Before I start describing the problem, I’ll beg for your patience: 1997 SL SOHC, 186600 Km, standard transmission; in Toronto (Canada) weather; and I love this car. Last December, engine was having high idling speed (1550 to 2000 rpm) when stopping at traffic lights, slowly coming down to its normal (approx. 700-900 rpm) after 4-5 seconds. As Emission Test was near, and this symptom had been identified in this forum as defective Idle Control Valve, (fearing that this idling could hamper passing the test), I made the decision to request my mechanic to replace it. After he replaced it, the car would not start; so he restored the old one, and the high idling rpm “seemed” to disappear, But a new problem appeared: ERRATIC startup (here is where the request for your patience comes in): There is NO PROBLEM to startup: a) In the morning, b) When the engine has been OFF less than 30 – 45 minutes, or c) More than 1.5 (one and half) hour. The problem appears when engine has been OFF for about one hour. Then, the engine cranks (what I believe is indicative of battery OK), but no starting. I suspect it is something related to fuel system; and the scenario is further complicated because the failure seems more frequent when the gas tank is below ¼ of full (approx 10 liters). So, my questions are (provided my assumption of fuel system is correct): Is it there any VERY SPECIAL precaution in removing / mounting / exchanging the ICV? Is it possible that other factor is provoking this failure? I want to have this problem solved, as I fear it may impact the integrity of the starter motor. Your ideas / suggestions / corrections will be very much appreciated danielR
When did you replace the fuel filter last? Did you use any 'drygas' to remove water in the fuel system? Water will plug the fuel filter. Do you get fuel pressure at key on? Do you get fuel delivery (measured with a bleeder at the fuel test port) during cranking? If the car has a crank angle sensor, have it checked. Defective coolant temperature sensors (CTS) and EGR valve leakage will all contribute to high idle RPM. Consider that while the idle air solenoid (IAS) was being changed, a flaky harness was disturbed. Check for abraded wires and corrosion on the connectors. Grounds, Grounds, Grounds! Older cars have a propensity for loosing good chassis ground connections. The cranking current (about 300 Amps) can upset things if there is a flaky ground somewhere. Oppie '01 lw300
I agree with all Oppie said but especially that poor chassis ground connections can cause all sorts of strange and difficult to diagnose and explain problems, many of which can be intermitant and temp or weather dependant. IMHO a COMPLETE cleaning of ALL chassis ground connections is one of the most important preventative maintenance operations that should be performed on all vehicles and should be the first task when looking for the cause of ANY electrical problem. Saturn S series cars have several chassis grounds, IIRC there are 2 on the engine rear above the clutch housing (IIRC the engine rear bolts require a deep socket and probably significant force to break free due to probable galvanic corrosion inside the engine block, which may be the cause of poor electrical connection, use anti-seize on threads when reassembling) and 2 or 3? located on the left inner fender body sheet metal at least 1 of which IIRC requires the removal of the battery tray for access. All of these connections should be disassembled, cleaned (small wire brush & sandpaper)and IMHO greased to prevent or retard further corrosion. I use any coppercoat anti-seize thread compound as the anti corrosion grease, but any grease should work. This job will take 1-2 hours to complete properly but requires only basic tools and can be scheduled for a sunny warm and dry day and easily combined with an oil change. Both jobs can easily be completed by any driveway mechanic. Good luck, YMMV
I had something very simular to your problem with the fast idle but it did not affect starting. I found a large split in the large vacumn line that comes off the top of the valve cover and goes down behind engine. Fixed that and the problem went away . . . .
