Weird problem during engine starting, need advice on what to do

I have a 2002 C4S which recently developed a problem that I simply have no idea how to solve. Most of the time the car starts normally and drives just fine, no codes stored in the system. One time I had to stop by my house for a few minutes, the car was at operating temperature, I locked it in my driveway and came back in 5 minutes and heard 2 horn honks upon unlocking. I now know that this means an error is stored in the alarm module, but at the time I thought that was weird, proceeded to start the car and it had a very hard time starting. The engine was misfiring heavily and vibrating and just couldn�t start. I then went into my house to get my durametric, connected it and saw misfires on several cylinders (cylinder 3 and 5, as well as the p0300 for multiple cylinder misfire) and also a code 61 in the alarm module. I deleted all the codes and the car started perfectly. Then the car was fine for about 2 weeks (it is a daily driver). Then again I come out to the car, hear the 2 honks, and the car again has the same starting problem (heavy misfire). I can give it some gas and then it will start, but heavy sputtering and flashing check engine light, so I let off the gas and the engine stalls. I don�t have my durametric on me at this time, but I do have a wrench to disconnect the battery. I disconnect the battery and connect it back and now literally maybe 30 seconds after this horrible starting attempt, the car starts no problem. This has occurred several times and now I view those two horn honks as the dreaded omen of a no start condition. I have a small generic OBD scanner and every time it happens I scan for the misfiring cylinders and they constantly change. Sometimes it will be just a single cylinder misfire, sometimes will involved 2 or 3 cylinders, but the actual cylinder number(s) seems to be random. As for the alarm module, I�ve seen codes 61, 47 and 54, although not all of them come up every time. The car is a daily driver and I can�t recreate the problem on demand, just seems to happen randomly and so far it was only when the car was at operating temperature, never when the engine is cold. This is a total puzzle to me and seems that resetting the errors by either using durametric or simply briefly disconnecting the battery solves the problem and the car starts perfectly. I�ve checked cam deviations and they were less than 3 degrees for both banks, and when the engine is running the smoothness data appears to be perfect. All 6 cylinders are contributing fairly evenly. Any idea of what it can be? The first time it happened I thought it was a failing fuel pump and the car just not getting fuel, and since the pump was still original I replaced it along with the sending unit. I also recently replaced the coolant tank because it was original and started developing a crack. It has been about 500-600 miles since those two things were replaced.

Does the misfire problem start after the engine becomes warm? Ie. It starts/runs fine when cold???

Let me clarify. The misfires only occur during startup. Once the engine starts, I can drive for an hour, be stuck in traffic, etc., no misfires, runs perfectly. I've never had the misfire starting a cold engine. Even starting a hot engine the only time I've had the problem is when I get an alarm error and 2 honks upon unlocking the car. For example, I generally get gas close to home after ~30 minute drive from work, so the engine is fully warmed up. Since I don't lock the car during a gas stop, I've never had a problem starting at a gas station. However, if I stop at my house or a store after a long drive, arm the car, come back unlock it and hear the 2 horn beeps, then I have a no start condition. So far the misfire (and I must say that is a very bad misfire, the engine generally just sputters but won't start) has only happened when I get an alarm error. That may just be a coincidence, but a very weird coincidence. My solution for this no-start condition has been to reset the modules, which I can do via durametric or simply battery disconnect. Also, I don't know if I can even bring something like this to a mechanic because I can't recreate a problem. Can go a week or longer without it occurring. I honestly hope that it gets worse and becomes more frequent and reproducible so that I can at least try to diagnose it better, otherwise I am completely stuck.

I suspect that it has something to do with the alarm module (I believe that is is under the driver's seat). This problem only began when I started getting alarm errors and every time this start-up misfire occurred there was an error stored in the alarm module. I stuck my hand under the seat to see if water has gotten in there somehow, but feels dry. The car is also stored in the garage all the time (home and work), and I don't recall the last time it has been in a rain. Lastly, since it is getting cooler now, I probably haven't used the AC for about 2 months or so, therefore I don't think there should be any source of water under the driver's seat. Does anyone know how the alarm module controls the engine? I know that it can prevent the engine from starting, but I would imagine that if this was a security issue then it would not allow the starter to crank. The engine sounds like it is not getting fuel when this heavy start-up misfire occurs. Can the alarm module shut off some injectors or ignition coils but allow the starter to crank? Seems very strange.

