Millenia S stalls after gassing up

[OldMazzies]

Thanks for the page number - it was the pressure regulator I was wondering what it was and was not identified on the other diagrams.

Well I had another set-back. Pulling the test hose off the bypass valve the upper port broke off! I should have cleaned more of my garage so I could have fit the whole car inside. The door has had to stay open and so it's been cold - at least that's what I'm blaming the breakage on. With no time to get another, I've epoxied it back together. When it's set a bit longer I'll put cable ties above and below the port. In the process I got a good look at that white screw adjustment. Mine was set to full-open, by the way.

The parts from Mazda came today. I'm going to be very careful with the rest of the hoses I move and I'm out of time to check all the solenoid valves.

[bigb]

hopefully the epoxy holds, but if not I have you covered. I have a spare bypass valve

[OldMazzies]

Thanks bigb. The way it broke it's able to get some extra support from cable ties I put above and below the upper port. We'll see how long it lasts. It actually proved quite helpful to have at solenoid valve & bracket removed, as it made some lines I still hadn't figured out more accessible.

And I just found and fixed the last problem with my vacuum lines!!! The little hose that was supposed to go to an outlet off a metal line to the purge solenoid valve was disconnected. I spent the longest time puzzling over where 3 lines went underneath a bracket - two metal and one rubber. I finally took a look with a mirror and though I still couldn't see the end of the rubber hose, I finally concluded it had to be going nowhere - and I was right . The end that was supposed to go to the outlet off the metal line was stuffed under the a bracket so it couldn't be seen. The line that heads off with the fuel lines (I'm assuming to the purge canister) was connected to that metal outlet instead, so the purge solenoid was out of the picture entirely, and the system was acting as though the purge solenoid was always open. I'll bet filling up the tank pressurized some that portion of the vacuum line and as it was relieved it caused the engine to stall.

Of course the proof will be at the next gas-up, but I think it makes sense.

[OldMazzies]

I put about 80 miles on it before clearing the DTCs, and no codes were pending, which I think is a good sign.

I can accelerate normally again, so bigb, I think your IAC valve did the trick on that problem.

We probably won't gas up for a few days to check if the one-time stall is gone after filling the tank. I'll be very surprised if it's still happening. The canister line error would still be simulated if the purge valve is stuck open, but I don't think that would make the car stall. I think some of the other misrouted vacuum tubes played a part in that.

If there is still an EGR problem reported, that will take 200+ miles for the check-engine light to come back on if it acts as before. It would be interesting to fully analyze the effect of the vacuum lines being hooked up how they were, but I don't know enough about engines and what each sensor and actuator does to be able to do that well. A closer look showed that I had been mistaken about the connections to the EGR Solenoid Valve (Vent) being reversed (#6 on the earlier post). So actually the EGR portion of the vacuum hoses were all hooked up correctly.

[OldMazzies]

Gassed it up today and there was no stall afterwards! Differences in how the car has performed are:

• mpg with about 1/3 tank doing highway miles was about 19 mpg. It had measured about 13.5 around town mpg before. My last gas-up that was almost all highway miles was only about 21 mpg. Earlier it had been getting 25 mpg for highway miles, so the IAC and air bypass problems account for the drop in mileage I think, not just the misrouted hoses.
• idles at about half the rpm it used to (I hadn't mentioned this before, and I didn't really notice it being a problem - no shifting problems like some other had reported with the P1540). Surely this must help the around-town mileage.
• the check-engine light is still off. After going through another tank or two of gas I'll have a better idea if the EGR & ABV systems are really working correctly.

Why did it stall? I'm thinking it was for one or both of these reasons, which I think both make sense the way the vacuum hoses were incorrectly connected:

1. Filling the gas tank put positive pressure into the vacuum circuit connected to the MAP sensor, causing the engine computer to receive a measurement that indicated a lot more air was going into the engine than there really was, and therefore sending more fuel than it should have in an attempt to mix the air with the right amount of fuel.
2. Gasoline fumes, due to positive pressure from filling the gas tank, passed from the newly-filled gas tank through the charcoal canister through a mis-routed vacuum line into the hose to the rear intake manifold. These extra fumes made the fuel/air mixture too lean, causing the stall.

What do I wish I had done differently?

• Not begun forcing the engine through the stall by rev'ing it upon first ignition after each gas-up. In hindsight I think nothing was getting damaged by letting it stall - it was just inconvenience to have to restart one or two few times to clear the problem condition. I'm not sure, but forcing the engine through the stall may have damaged my IAC valve. If it hadn't been for bigb's good price on a used one, I don't know what I would have done about that. Those are incredibably expensive. Of course I'm keeping the old one in case it really isn't damaged or can be repaired and sold someday.
• Used the vacuum line diagram I had downloaded from milleniatech.com last summer rather than waiting until the problems got worse.
• This problem happened at the first gas-up after the engine was rebuilt. I should have insisted that there was something wrong and asked the shop that rebuilt the engine specifically if they verified the vacuum line routing with manufacturer information, and if not, insisted that they do so. When I verified the lines, I found that there was one hose that had to be extended by a few inches to reach the correct place (the one to the front intake manifold), and I think it was not one of the lines they replaced with new hose. I'm guessing that in their haste to finish the R&R at the end they just hooked the lines up as they had been before I bought the car.
• Better yet, I would have asked for all of the vacuum lines to be replaced when the engine was rebuilt. I would have provided the 21' of 4mm silicon hose and ordered brass tee's as I've seen recommended by some on this site. Hopefully it will still be a good long time before my vacuum hoses fail, but they were so accessible during the rebuild that it was an ideal time to do it and would have saved a time and money long-term (assuming it was done right, of course).

