What's causing my rough idle?
 

Symptoms:

Not passing smog - gross polluter

Noticable richer smell in the exhaust.

idle is usually rough, however there has been a few instances where it would be smooth for a few minutes.

Vacuum is still the same

The car is running stock ECU with 1300cc secondary injs so I can't imagine that detonation is even an option, especially when I'm not boosting more than 5psi.

Electrical?

A few services ago (I take the car to my local rotory gurus) I was told that my LIM gasket was torn and on it's way out. I could expect the idle to get rough, and soon it would need replacing. Stock engine at 120'000 kms (75k miles)

Sure enough about 10 weeks later the idle starts getting lumpy. It was similar to what you describe - the car would idle fairly smooth, then occasionally would go through a patch of stumbling, with the revs jumping around by 200ish rpm. so I take the car in to get the LIM gasket replaced. Turns out I had a broken wire leading to my O2 sensor as well.

Both now fixed, and the idle is nice and smooth again :)

I can't give you advice on how to check the LIM gasket unfortunately, but maybe somebody else can?

How do you get the oem ecu to properly controll the 1300s? Last time I checked it wont. Mike

Exactly, although that really shouldn't effect idle when the secondaries are offline. Bad idea though, I don't see the car is even drivable....

Did you remove the air pump?

if it's idling rough, that isn't the smell of rich. raw fuel burns your eyes--too much fuel just gets you high. when you're misfiring, it's basically the same effect as running lean.

what did you fail for--HC's, right? can you post the sheet?

Well, it's my understanding from the MAzda mechanic that it's only an issue under WOT.

At cruising under 3k RPM with little to no boost the secondaries shouldn't be activated.

The stock ECU is temporaly in there for my smog test.

Otherwise I would have my PFC installed.

It's drivable.

Airpump and everything that is smog related is still on the car.

I changed O2 sensor and plugs and no changes in symptoms.

1867 RPM Max allowed HC ppm is 88; gross polluter is 275 and I measured 317
2912 RPM Max allowed HC ppm is 53; gross polluter is 225 and I measured 390

1867 RPM Max allowed CO% is .52; gross polluter is 2.02 and I measured 1.38
2912 RPM Max allowed CO% is .50; gross polluter is 2.00 and I measured 3.62

1867 RPM Max allowed NO ppm is 704; gross polluter is 1936 and I measured 727
2912 RPM Max allowed NO ppm is 738; gross polluter is 1736 and I measured 310

The factory cat has 30k miles on it.

I get rough idle from time to time too, clutch in or out, comes down to 500rpm for a few seconds than the ecu says uhh no and puts it back, i do have a code 40 for purge though, gonna replace it

the stock ecu can't work with 1300cc.you have to have the pfc to control the fuel and you have the change the data for injector pulse value and injector volume.i had the same problem b 4

The smog test requires that the check engine light be present at start up............. something that the PFC fails to provide.

At idle the stock ECU should be able to handle the 1300cc since it's not being used.

I suspect the same for the under 3k RPM smog test where there is no boost involved.

with that smog sheet, it looks like you're running too rich with WAY too much timing, but i'm not sure how that would happen with the stock ecu.

i'd find a different smog place that will overlook the cel and try it with the pfc.

same results at another test only center.

The timing is untouched.

So perhaps my coil and/or plug wires are bad and part of the reading is a result of uncombusted or incomplete combusted A/F?

You have to pass smog in FL? I didn't think they had any smog regulations.

In 1996, at its first smog check, our stock 1994 FD (purchased new, worked on only by Mazda dealers) with only 19,657 miles, was a gross polluter. At that time, dynos were not used in CA; just idle and 2500 rpm tests in neutral. The initial results were:
HC Spec at idle: <130 ppm. Measured: 552
CO spec at idle: <1.0%. Measured: 3.41%

(The car passed at 2500 rpm with HC = 71 and CO = 0.01%.)

The problems, as reported by the Mazda dealer tech who worked on our car were:

1) Pinched wire (open circuit) at solenoid controlling Air Bypass Valve.
2) Vacuum hose misconnected.
3) Vacuum hose deteriorated and leaking.
4) "Other various vacuum leaks."

The State Test-Only stations at that time did not provide reports, so I never saw the results of the retest... only "Pass." However in 1998 with 38,070 miles, this time on a dyno, the results were:

15 mph - HC spec <125 ppm, measured 2; CO spec <0.52%, measured 0.02, NO spec <9999 ppm, measured 52
25 mph - HC spec <75 ppm, measured 8, CO spec <0.50%, measured 0.08, NO spec <9999 oom, measured 44.

In 2000, at 54,731 miles, HC and CO specs were the same as1998, but NO now realistic at <1436 (15 mph), <1236 (25 mph). Measured values at 15 mph were HC = 6, CO = 0.21, NO = 62. Measured values at 25 mph were HC = 11, CO = 0.14%, NO = 86.

But in 2002, only 12K miles after the "60 Kmile maintenance," at 72173 miles, the car failed again; only on CO at 25 mph. Same spec as earlier year, but measured CO = 0.61%. Tried MANY fixes including new Bonez cat, which was better on CO but now failed HC. Wound up buying a new stock cat for $1265.62, which fixed the problem.

Point in all this is that emissions test results are very mysterious and so are the cars that are tested. My Miata flew through the tests 2nd time back, but just squeaked by last time. No changes and very few miles in between tests. What is needed is maybe a test facility that will test the car, adjust things, change stuff if necessary, and retest, all at the same place. Dealers will work on a car, but decline any responsibility for failing smog by another station after their work, and they won't test the car either! And test-only stations are not allowed to work on the car! Talk about your "Catch-22"...

just to make sure-- was the motor warmed up before the test?

these days the smog tech is allowed to "preheat" the cat...
hold the rpm's around 2500 rpm for about a minute before the test to get it lit off. your sheet kinda looks like it was done on a cold engine.

--too rich
--not burning enough fuel
--too much nox

1) typically, (very loosely) when the hc goes up, the co goes down and vice versa
2) cats eat a lot of nox. rotaries don't produce all that much nox--it's the nature of the beast, so for you to fail for that also, i'd say the cat wasn't working.
you see how much your nox went down when it switched to the higher rpm test?

those are some of the reasons i'd kinda lean toward maybe the shop didn't warm it up first.

I float between FL and CA.

Right now the FD is in CA.

No it was VERY warm. I was driving it for a while and never turned the car off just minutes before pulling it in.

that's interesting... maybe the cat is just bunk.

naw.. i've got cleaner readings than that with no cat before.