I would venture a guess - the alarm/immobiler is a go/no go situation. Since this only happens when the horn beeps twice, have you checked everything, i.e., fuel lid, center console, glove box, etc.? There is something setting off the two horn beeps, an open zone somewhere. Also, does this only occur after unlocking the car? Have you tested to see if locking/unlocking causes this? What happens if you leave the car unlocked in your garage?

So this has only occurred after unlocking the car with the two horn beeps. It has never occurred if the car has not been locked. Whether it is in my garage, a gas station or anywhere else, if the car was not locked then so far I've never had this start-up misfire occur. I tried looking for some zone being open somewhere, but I can't replicate it. The last time this occurred was yesterday evening. The alarm module had code 61 stored which is for the driver's door lock. However, after resetting the alarm code, I've tried for a while locking and unlocking the door, locking and unlocking holding the handle up, pushing or pulling on the door slightly, etc, and can't get that alarm code to set again. Also, what I can't understand that even if there is an intermittent issue with a microswitch somewhere, and lets say when the car is armed it open a circuit and sets a tamper code. Then it honks the horn twice when I unlock to inform me that tampering has occurred, why would that impact engine starting? If I unlock the car with the remote and the system is disarmed, what does that have to do with starting the engine? Lastly, if the alarm wanted to prevent me from starting the engine, I suspect that it would prevent starter engagement, and the starter cranks without a problem.

I suspect the alarm and the starting are related, and I suspect the ignition switch. More precisely I suspect that it's not detecting when you pull the key out and the car is staying somewhat powered up, which may be draining the battery.

Theprf, thanks for the suggestion. I am still not convinced how the ignition switch may be involved, but it is a relatively cheap part and likely to be original, so I'll replace it when I get a chance. Does anyone know if the alarm module is involved in priming the fuel pump? Perhaps the pump isn't priming when it is supposed to and causing misfire during starup?

No one really understands the ignition switch. Its the single most item that causes random issues with our cars. The next thing i would do is investigate the alarm modual under the seat. If its wet or been wet it can cause all kind of issues. I dont think it controls the fuel pump. It has a relay thats kicked on by the ignition swich and the dme. It doesnt really sound fuel related but u can put a presure guage on the fuel system to monitor the pressure the next time u have an issue.

Does anyone have a wiring diagram of the ignition switch? I would like to know specifically what it connects to.

It connects to everything. Every electrical circuit in the car is routed through the ignition switch.

ignition switches do cause problems with these cars but never heard of similar issues you are dealing with from the ignition switch. I believe your issue is caused by the security module under the seat. We have a local company that can scan the codes in the security system and reprogram if necessary. that is what I would do since it only happens when you are unlocking the warm car and get the 2 honks. I would guess a solder or connection on the circuit board makes a break in connection when it warms up and causes the fault. does this happen every time that you drive the car for over 30 mins and lock the car and unlock? can you replicate the fault?

You could always send your ECU, Immobilizer, and key to ECU doctors (Specialized ECU repair) in Florida. If anything is amiss, they will surely find it. They can run all sorts of electrical diagnosis on those components - much more advanced equipment - than we can. Piece of mind if anything.

Turbodan, So far every time that it happened was after I heard the 2 honks upon unlocking it and the car was just recently driven and up to operating temperature. This has occurred 5 times over the last month and a half so the conditions may be a coincidence, but that is all the data I have to go on. The problem is that I can't replicate the issue on demand. It occurred once yesterday, and that was about a week and 200 miles following the prior occurrence. Also I was thinking that this problem roughly correlates with the temperature getting cooler and me starting to use the heater instead of the AC. I have no idea if there really is any link to this or it is just an irrelevant coincidence, but perhaps some connection was close to failing and with the heater running this warmed up this bad connection causing the intermittent failure. I think the heater can potentially warm up the ignition switch since there is a heater duct right below it or perhaps the alarm module with heater blowing on the floor. However, because the problem is very intermittent and I can't reproduce it on demand, this makes it difficult to diagnose. I decided that my next step will be to replace the ignition switch (I already purchased it). I figured that it is cheap and the one in the car is likely still original, although I don't understand how the ignition switch would cause something like this. I will try to find the time to replace it this weekend and then report back with an update. Again, since I can't reproduce the problem, I think that I would need to go a few weeks without an issue to confirm a fix, or I might find out very quickly that ignition switch replacement was not the issue.

