[spydergear]

Today, I went for my smog test. I prepared for this by changing the oil and filter, the air filter, and the plugs (gapped at .028"). I also filled up with Chevron last time for the Techron. As you can see, it failed rather spectacularly. Even though I drove about twenty-five miles to get to the smog station, the car sat for about a half hour as I waited in line behind two BMW's. I think this is why it failed the 15 mph test - I'm assuming the cat was just too cold. It passed the 25 mph test at a higher RPM in the same gear. BTW, it is a 1990 SPG Turbo.

I had the same situation twenty years ago with my '85 Turbo. They replaced the cat and it still failed. Maybe kind of embarrassed, the tech took it out and drove it and then it passed just fine.

Two years ago the SPG passed both tests at about 1800 RPM.

Anyone have any opinions?

Here is a picture of the results:

 

[jvanabra]

I don't think sitting for 30 minutes would account for that. Usually, the techs will let the engine run while they do the visual... I think they are supposed to do that to ensure the engine is at operating temperature. If they didn't maybe, but frankly to me it looks like there are some other issues, maybe including a weak cat? If you're confident everything is dialed in, then I guess you could retest it... For whatever it's worth, the smog place is six miles from my house.... I jet over there Saturday AM rain or shine and it gets done. A hard run may help a marginal cat get the job done, but if everything is in good shape I don't think it should be necessary.

 

[spydergear]

As a first step in this, I imagine I should check for vacuum leaks, as before the test two years ago, I replaced the breather hose which crumbled upon removal. Maybe I should pull the codes - I haven't done that in a long time.

Recently, I did find another leak at one of the brass fittings on the throttle body that goes to the AIC valve. I used 518 on that (it was what I had available).

Does anyone think a smoke machine is useful? There are certainly a lot of places to get vacuum leaks on a Turbo model...

 

[catlike]

I think a smoke test could be a useful thing to try as a part of smog diagnosis. I just went through the CA smog dance with my '85 SPG, failing three times before passing. The dance went like this:

1. Smog test failed, High HC + High NO on the 1800rpm test
2. Mechanic did smoke test for air/vacuum leaks. Replaced turbo boot and some grommets. Checked and adjusted ignition timing. Did compression test. Oil and filter change.
3. Smog test failed, just over the limit on HC and just below on NO
4. Ignition service: new distributor cap, rotor, leads, spark plugs
5. At another mechanic, failed smog on their first try. Mechanic noticed engine running a bit cold, replaced thermostat with 89 degree part (I have records of previous owner putting in 82 degree thermostat). I believe they also calibrated the Mass Air Flow meter (manual process on LH 2.2). Smog passed on second try with really good numbers.

This is an LH 2.2 car so there is no way to get error codes from the fuel injection system. Since yours is LH 2.4 checking for codes sounds like a good idea!

 

[jvanabra]

That is exactly what I would expect smog troubleshooting on a c900 to look like. All the boxes checked.

A smoke machine is always useful - it's the easiest way to check for leaks. But, it's not the only way - spraying joints & such with carb cleaner will generally expose leaks as well. It's the low-tech approach for sure, but it's worked for a very long time. You can also make use of a pressure pump (the Mityvac Silverline is great). If you've got a decent air size air compressor, you can use that to actually pressurize the intake to simulate boost.

IMO, the devil is in the details on these cars. LH is pretty dumb by modern standards... it won't throw a fit if things are off, but also it has no way to adjust for things that are.

High HC and high CO means it's running rich, or the ignition system is not up to spec. High NOx suggest lean running, or high combustion chamber temps as a result of incorrect ignition timing or carbon buildup. The fact these two are at odds with each other kinda points to multiple issues, like a major vacuum leak that has caused a very rich mixture that the ignition system can't cope with, or a lazy O2 sensor that's not accurately reporting. Could even be something like clogged injectors that aren't properly atomizing fuel, etc.

 

[ReverendPaul]

Man that is an abjectly terrible test! They also labeled it gross poluter (2x normal emissions) so now you'll have to go to a star station (I haven't lived there in 5 years so that part might have changed) to re-test.

An iffy cat would try to be squeezing by and only failing on NOx.

Generally:
Hydrocarbons on cars that old have to be really high (like bad cylinder high) to fail. I've passed with cracked heads/ exhaust valves before (I needed to get tags to not get towed by the apt. complex and got a miracle).
And high Monoxide indicates engine management is just plain wrong...

BUT

That NOX is so crazy high and CO2 (efficiency) is OK so i suspect your catalyst is now little more than a straight pipe. It probably wouldn't even be hot with a thermometer because there's little left inside it.

The good news is a 3 way cat (you have to update per CARB) would hammer the worst 1990s car into compliance on HC and CO.

So just have to get a new cat. If your in the IE go to the Sex-exhaust place in fontana and its $500 out the door.

