25mph ,thus causing the test to be aborted,I've seen that before.Show him the pretest and tell him to run the test again in correct gear and rpm.

Good luck.

[vash]

my guy did it in 3rd, and rode my brake? i think..i didnt ask him to do this, but he said it helps on these old porsches. maybe it was in 2nd and then he held the brakes..dunno. i know he rode the brakes.

[juanbenae]

Originally Posted by vash

i know he rode the brakes.

early or late in?

[emcon5]

Originally Posted by Peter Zimmermann

What was your pre-cat exhaust sample CO% on a hot (front cooler just opened) engine? It should be 0.4-0.8% with the oxs sensor disconnected and the idle at 900 rpm. 1981-83 cars are happier at 0.4-0.6%, which can "steady" the idle a bit.

I don't know Peter. I set the mixture the same way I have before every smog check so far, I leave the O2 sensor plugged in and adjust it until the Frequency Valve duty cycle is as close to 50% as I can get it.

Originally Posted by love911

It's the law ,smog tech's should start the test in 2nd gear and proceed to the 3rd gear if the vehicle is equipped with manual transmission unless the rpm exceeds the limit ,which in that case they usually start with 1st & finish in2nd, which in your case as it shows and all the porsche 911's the tech in the testonly station should have used the 3rd gear for the 25mph portion of the test,and yes the correct gear makes a difference on the CO reading as you see in the pretest.Maybe the testonly guy doesn't know or is not confident enough to change the gear in fear of not recovering the required speed in time during the 25mph ,thus causing the test to be aborted,I've seen that before.Show him the pretest and tell him to run the test again in correct gear and rpm.

Good luck.

Do you have a document or reference for that? If so, I will print it out and take it in with me.

Sadly I know all about the aborted tests, the pre test guy did that 4 times.

[Peter Zimmermann]

Originally Posted by emcon5

I don't know Peter. I set the mixture the same way I have before every smog check so far, I leave the O2 sensor plugged in and adjust it until the Frequency Valve duty cycle is as close to 50% as I can get it.

The only thing that I can suggest is to set the CO% with a CO machine. You might have to contract a shop to do it. I have never had a problem with a smog test (SC or 3.2 car) by checking the CO% using the factory method (machine) as outlined in the Technical Spec books. Engine hot, idle 900 rpm, OXS sensor disconnected, exhaust sample taken pre-cat using the test nipple on the side of the cat.

[87 930]

Usually, but not always, it's better to run the car in a lower gear for the test. The car runs more efficently and you're sending hot gases back to the cat to get it hotter and lower emissions better.

Cat temp is super critical when testing and often the reason why one smog test results will vary so much from one done an hour later. That's my guess why the O2 readings are slightly different between the tests. I don't think the test-only site got the cat hot enough (or the first guy got it hotter). The tech must follow a certain protocol to properly heat the cat.

The cat stores O2 under certain conditions to lower (reduce) some bad emissions and releases 02 under different conditions to lower (oxidize) other emissions. Heat makes the cat perform these processes easier. If the engine is operating properly, without misfires or radically rich mixtures, you can't get the cat too hot from the gases produced by the engine. Hotter = better. If your AFRs are too rich (hi CO), or you have a misfire (cylinder not firing and raw HC fuel entering cat), the cat will attempt to consume the extra fuel in the exhaust and work itself to death. Its chemical reactions will overheat it and it'll melt itself. If you had high O2 levels and failed multiple gases (especially at lower rpms), the cat's probably failed and should be tested.

High CO and low O2 means you're probably a bit rich. In your case, not much. The ignition timing affects HCs. Just make sure it's within your allowable timing window and focus on the CO problem. If you can, try to lean it out a bit. I'm a CIS guy and it takes 10 seconds to do it on my Turbo. Get the cat HOT, and test it in second. Good Luck,

[emcon5]

My understanding of CIS-Lamda is that as long as you are close, the base mixture setting isn't all that important because the Lamda system will try and bring it back to stoichiometric with the frequency valve, except when the engine is cold, or at full throttle where it goes open loop. Is this incorrect?