I would have a good shop load test the battery. On both the Porsche and my BMW if the battery gets tired,,, everything goes completely crazy in the car. Both Marques have the same odd after shutdown electrical activity that can drain a tired battery. Starts out subtle and gets steadily worse..

pdxmotorhead, thank you for yoru response. I didn't mention it but the battery is about 1 year old. The alternator output is fine (starts out at ~14V when cold and then slowly declines as the engine warms up and stays around 13.4 - 13.5V at full operating temperature) and the starter cranks at normal speed so I think that battery is unlikely to be an issue here.

This happened to me once. I thought my car was going to blow up. But! I also had just put my ignition key on a shiny new (and delightfully expensive looking) key fob while leaving clumsy immobilizer gizmo on old key fob in junk drawer. So. I retrieved immobilizer, put it a little lunch baggie, and threw it in the glove compartment where it now rattles around when I go over small bumps or turn sharply. But! I don’t think my car is going to blow up on start up anymore.

Dagvik, Thank you for your response. Can you please explain what you mean and how you solved the problem. I don't understand what you are saying.

My car came with one key (with lock\ unlock remote buttons) and a weird little plastic fob thing on one ring, the fob thing evidently has something to do with an anti-theft system. The car starts perfectly with the fob in the car and produces similar symptoms to your issue when the fob was in my junk drawer. But maybe my car key is weird / a replacement and requires this fob that other 2002 996s do not. I bought the car 10 years ago and know nothing of its history. Do you have a two part key? Either way, I agree with others that it has something to do with the alarm / immobilizer.

Dagsvik, I have no idea what the fob you are talking about is. I just have a key with two buttons, one for the doors and another for the trunk. You must have some aftermarket security system of some sort.

Does it honk once upon locking the car using the remote locking? If so find out which lid/door causes the alarm system to notify you about it being not shut by opening one at a time and lock car via remote with the one lid/door open. If it honks twice, that lid/door's micro switch is fine. Yet if the car honks only once, that's the lid/door with the faulty switch. Start with the fuel flap. Perhaps your troubles are related to the evap control system, even though it shows no such codes What's the battery voltage when cranking? Edit: just read that engine cranks at normal speed, voltage will be fine then. (If voltage drops too low for normal operation (e.g. when alternator no longer charges), ABS/PSM & battery warning light starts coming on and voltage meter inside our beloved 996es show how low it actually is)

Hardtailer, I don't recall hearing any honks upon locking the vehicle, but admittedly I wasn't paying that much attention to that so will do so going forward. Separately, what is a fuel flap?

Quotes source post: 17765046

Just the little door for the fuel filler cap

Why would the fuel filler door have anything to do with this problem or engine starting in general? Is there an alarm switch on it? I didn't mention it earlier but the locking mechanism on the fuel filler door is broken on my car (permanently stuck in open position). It was this way when I bought the car 3 years ago and I had no plans of fixing it unless I was replacing the vent valve. I figured that if I ever do replace the vent valve then the locking mechanism is right there, otherwise too much work for insignificant improvement in my view. Many cars (mostly american) have no locking mechanism on the fuel door from the factory.

Hey guys. FWIW, my car came w/ 3 keys back in 2002. If there was a 4th (plastic) I can't find it.

Quotes source post: 17765219

That's how they came, one valet key and two with remotes.

Quotes source post: 17765147

yes the security modual reads all of the doors, windows, hatches and more. Although i personally dont think this is your issue, i guess it can be? I still think the modual itself has an issue. the modual will store codes and report where the issue is. Go to a decent shop with a decent scanner and see whst it says. Everything here is just a guess.

Fuel filler door doesn't have a security sensor, just a lock.

Quotes source post: 17766275

I thought it knew if the lock was extended. U r probably right. Well the rest of what i said should be correct...maybe .

@OP: Since it's such an easy check, just find out whether the alarm system honks once when you lock your car using the remote. Thinking behind the fuel door is the following: -obviously the alarm system is playing up one way or another. The fuel door (lock or microswitch we don't know for sure) could be the cause for this -evap system has an effect on fuel mixture which in turn can affect starting, which your car sometimes has trouble with. - fuel door, cap and evap system are all close together, so start there. Checking the door you might notice something is wrong with the cap, independently of any alarm issues. How do you lock and unlock the car? Always via remote or sometimes opening using the key to turn the door lock?