But that car is WAY past "squeezing" through. It's in the top 5% of worst polluting cars on the road.

 

[spydergear]

Today, I tested the Oxygen sensor with my old Harbor Freight, Fluke copy, DMM. I doubt that this is as good as pro equipment, but it does have a moving-bar graph under the voltage value that moves fairly quickly.

First, I thought I'd test the heater circuit. So, I pulled the connectors apart. I found that the heater connector had corrosion on the terminals (boot has tail broken off) which I cleaned with Deoxit. The main-sensor wire has stainless terminals, I believe, but the insulator body of the male connector on the car's harness crumbled apart into several pieces. I dug those out, reconnected and made my measurements.

When I last tested this, shortly after I bought my SPG, it would only read .5 volts, unlike my '85 that would vary up and down. At the time, I thought that that was weird. Now, I get varying voltage from about 0.2 something, to around 0.75. It speeds up when I raise the RPMs. Plus, now the heater is getting 15.0 volts (maybe the alternator is over-charging?).

Looking back in my notes from about six years ago, I find that I got fault code 12112, which the Bentley manual says is "Oxygen sensor self-compensating circuit problem (incorrect air-fuel ratio at idle)". The Bentley has no corrective actions for this, but I assumed it meant I had a vacuum leak. Now, I find the Factory Manual has a more explicit explanation:

I really don't know if the Lambda sensor was working or not, but it appears to be now. Whether it moves fast enough, I cannot say. And, over the years, I have corrected a few vacuum leaks - but perhaps there are more?

I also tightened the clamp on the MAF that I removed the replace the air filter and I sprayed some Deoxit on the connector terminals, even though they looked clean.

I have questions about the distributor: It seems to lack any centrifugal advance, but the vacuum advance/retard unit has to move this mechanism to get the correct timing if I'm understanding it. How do these coaxial shafts that move the rotor stay lubricated? And could they stick creating static or incorrect timing? The distributor that is in there now is a used unit - are there critical differences in the vacuum units, or the distributors themselves, among different B202s?

On the cat: I'm not saying you are wrong ReverendPaul, but for now I would like to follow other leads first. If it comes to a cat replacement, I've replaced two on the '85 and they made little difference. Currently, there is a new one on the '85 I could use. I foolishly let the muffler shop torch the SAAB cat off that car the second time. But I still have the original cat from the first replacement that I could also try.

 

[jvanabra]

There is definitely a spec for O2 sensor voltage swings, but I'm not sure you could get to that level of science with a multimeter. The voltage should bounce back and forth between <.45v and >.45v quick enough it's hard to read, about 1-2 times per second. The swings should be relatively extreme and relatively balanced, like .2v to .8v. If the swings are narrow (.4v and .6v) or unbalanced (.4v and .9v) then either the O2 sensor has become biased or the mixture is too far off to be corrected. You can further test by adding air or fuel (like propane?) to force a mixture change... if the sensor reads it, you can be more confident the sensor is fine and the mixture is not, or vice versa. The heater is always battery voltage - if you are charging at 15v something is wrong!

There is no advance in the distributor. Timing (and dwell) is controlled 100% by the ignition control module. The vacuum module is a fixed relationship - retards X degrees at Y boost. It is not variable. There are two modules - the common one (which I think is 5 deg at 5 psi) and the "175hp" part fitted to '90+ SPGs and included in the Tuning Kit, which IIRC offers less total retard or maybe a more gradual phase in? I'm not sure. The factory distributors were -008 (standard) and -010 (175hp) but of course vacuum modules can be swapped around. I have had the plate the vacuum module acts on get sticky, which results in wacky performance and was a PITA to troubleshoot! It's easy to test with a mityvac and the cap off, or with a timing light.

 

[spydergear]

I will just have to take it on faith that the oxygen sensor is working with my present equipment. I suppose one of those eBay oscilloscopes would be better, but I don't have one or a laptop. I do have three months of free smog retests until it passes however.

Perhaps I'm misunderstanding you, but are you saying that the vacuum unit cannot advance or that there is no centrifugal advance or that the ignition amplifier controls the advance? I hear that it can vary dwell - that is the charge time for the coil? I know in a points system when you vary the gap, and thus the dwell, timing changes...

I did take my MityVac to a spare vacuum unit and at 15 inches of mercury it pulled the arm in one millimeter. I also tested my complete original -010 distributor and it made the sensor plate and the hall sensor move clockwise a noticeable amount.

This from the 87- Factory manual. They list the 007 distributor for a B202 Turbo, but perhaps that is a European version. Anyway, it does show that it also advances under vacuum:

If so, I can't find any timing specs for the -010, how much the ignition amplifier contributes, or how they vary from other part numbers. Also, not sure how to lube the sensor plate if sticking is a factor since the rotor is usually glued on. Perhaps I could take the vacuum module off and spray something in there. What did you do to un-stick yours?