[bigb]

glad to hear everything went well.

I'd hate to lose another KJ engine to something so stupid as a vac line. I don't know what these engineers were thinking sometimes lol

[OldMazzies]

Well, pulling it in to the garage last night, the check-engine light was on again. I got the code read this morning. It was a P0421. If memory serves I never caught it alone before. I'll give it another 500 miles before concluding the problem is now isolated to only this DTC.

I reread TSB Item Number 01-003/01 regarding overly sensitive rear o2 sensors erroneously storing these codes on some Miller Cycle engine equipped models. I notice that my model is not specifically listed in the TSB. The TSB says it is applicable to all 1997-2000 Millenia S and was last issued 02/28/2001. Also I noticed on RockAuto.com that the O2 sensor offerings from Standard have a part listed specifically for 03/05/2002 and later for rear downstream. The Mfg date of my car is 05/02. Does anyone know or care to speculate whether or not some manufacturing change effective 2001 or 3/5/02 occurred (could be an O2 sensor change or an update to the 2.3L ECU firmware) that makes the rear O2 sensors less likely to be the cause of a P0420/P0421? These codes are originally designed to determine whether or not "rear (RH) bank manifold catalyst" is defective.

I also reviewed posts to my earlier thread when I had 4 DTCs active. Applicable details are that my engine has been overheated (that's why the car was for sale cheap and it had to be rebuilt). The overheat could have affected the catalyst, the O2 sensors, or both. The shop where the engine was rebuilt had it on the computer and readings seemed normal, however. Visually they look OK or they would have required that I replace them along with the many other parts renewed at the time of the rebuild. They have checked for exhaust backpressure and didn't find it abnormal.

[bigb]

since you have been driving for a long time with incorrect fuel mixture due to your issues above I suspect your cats are all crudded up.

I recommend clearing the 421 and 431 codes if they come up for the next 6 months so that the engine has time to "clean" itself out and burn off all the **** that is in there.

just note the dates when you get the code. if the dates come at the same interval even after 6 months then it is either the sensor that is going after 10+ years... or driving rich for so long has destroyed the catalyst. a back-pressure test or inspection with a video boriscope could confirm this. with any luck the code will become less frequent or disappear altogether.

don't worry too much about this code as it is more informative than anything else.

[OldMazzies]

I disconnected the battery to reset the computer and clear the DTCs. I've wondered for some time whether clearing the DTCs with a code reader is identical to leaving the engine computer disconnected or not. I'm thinking that disconnecting it from power might force the computer to start fresh on its estimate of some engine parameters and cause some "relearning" that might not occur if a handheld DTC scanner is used to clear the codes. Any second opinions?

I also did something I've been intending to get around to for a long time: comb the front intercooler fins. Last week before putting things back together I had done it to the rear intercooler and changed the foam sealing the fresh-air tube, but the front intercooler was in worse condition with many badly-bent fins. I used the 15 tooth-per-inch radiator comb out of the radiator comb kit I bought last summer. FYI, I discovered that the toothpick end of a plastic "flosser" was extemely useful for quickly bringing the most badly bent fins into rough alignment so that the fin comb would fit in. A small screwdriver was also helpful but a bit dangerous since the fins are so delicate.

So this makes me wonder what effect would a difference between front and rear intercooler performance cause to the engine? The front one was bad enough that a mechanic looking at the car for another reason commented on it and said that it was probably significant.

I'll keep track as you suggest, bigb, of how quickly P0421 comes back and watch for other codes as well.

What back-pressure test are you referring too? About a year ago when I took the car to the machine shop that rebuilt the engine they did some initial checking due to the DTCs, and P0421 was one of them. The did test for back-pressure and told me it was normal and the cats weren't plugged. I don't know exactly how they did it, but I understand it involved removing the O2 sensors and testing pressure at those points. They didn't tell me the numbers and how the measurements compared to what they are supposed to be. Does that sound like the right way to verify the cats weren't plugged? They warranteed the engine for two years, so they have incentive not to let it fail. The first few oil changes were mandatory to do with them so they make sure it was done and could have a chance to catch any major problems early.

[bigb]

clearing DTC's and unplugging the battery do the same thing. I'd use the reader if you can just because it doesn't wipe out all the other stuff like your radio presets and such.

and yes, that's what I meant by exhaust back pressure test. if you've already had it done and the cats aren't clogged then I'm certain some hard driving will help made the codes go away if not less frequent. if it still bothers you replace the sensors.