So quick update. I think that this may not be alarm related because I had the problem recur yesterday (engine at full operating temperature, trying to start but simply can't catch and misfiring heavily). I did a code scan using my handheld Foxwell scanner and there were no codes in the alarm module. Perhaps the alarm is a separate issue. So I then decided to do a crude test of the fuel system from the driver's seat. I know that the fuel pump was delivering adequate fuel pressure and volume just a few minutes ago because shortly before this I was driving with the scan tool connected watching fuel trims, and even when I did heavy acceleration, partial throttle and wide open throttle, the total fuel trims were very good (within 10% of zero). So I figured that I will activate the fuel pump with the scan tool, let it run for 10-20 seconds, and then immediately try to start the engine to see if that makes and difference since that was an easy thing to do right from the driver's seat. I tried that and it made no difference. Again, this is not a perfect test of the fuel system, but I decided to test a few other things as I don't have a way to connect a pressure gauge to the fuel rail. I then decided to do a simple test for a vacuum leak around the intake manifold using brake cleaner. Once the engine cooled down (since it would not start hot), it started up and I found a vacuum leak. See picture below where this elbow (circled in red) connects. When I spray some brake cleaner in that area the engine almost stalls. Where do these to hoses go? So then I looked closely how the fittings are attached and disconnected one (with the engine off). They are easy to remove by hand with no tools required. Once the hose is disconnected I couldn't get the engine started and had similar sputtering that I experienced before. I started looking through the Porsche parts catalog to figure out where the lines go and I can't figure it out. What the catalog shows is a different arrangement, like in the second picture where the line coming from the left connects directly into manifold above the alternator while my car had a blockoff plate there and instead connection to the hose on the right. See the second picture below that I found on the internet regarding the other way this line is connected in some cars. If someone can explain what these lines are, where they go, and how even a small air leak in this 90 degree joint can prevent the engine from starting. Disconnecting this hose and trying to start the engine causes heavy sputtering (like the engine is trying to start but can't). Since the plastic elbow is leaking, I will look for a 90 degree rubber hose to replace it to make sure there is no more leakage, but I am not sure if this is really the cause of my intermittent hot starting problem or an unrelated issue I just discovered. This is the setup on my car. 2002 C4S Automatic. This is the setup shown in the parts catalog and the pictures I see on the internet. The top pipe connects to the intake manifold (circled in yellow). Ignore the vacuum gauge, just in the picture that I found.

So I've done some more digging and found that the difference between these two hose routing has to do with the transmission choice. Since I have the Auto, I have this elbow connector and the additional hose. On the manual transmission cars this hose connects directly to the intake manifold. I don't understand why these differences exist so perhaps someone can explain where these hoses go and why this venting system is different on auto vs. manual cars.

I think that is the vacuum hose that runs to the brake booster (vacuum assist) in the frunk. Might check the brake system and brake booster. There is also an O-ring on that connector in the engine bay. Could be it is just the O-ring leaking. #10 here http://www.autoatlanta.com/porsche-p...section=604-00

I doubt that this is going to the brake booster because why would that be different on the automatic vs. manual. On the manual it connects directly to the manifold while on the auto it doesn't go to the manifold but goes elsewhere behind the engine.

Quotes source post: 17775255

it is 100% the brake booster vac supply. Its not different for an auto or 6 speed. It is different through '98 to 2000. Im not sure exactly when the changes were made but the new parts can be used and made to work.

De Jeeper, I was surprised as well that it would be different between manual and auto transmissions, but it is. I looked at eBay listings of 2002-2004 cars and then sorted them by manual and auto. Most listing have engine bay pictures, and you can see that in a manual cars the brake booster hose is connected directly to the manifold. In automatic it is connected the same way my car is connected. I also figured out the specific part number. The automatic car has a part called venturi tube (part # 99611017051). This part has a hose connecting to the brake booster (which is where this 90 degree angle piece goes) and the other hose connects to the rubber air hose between the air filter and throttle body. If you look up this venturi tube you will see that it is listed for M249 only (automatic cars). This part does not apply to manual cars. Also, the auto cars have a nipple on the intake hose (Porsche calls it air guide) which is where the venturi connects. That is why if you look at the part number for the air guide, they are different for manual (99611022157) and automatic cars (99611022158) . Look at the pictures of these two part numbers and you will see the difference. I am now puzzled why the brake booster hose is different between the two transmission options, but it definitely is different.

