Home › Forums › Stay Dirty Lounge › Service and Repair Questions Answered Here › Cold starting problem 99 Caravan 3.3
- This topic has 36 replies, 7 voices, and was last updated 11 years, 8 months ago by college man.
-
CreatorTopic
-
January 3, 2013 at 6:23 am #488212
Very frustrating issue occurring with my 99 dodge caravan, 3.3L automatic: as soon as the outside temp dropped down to about -10C in November, the van refused to start. It turns over and over and over, but will not start. Gas supply is good, spark is good (replaced coil pack). When the weather was warmer, it started and ran fine. I’ve noodled around on Dodge forums, and found it to be a common problem (not necessarily during cold weather) but no one seems to have come up with a solution. Any members had and solved this issue? Any suggestions, anyone? Please!!!
-
CreatorTopic
-
AuthorReplies
-
January 4, 2013 at 6:14 pm #488448
[quote=”Digger89L” post=43835][quote=”johnbkobb” post=43824]Do you have winter blend gasoline where you’re located in Canada? I thought P0300 for Chrysler is multiple misfire (more than 1 cylinder misfiring during the same cycle) code? Check the compression of all the cylinders (dry and wet tests).[/quote]
Apparently all gas sold in Canada after a certain date in the fall is “winter blend”. Again …not sure why winter compression would be much different than summer compression. Can you explain that?? Will do a compression test once I can get it running and into a shop.[/quote]
Compression is compression whether it be taken in the summer or winter it doesn’t matter. The vehicle does not have to run to do a compression test. Low compression can cause misfire and you seem to have multiple misfire problem if you have a P0300 code. You have not told us what the fuel pressure at the fuel rail is yet. Unless we know what that is and what the specs are we do not know whether in fact the question of are you getting enough fuel has been answered. If the proper pressure is not available to the engine it will not start or run. Just hearing a pump run or seeing some gas come from the Schrader valve on the rail when the valve pin is pushed DOES NOT mean your fuel pressure is what it should be for starting and running this van. I’ve got a 1990 Caravan and the only time it hasn’t started since I’ve owned it was when the fuel pump went bad just after a fill up at the gas station probably 13 years ago.
January 4, 2013 at 6:22 pm #488452Scanner Danner I just have to check dailey for your new you tube videos.!!!
http://www.youtube.com/watch?v=yjQY613sql4 Scanners last video may give poster a little thought.January 5, 2013 at 4:38 am #488595You don’t have to change spark plugs every time it doesn’t start. But people generally neglect to ever bother to do them and the wires with it that very well could still be factory. So if you don’t know for certain when they were changed do them. If you’re just going to throw parts at it, throw parts you’re going to need anyhow. Several of the guys are mentioning that fuel pressure must be in spec and that is completely true. Doesn’t sound like you have checked it yet. The fitting is very easy to get at on this and a cheap fuel pressure gauge with the universal schraeder fitting is what it takes, so no pricey adapters. Since you have spark we know your crank sensor is working, as the signal from that is what the PCM uses to trigger the injectors and the coil. Your fuel pump will run for 2 seconds when the key is first turned to the on position, and then will also run when it sees a signal from your crank sensor. So while you are cranking the pump should run if the pump and related wiring are good.
Its not enough to have fuel pressure. You have to have enough fuel pressure. Its not enough to have spark. You have to have a hot enough spark. Its not enough to have compression, you have to have enough compression. You need to test for these things to find out what you are missing or you are just wasting your time and money to still have a vehicle that doesn’t run. You need to test it. You can look for a crank signal on a vehicle without a tach by using a scan tool that can view data and watching the RPM pid. If you suspect a bad coolant temp sensor you compare the readings of the IAT and the CTS pids to see if you are reading way off. But the key is to actually test it instead of guess.
January 5, 2013 at 6:41 am #488626Okay, guys. Very much appreciate your patience and continuing input. I’ve got my action list: change plugs; change fuel filter; buy a fuel pressure gauge; and in the meantime do a compression test (have the gear). I do have a logistics problem right now: I live in a small town without mechanic services, but I do have access to an indoor heated shop about 20 miles away, so I have to arrange to have my van towed there (ABSOLUTELY hate working on a vehicle outdoors when its -15C). I will report progress. HOWEVER …still waiting for someone …anyone …to explain why this would happen only when it got cold out …????
January 5, 2013 at 2:46 pm #488714My first thought was coolant sensor. Its value changes with temperature of its surrounding. When temperature changes it has a changing resistance, which then comes back as a input to pcm.