 

[jvanabra]

There is no mechanical advance on the distributor. The advance is electronically controlled by the ICM, along with dwell. The only thing the distributor itself does is generate the tach signal and funnel the spark from the coil to the plug.

The "vacuum advance module" isn't primarily a vacuum advance module. It's a pressure retard module. At ~5psi of boost (positive pressure) it retards timing ~5 degrees. There could also be a slight movement under vacuum, but I've never personally observed it. The primary purpose is to retard timing under boost. People call it a vacuum advance module because it looks like the vacuum advance module you'd find on old NA car. On the 16vT, it's a pressure retard module.

 

[spydergear]

Interesting. So it is just like my Corvair Spyder, except the advance is electronic. Maybe that is why I haven't noticed any difference in timing when I disconnect the hose to the distributor (at idle).

 

[Rodentmaster]

Just a side note on the exhaust lambda probe.....
Jim M posted some years ago that lambda probes rarely fail outright but more that their response times slow..... they become lazy.
I can say that I did see such response issues when I owned the last 900 but more recently, with the diesel car I have now.
When I went to swap the probe on the diesel, the dealer part supplier virtually repeated the same words describing the side effect Jim noted......response time.

 

[jvanabra]

Interesting. So it is just like my Corvair Spyder, except the advance is electronic. Maybe that is why I haven't noticed any difference in timing when I disconnect the hose to the distributor (at idle).

Yes. The factory calls for you to disconnect the vacuum line, but I think it's out of an abundance of caution. The plate that the module controls generally only goes one way.

 

[ReverendPaul]

On the cat of the 85... I'd sure try it (worth a retest to save money), but don't be surprised if on your car they ask for the actual CARB cert. Unfortunatly Gross Poluter gets you the fine tooth comb and he may want proof it's CARB not EPA (doubtful he will but I myself would ask as Star HAS to fail people).

As said the distributor is "decorative" its almost like the little square ignition modules but with the added advantage of moving parts to break. You can mess with base timing on those rigs; however, timing is part of the smog test (and you passed).
Likely If it's advanced forwards or backwards from where it's now in spec it'll fail. The machine also checks the advance for you.

The test is actually telling you a lot (we use exhaust gas analyzers outside of smog at shops to diagnose performance)

The 14% CO2 means its igniting/ mixing well. But... the 02 is a tad low (0 at 45 is low) so a lean (vacuum leak) is out... which means the NOx is at least hitting from catalyst efficiency somewhat (iffy cat).

To address NOX at this point the only thing maybe that could do it is filthy combustion chambers. Feed some seafoam down its throat...?

If HC and CO were low but NOx was up it would be ignition too advanced; that's not the case and the car passed ignition.

So If we assume the CAT is just iffy, your main lead is in CO. Carbon Monoxide is a rich run gas (the high HC) which the low 02 backs up. (but it's really not often the 02 sensor that causes that).

Carbon Monoxide tends to come from faulty engine management at the sensor level.

I would make sure to scan you MAF, Idle air Temp, Engine Coolant Temp (all doable with a multimeter. Maf to a point). If they think its 68 out when its 104 they will mix too rich.

But again, the 0 O2 could just mean the catylyst is doing nothing and those HC/CO #s are not bad for a straight pipe as 14% Co2 means it's combusting well.

If you had low CO2 I'd suspect more; thats actually really good for a car that old (can tell you just changed the plugs).

All that said, the 1 item that can completely bypass all of this is leaky turbo seals. If you stab the had at idle hard 3x while the turbo is hot and it smokes thats the issue right there. The oil would starve o2 out and the cat just wouldn't fire letting Nox and Co run right through it.

I'll also just defend the assumption its a cat again off the fact its a 90 something. The tailpipe limits are very loose and you can even potentially pass with a dead cylinder.

So something like a bad ETC is unlikely to mess it up so bad it fails. Only the MAF can really tank it that hard and its not running that badly I assume.

 

[jvanabra]

On the cat of the 85... I'd sure try it (worth a retest to save money), but don't be surprised if on your car they ask for the actual CARB cert. Unfortunatly Gross Poluter gets you the fine tooth comb and he may want proof it's CARB not EPA (doubtful he will but I myself would ask as Star HAS to fail people).

Can you clarify what you mean here? CARB EO'd cats are physically marked - all a tech needs to do it slide their mirror underneath and look.

There is no requirement for any smog stations to arbitrarily fail anyone. They don't even have the ability to do that.

 

[ReverendPaul]

Can you clarify what you mean here? CARB EO'd cats are physically marked - all a tech needs to do it slide their mirror underneath and look.

There is no requirement for any smog stations to arbitrarily fail anyone. They don't even have the ability to do that.