So I did a bit more internet searching (amazing what you can do in google) and I think that I have the answer on the difference in the brake booster setup between the auto and manual cars. You can always tell the difference between the manual and auto car engines based on the flywheel vs. flex plate attached to them. I am certain that the brake booster hose setup is different based on transmission. On the manual, it is very simple, the brake booster hose connects to the intake manifold on bank 1 right at the back of the engine (by back of the engine I mean the side of the engine readily accessible when opening the rear trunk). Basically the booster hose is connected to the intake manifold closest to cylinder #1 with a fitting right next to the alternator. Since the engine uses the same plastic manifolds on both sides, there is a similar port on the front bank 2 of the engine (by cylinder #6). On the manual cars the port by cylinder #6 is simply blocked off. On the automatic cars Porsche ended up doing something very different. The port by cylinder #1 is blocked off. However, there is a venturi tube ( part # 99611017051) that has 3 connections. See below picture for the venturi part installed on the engine from an auto car. The yellow tracing is the hose that goes to the brake booster. The blue hose is a connection to the intake manifold (after the MAF). Again, this connection doesn't exist on manual cars. Now if we look at the backside of this engine you see that this venturi tube is connected to the intake manifold by cylinder 6. Here is my speculation of why Porsche did this, although I am curious to get other people's perspective. I think that the goal of this is to boost vacuum in cars with auto transmission. You basically have a venturi tube between the manifold and brake booster. Instead of all the air into the engine coming through the throttle body, you are allowing some air to pass through this venturi (hose in blue outline in first picture) directly into the bank 2 manifold. I am sure that the venturi is designed such that the flow of this air creates additional vacuum on top of what you would normally get. Not sure what the rationale for doing this was, perhaps the auto car is heavier and they wanted to give it more brake boosting and this was cheaper that putting in a larger brake booster. Or perhaps if you are running wide open throttle (low vacuum) and then let go the gas and quickly transition to the brake, the auto can't spin down fast enough relative to the manual due to the inertia of the torque converter and other internal transmission parts so they wanted to create some additional vacuum for higher RPM situations. Obviously this also means that in the auto cars there are more sources of vacuum leaks. Also, any vacuum leaks in this system will likely have more impact on bank 2 fuel trims than bank 1. Something interesting to keep in mind when diagnosing manual vs. automatic intake vacuum leaks.

My comment was directed to the elbow piece that u refferenced earlier that was the source of the leak. This is a common place in our cars for a vac leak. Yes the plastic tube for the booster and how it is roughted is different http://www.autoatlanta.com/porsche-p...-05�ion=107-10 See part #22 for the elbow.

De Jeeper, the link you send doesn't work (at least for me).

http://www.autoatlanta.com/porsche-p..._911_parts.php Try this. 12th drawing down in the left, part #22

THanks, that works. I was actually looking at this part and thinking that instead of using another plastic fitting I will look for a similar sized rubber hose that I can then use hose clamps on. If I can't find one then I'll buy another fitting. The fitting may be faster/easier to connect during manufacturing, but I think a rubber hose elbow with clamps will probably produce a better seal. Separately, I wanted to find the part number for the pipe that this fitting connects to that goes to the brake booster. However, I can't seem to find that brake booster pipe on any parts diagrams. Let me know if you may be able to find it.

http://www.autoatlanta.com/porsche-p..._911_parts.php 3rd drawing down in left 5th drawing on left and right All 3 drawings have sectiins of the booster vac line

I clicked on the link you send and I don't see it. 3rd drawing on the left is a piston with rings. 5th drawing on left is intermediate shaft with chains. and on the right is the oil pump. If you see the booster vacuum line, can you post the part number.

You Tube If it's not the 3rd from the top, perhaps it's the 3rd from the...

I don't see it on any of these diagrams. If someone does see it, please post a part number.

http://www.autoatlanta.com/porsche-p..._911_parts.php dont no why the links r not coping correctly u can always go to the autoatlanta and pick your model. All the PET drawings r there, look under the "brake" tab. Some of the parts vary with model so u need to learn how to look up your own parts.