January 5, 2013 at 4:23 pm #488718Like I said, I saw this on a lot of Caravans that needed plugs and wires. Don’t bother testing the wires, just replace them. You’re more than likely going to wreck them getting them off the plugs anyway. I had 13 vans in that fleet. Some of them had even been driven a bit in the night shift, ran fine, no issues. They sat out when the temps got cold and BAM, no start. Wouldn’t even try to start. You’ll find out when you get those plugs out. Get Autolite plugs for it and nothing else. The Autolite part number is 646. Gap them at .050 and put a little dielectric grease in the plug wire boots. You need a hotter spark the colder the air gets. As the plug wears its gap opens, requiring higher KV to fire hot enough. The wires deteriorate and reduce the amount of KV that can get through just like squeezing a garden hose reduces how much water you get out of the end of it. Eventually as you keep dropping temps down you will reach the limit where the spark can no longer jump the plug gap with enough intensity to heat up the fuel to its flash point and ignite it. It’ll make a signal on a spark tester, but it won’t light gasoline.
This is what may be the very scenario for you. I know what you mean about not laying under it when its cold out. I’ve lived in the north my whole life and I refuse to be crawling around a snowbank to work on a car.
January 10, 2013 at 8:55 am #490395UPDATE: Okay ….it was -3C here for the last 2 days (that’s +26F) so two days ago I plugged the block heater in (still not able to tow the van to a shop) …and left it plugged in for two days. This evening at 10:15 PM I went out and checked the situation. The top of the engine was actually warm to the touch (we don’t “normally” plug in the block heater until its in the -25C range (-13F) around here) …and the motor started right up!!! Runs like a top!! Smooth …no misfires …nothing. Put the OBDII code reader on it …NO CODES whatsoever. So …now where do we stand? Spark plugs and wires ..etc? How many votes do I get for Coolant Temp Sensor? The Dodge shop manual says: ENGINE COOLANT TEMPERATURE SENSOR—PCM INPUT
The engine coolant temperature sensor is a variable resistor with a range of -40°C to 129°C (-40°F to 265°F). The engine coolant temperature sensor provides an input voltage to the PCM. As coolant temperature varies, the sensor resistance changes resulting in a
different input voltage to the PCM. When the engine is cold, the PCM will demand slightly richer air/fuel mixtures and higher idle speeds until normal operating temperatures are
reached. The engine coolant sensor is also used for cooling fan control.January 11, 2013 at 3:02 am #490588keep us posted if the cts solves the problem.
January 15, 2013 at 5:13 am #491834To be honest I was thinking coolant temp sensor from the time I read your first post. Everything you mentioned in your last post is true. If the computer doesn’t see an accurate coolant temp reading it can’t properly calculate a fuel mixture and in some cases the engine won’t start. If it’s cheap you might just want to go ahead and replace it. I’ve seen sensors that just weren’t giving accurate readings but still had resistance. To know if it’s giving the correct resistance you need to measure it’s resistance at a range of temperatures and compare it to spec. That said if it’s cheap enough at that milage why not give it a shot to see if it helps the situation.
Keep us posted.
January 15, 2013 at 5:59 am #491861Yeah …I started phoning around here locally (in Western Canada) for a replacement CTS …prices ranged from $48 to $87 for aftermarket parts. P***ed me off, when they sell on line for as little as $12 plus shipping. I’m in no huge rush, so I ordered one online for $24 including shipping. Will pop it in as soon as it arrives, and report progress (if any).
January 19, 2013 at 2:18 am #492865Keep us posted. Also, not all parts are created equal so be sure to go with quality. In fact, that’s the reason I made this video.
February 26, 2013 at 8:21 am #503316Well ….finally got some time and some decent weather (-4 C) today, and was able to install the new coolant temp sensor (after testing it in a jar of hot water to make sure it was working according to the specs in the shop manual). I purposely didn’t plug in the block heater so as to ensure I was working with a cold (but not frozen solid) engine. I did charge the batter for 4 hours to ensure it was topped right up. Good news ….bad news: the sensor is working as it should, but the engine would still not start!!!! I have it on the block heater now, and will test it again in 4 or 5 hours. Remember: it started before with the old temp sensor after being on the block heater for several hours. Still a major puzzler. I’ll report back.
March 3, 2013 at 9:22 pm #504338Well, to follow up on my last post (5 days ago …no one apparently noticed): after changing out the Coolant Temp Sensor (CTS) for a new one, the cold engine would still not start …so, I put the battery back on the trickle charger, and plugged in the block heater, and left it for about 10 hours. Engine started right up …ran smooth …no codes. (just as it did with the OLD CTS) So, my conclusion is that the problem is NOT the CTS, as many of us suspected it might be. Back to square one. BTW: after I went through the minor trouble of installing the new CTS, I had one of those DUH! Moments: I could have simply left the old CTS installed, and unplugged the plug ..then plugged it into the new CTS to test it !!! Oh well.