Ah that's changed then (like i said its been YEARS). They used to look up the cat with paperwork but then added the whole "must be current model year" thing. As I understood it a CARB cat expires each Jan 1st? Something about stopping people from reusing them.

There was also something you got when you took off the old cat (like you wrapped a paper around that and it sat forever in the shop!). They didn't want us replacing good cats/ something about trying to catch modifying. Idk I took only 2 not the 3 classes for the cert and it's now usless info for me (I'm in New England now). And I've never faced gross poluter so all I ever faced was just a regular re-test. I only messed with Star once and it was an easy pass (97 accord). But i did have to see the ref once (little fan in an S-10 airbox you couldnt get anymore). Him being my professor from community college made it fast and easy.

What I more meant on "looking to fail" was that while a star station can't arbitrarially fail you, the techs do really have to watch the "follow up pass rate" FPR in state lingo. If they have more than .4% of the cars they pass fail the next inspection they can lose their certification.

So that just means on a 90s car they are gonna hammer hard. It has to be good enough to pass next time so visual inspection is given a priority. They have no reason to say "na that's close." They have every reason to say, "that vacuum hose is torn, fail and retest." Or in this case, "did they slap an AZ/ junkyard cat on the car?"

I believe if they fail then retest you it doesn't impact FPR the same? But I'm not sure.

And even if its stamped with a CARB #, they can then look up the # and see when it was issued (if they do... some of that is just arbitrary). As I'm aware something like a 1-2 years ago sale date on the cat would be enough to fail it.

But i said try it, just dont be amazed if they want the paperwork. As I know junkyard cats are a no-no. It has to be new and current model year design.

I've never done anything but just buy a new one at only $500 (about the same as tires)

 

[jvanabra]

Since 2009 or so everything is electronic - there is no paperwork, and there is no ability for a tech to even know whether a car passed or failed until the computer tells them so. Used catalytic converters are illegal at the Federal level - they can only be bought by a few licensed companies, and "recertified" by even fewer. I'm not actually sure if any of those can even be sold in California, but in any case you certainly cannot buy a cat from a junkyard legally. The database of past and current EOs is published online:

Aftermarket Catalytic Converters

off-road, auto emission regulations,replacement parts, executive orders, automobiles,motor vehicles,on-road,off-road,automotive emissions,automobile emissions,auto exhaust,car emissions,exhaust emissions,auto emissions,automotive nox,automotive standards,clean cars, vehicle emissions,car...

www.arb.ca.gov

I won't lie, it took a while, but the whole thing in California is pretty streamlined and easy to stay in compliance. The only rough part is shops that can't for whatever reason RTFM and screw around with the visual portion. When you have an old car, it's really key IMO to not try and find the cheapest shop. They're the most likely to employ low-grade techs who aren't motivated to do their job right. Those places are just as likely to give you a false pass and a false fail, and that's just not repeatable.

 

[jvanabra]

Also, Paul mentioned up there somewhere to hit the cat with a thermometer... that's DEFINITELY an easy step. If the cat isn't 700+ degrees on the surface, it's not doing anything. Exhaust should be a little hotter after the cat than before, and the harder/better the cat is working the hotter it will be. If the cat is cool, it's dead... pretty much period.

 

[spydergear]

Just a side note on the exhaust lambda probe.....
Jim M posted some years ago that lambda probes rarely fail outright but more that their response times slow..... they become lazy.
I can say that I did see such response issues when I owned the last 900 but more recently, with the diesel car I have now.
When I went to swap the probe on the diesel, the dealer part supplier virtually repeated the same words describing the side effect Jim noted......response time.

I do have a brand new, in the box, Bosch Lambda sensor that I've had for several years. I wouldn't be surprised if this car still had the original O2 sensor. Maybe if I clear the codes, and recheck, I might find out if my cleaning of the connections has changed anything.

 

[spydergear]

I can verify that the smog tech has no idea how the dyno test is going. I thought there might be something on the screen that indicated that, but all I saw were indicators for speed, RPM and duration.

I go to a "Test Only" station. Probably out of force of habit. I tried a cheaper place that also did repair and it was not a good experience. The tech did not seem to know what he was doing, gave up, and told me to go elsewhere. And lets face it, there are very few places that work on C900s, let alone understand them.

I do have a cheap digital thermometer that works well if you get really close as it tests a broad area. Thanks for the suggestion. I will try it on the cat - that would be simple to do, although I might have to jack up the car. These SPGs are low!

Also, I think I will try another bottle of Cataclean. I have no idea if this stuff works, but I did pass after using it two years ago. Of course, correlation does not prove causation. But perhaps it will clean the injectors and hopefully the combustion chambers at least a little...

I also do suspect the MAF. In that respect, I do wish these were adjustable like the LH 2.2s. And the turbo seals - yet another area to look at!