Good point. Similarly, the EGR is a calibrated vacuum leak and is compensated for. If the EGR valve sticks open, depending on the type of valve (some incorporate a position feedback), the manifold vacuum goes to hell. Some engine management systems have a manifold vacuum sensor that looks for a certain level of vacuum during cranking in order to self test. It has to pass this test before injectors are enabled. On my lw300, it has to crank at least a half second before it will catch. I have had other EFI cars that (especially when hot) would kick over on the first revolution of cranking. But I digress... If you have one of those cars that does a self test before enabling fuel injection, any vacuum leak may be inhibiting the starting. If it's marginal, it may start and idle high. You can pinch the vacuum lines while the engine is running and see if the RPM changes. Also a vacuum gauge applied to the manifold port should read at least 20" and steady on idle. If you haven't changed the PCV valve, do it. It too is a calibrated vacuum leak and if the pintle sticks open, will bypass too much air. If all else fails, have the catalytic converter and muffler checked for backpressure. Not unheard of for an older car to shatter the honeycomb structure inside the cat converter and wind up plugging the muffler. This would also give you limited power for acceleration.
To ALL; Thank you VERY MUCH for your comments, hints and recommendations. I will discuss them with my mechanic, and will keep you informed of future developments, BTW; I systematically change the PCV valve before the emission test. Peharps, this last I installed is faulty. Will replace it and shall see results. Thanks again. danielR.
danielr wrote: About 20 years ago I had a 1978 BWM 320i. It started exhibiting a very similar symptom. The interesting thing was that the 1 hour + translated to going to dinner with a family member. If I went to work, car sat for 8+ hours. If I was doing errands, usually less than one hour. On a date, usually over an hour and a half. But in one month when my sister and then a couple of weeks, brother, came to town, the car would not start after dinner. The problem was the fuse for the fuel pump. It was slightly corroded and at a certain temp would not make a proper connection. Apparently the hour was enough for it to cool down to the point where the connection was bad. A piece of emery cloth to polish up the ends and the problem went away.
hello everyone; i discussed the post and your answers with my mechanic. he lifted the hood to have a look at the ICV area, and showed me a leak in the fuel system. there is a black bakelite right angle which couples from a rubber hose into the (apparently aluminum) front interface of the fuel rail (looks as a manifold and runs behind the SOHC engine block, delivering the fuel into the cylinders). IT IS WET. we ignore if what is leaking is the bakelite-aluminum interface, or the bakelite-rubber interface. he says that the best would be to CHANGE to whole fuel (manifold) rail (C$ 350 + labor). because a coupling bakelite-aluminum is flimsy and could be location of the leak. he did not touch it, nor explore the area. the part must come from USA location. a personal note: speaking as an (ex) design engineer (i am retired now), my duty was 5% inspiration and (compulsory) 95% perspiration, to deal with the frustrating task of fixing / improving the (first pass of) design to make it work. apparently, for this mechanic is more cost effective to replace instead of researching to fix. i would not oppose to replace unless IT IS STRICTLY NECESSARY. searching the web for fuel rails for saturn, led me to stores/ warehouses which offer “MR GASKET PUSH ON RUBBER HOSE, BLACK -- -8 AN, 4 ' Long, Designed To Be Used With Mr. Gasket Push-On Swivel Fittings, Can Withstand Up To 250 PSI”; for U$S 23.00………… it is only the hose, but gives an idea of possinle alternatives once the REAL ROOT CAUSE of the problem is located. so, my new questions are: a) as there were previous replies mentioning drop of pressure in fuel system as possible culprit of the erratic startup problem, is it possible that this leak to be the cause of it? b) i don’t know the designation of THAT black aluminum angle, neither if it is a coupler between the rubber hose OR the front end (affixed to) the fuel rail manifold; so i can not seek it online. all suggestions, references, comments, advices will be greatly appreciated. danielR.
Hello fellow design engineer (I'm an EE). I would get another opinion. Yes, the fuel leak has to be fixed but whether replacing a whole assembly instead of doing an appropriate minimal fix has to be evaluated. Besides, 'wet' will not likely be your root cause. I had Saturn try to find out why my L300 wouldn't start. Turned out to be an intermittent crank sensor but they surely charged me for a very minor leak in the fuel test port (which did not leak when the safety cap was on). I had damaged the schrader valve when attaching a makeshift fuel pressure gauge... Point is that lots of techs go for the obvious rather than looking for root cause.