So, now my grey matter is challenged again: what other sensor or component on the engine block would be effected by a 20 or 30 degree change in temperature? What other electronic component attached to the engine block would “expand” with warmth and re-connect a bad connection in a circuit board …and “shrink” with the cold so that it would keep a bad connection? Hmmm: crankshaft position sensor (CPS). So, my next challenge will be to replace the CPS ..which is little harder to get at than the CTS, seeing as how my van is surrounded by 2 and half foot snowbanks in my drive way.
Just to re-cap for the late-comers: van won’t start when its -5C or colder; WILL start if the block heater is plugged in for several hours …but that isn’t much help if I can’t keep it plugged in wherever I go; I have NOT changed the spark plugs, the distributor and rotor, the coil, the spark plug wires, the PCM, the MAP sensor, the throttle position sensor (TPS), the battery, the ignition switch, the fuel pump, the relays, etc., etc. Here’s why: the Shop Manual gives this info for “startup sequence”:
IGNITION SWITCH ON (ZERO RPM) MODE
When the multi-port fuel injection system is activated by the ignition switch, the following actions occur: The PCM determines atmospheric air pressure from the MAP sensor input to determine basic fuel strategy. The PCM monitors the coolant temperature sensor and throttle position sensor input. The PCM modifies fuel strategy based on this input.When the key is in the ON position and the engine is not running (zero rpm), the Automatic Shutdown (ASD) relay and fuel pump relay are not energized. Therefore battery voltage is not supplied to the fuel pump, ignition coil, fuel injectors or oxygen sensor heating element.
ENGINE START-UP MODE
The following actions occur when the starter motor is engaged: If the PCM receives the camshaft position sensor and crankshaft position sensor signals, it energizes the ASD relay and fuel pump relay. These relays supply battery voltage to the fuel pump, fuel injectors, ignition coil, and oxygen sensor heating element. If the PCM does not receive the camshaft position sensor and crankshaft position sensor signals within approximately one second, it de-energizes the ASD relay and fuel pump relay.
The PCM energizes all injectors until it determines crankshaft position from the camshaft position sensor and crankshaft position sensor signals. The PCM determines crankshaft position within 1 engine revolution. After determining crankshaft position, the PCM begins energizing the injectors in sequence. The PCM adjusts injector pulse width and controls injector synchronization by turning the individual ground paths to the injectors On and Off. When the engine idles within + – 64 RPM of its target RPM, the PCM compares current MAP sensor
value with the atmospheric pressure value received during the Ignition Switch On (zero RPM) mode. If the PCM does not detect a minimum difference between the two values, it sets a MAP diagnostic trouble code into memory.Once the ASD and fuel pump relays have been energized, the PCM: Determines injector pulse width based on engine coolant temperature, MAP and the number of engine revolutions since cranking was initiated. PCM monitors the engine coolant temperature sensor, camshaft position sensor, crankshaft position sensor, MAP sensor, and throttle position sensor to determine correct ignition timing.
So …my somewhat logical mind (forget about the CTS incident mentioned above) tells me that NONE of the other components mentioned in this sequence can be directly effected by the ambient temperature of the engine block. The CPS is the only component directly attached to the engine block …..so, that’s where my current attention is focused.
March 4, 2013 at 12:42 am #504371Check the wiring harness connectors for corrosion. Just had a P0118 and the problem was the ECT sensor pins were corroded but the 3 pin wiring harness connector that plugs into it was even more corroded than the plug on the sensor. Nothing was wrong with the sensor, just the pin corrosion. Cleaned both of them up and applied some dielectric grease and put the scanner to the OBDII plug and watched the coolant temperature rise from 30F to 200+F and the coolant temp sensor voltage drop from 5V to 2V. Even improved the cold start idle. Thanks for your update.
March 4, 2013 at 12:53 am #504378cts did not solve it. ok- usually a cps fails
when hot not cold.could still be suspect. The
only other suspect that is attached to the block
would be your ignition coil pack.I know you say
you have spark.But has spark been tested when it
won’t start? what I’m thinking is when the bock
heater is plugged in the ignition coil pack is fine.
but when not plugged in and cold.Thats when the effect
takes place. -
AuthorReplies
- You must be logged in to reply to this topic.