The odd thing is I took it back today, and it was actually slightly worse. He put it on in diagnostic mode, so it just displayed the numbers on the screen, and he said it failed even at 15 MPH now.

He insists he can't run the test in 3rd gear.

I took it back to the place I got the pre test, he did the same thing, but printed out the results for me..

There I passed at 15 mph, but failed at 25.

At 15: 13.9%CO2, 0.1%O2, HC 13, CO .81% NO 257 Only one close to limit (.99)is CO.
At 25: 13.9%CO2, 0.1%O2, HC 15, CO .84% NO 428 Only one over limit (.79) is CO, by a whopping .05%

[87 930]

I've owned carbed Porsches and now my '87 930. I'm assuming that your '82's CIS is like mine. Setting the idle mixture screw correctly is extremely critical. It mostly affects idle mixture but also midrange and high rpms. Tiny changes on the mixture screw move the airflow sensor plate, therefore the fuel plunger and strongly alters the amount of fuel delivered to the measured amount of air.

The cat can only do so much to clean up the exhaust stream; the engine has to emit a minimal amount first. The Lambda system is there to help "trim" the a/f mixture a small amount. The O2 sensor and frequency valve can only work in a narrow range, and help lean out or richen a small amount. The engine's a/f ratio must be close to stoichiomtric because if it isn't, the Lambda system can't bring the emissions into legal range.

A tiny CCW turn will lean out your mixture. I mean about 1/16th to 1/8th of a full 360 degree turn. Focus on how much you turned it if you want to go back to original position later. Your HCs and NOx will probably go ap a bit, but they'll still be fine.

[emcon5]

Went by the pre-test place again, and reset the mixture as Peter described above. It was a little rich, but I set the CO at .04%, and plugged the O2 sensor back in and it climbed to about 1%. I did the 15 and 25 MPH test in diagnostic mode, and the results were the same as yesterday. Passed at 15 MPH, barely failed at 25. The CO at 25MPH in 2nd gear swung between .84 and .85%, passing is .79%.

I am stumped.

In related news, anyone got a set of 3.2L heat exchangers they don't need?

[pplkook]

Dumb question - but I live in Santa Cruz, and to the best of my knowledge they don't do a Dyno test here (at least last time I did it).

Can someone from another location - say San Jose, go to Santa Cruz for regular old smog?

-Jack

[emcon5]

No, the renewal form says test only, so you have to take it to a test only or the tester will reject it.

I went through that when I bought my car, it was last registered in Sacramento which had the dyno testing, but at the time, the Bay Area did not, and there were no test-only stations to go to. I had to call and jump through a bunch of hoops for them to take the test-only flag off my car,

[trekkor]

Smog tests are not fun.

We got a pass on our BMW, but only within the very top of the absolute limit...

I'm thinking if the tech farted during the test it would have failed


KT

[stlrj]

Originally Posted by emcon5

25MPH: 13.5%CO2, 0.9%O2, HC 9, CO .92% NO 278 Fail 2384 RPM in 2nd gear

Any guidance on what to check to get it lower? I only need it to come down about .2%

The pre-test guy said his timing light showed the ignition a little over 6º advanced, in theory I could retard the timing 2 degrees and still be in spec, would this help?

Usually high CO is an indication of the mixture being too rich. Backing off the mixture slightly should take care of that. Backing off the timing might help too.


Cheers,

Joe

[emcon5]

I agree, but yesterday when I tried it on the machine, leaning it out didn't do anything, because the lambda system put it back where it wanted it. The problem is, where the lambda system wants it, is not passing.

On edit, forgot to mention, I checked the CIS pressures yesterday, everything was in spec.

[djdsc]

Are you paying for the pre-test? If it is free unplug the box under the passenger seat, before you get there and do your pre-test. There is no OBD to test for the connection.