Hi Oppie; Thank you very much for your insight. I am in the process of getting "that second" opinion, and get the leak fixed. I am trying to to avoid (as much as possible), going to the dealer shop. Here in Canada, they are extremely expensive.
I've been considering again getting the Autotap tool. Had considered getting this a few years back but didn't have a laptop at the time. Just got an email from autotap telling of their new product. Looks to be the same hardware but an entirely new software package. www.autotap.com For $200 and if you have a laptop, there is some impressive diagnostic ability. Not just a code reader but can read and display all the engine and BCM data. ABS and transmission codes are not supported (which is not a deal breaker but I'm not happy). Other thing is that aside from resetting the Check Engine Light, it's a read only device; you can't change any program values or exercise solenoids. Still, the data you can read a wealth of data. Good Luck, Oppie White Plains, NY USA
Last Monday I took the car to this "second opinion mechanic". He fixed the leak; extracted the black bakelite elbow, and readjusted its fit. He noticed that EGR valve looked relatively new (it was actually changed October last year, after repeatedly SES notices and P0404 DTCs), mentioned that probably, when they changed the EGR somehow might have banged into that coupling. There is an O ring inside which perhaps was not fitting well. When trying to start up, IT DIDN’T. He put his (professional) ODB scanner, which reported DTC 1599. Checked spark into plugs, and here wasn’t. He indicated that it could be crankshaft position sensor [CPS], or the ignition module or the computer. By then, I left to have him working without me scooping around over his shoulder (I believe it is very annoying for some people, and didn’t want to take the risk). Something he mentioned: He said Saturn was an invention from GM to confront Japanese penetrating USA markets. Then, they put together (IIRC) a Suzuki engine with some other GM not too vital parts (I don't remember the tranny make he mentioned), and the plastic body. I am mentioning this because he expressed his confidence with confronting the problem based in his familiarity with GM “automotive architecture”, and the possibility of swapping parts rescued from some other scrapped GM cars (BTW, he showed me his stock of EGR valves rescued from other “deceased”). Furthermore, he has the Mitchell Database System, with all manuals data accessible in his PC. Phone call Tuesday: He had tried another CPS, but no luck; next, he had asked to borrow an ignition module from a friend, and was awaiting for it. Phone call Thursday: No luck with module. Was checking for corroded connections into the PCM (computer); mostly because when connecting the OBD scanner, it displays 63 RPM even is the key has not been turned to turn on. He says that are generally about 300 RPM during initial crank. As a retired EE, I saw my good quote of anomalies originated from spurious charges that get stored in some (not properly terminated or loosely connected) ports in CMOS technology devices (I believe that must be the technology of this PCM). These devils are, usually, very temperamental and temperature, moisture, shock (and you name it) dependant. If all checked connections are OK, them we’ll have to conclude that it might be the PCM. He says it’s difficult to get another used one to swap, first because my car is standard, and second because Saturn is history, and few cars available nowadays. Next phone call will be tomorrow, we’ll go from there. YES, I have my fingers crossed!!!!!!!! danielR.
_________________________________________________________________ Car Fixed: Before categorically blaming the PCM, this mechanic checked all electric circuits leading to PCM, but found none deemed to be the culprit. Lastly, replaced the Crankshaft Position Sensor for another used one. Success!!. Before leaving this posting, I want to express my appreciation to all those who contributed with their comments and suggestions. Best regards. danielR.
Glad to hear you finally got a fix! The Crank sensor, while it has diagnostic codes associated with it, rarely seems to throw those codes when it becomes defective. It is the 'master' timing element for all power train controls. When it gets flaky, it's more likely to cause all sorts of 'phantom' codes that are totally unrelated to the crank sensor. As I understand it, the sensor is a magnetically biased coil that senses ferrous gear teeth on the crankshaft. Crank has a nominal 36 tooth pattern with one tooth missing to indicate TDC. PCM does a clock regeneration and missing pulse detection to detect shaft position. (as a fellow EE, you should appreciate that...) Best Regards - Oppie
