speedin

Vehicle: Farnbacher Loles GTR (Boxster with a 3.8l M97 X-51)

Story: Spent the morning and afternoon enjoying a spirited mountain drive, car received a thorough workout, then drove an hour back home cruising on the highway. Car ran perfectly all day and pulled strong. On my way back I stopped off at the store just next to my house.

Problem: After coming back out to the car after about 20mins the car started then within a few seconds started misfiring terribly. Being as close to home as I was I limped it back to my garage.

Knowing that the hottest the engine was going to get was that period while I was in the store my first thought was that due to the high heat something electrical kicked the bucket while I was in the store. For reference it was about 80*F.

Came home, read the codes (see Figure 1 below) then cleared them. Tried letting the car cool off overnight, next day same issue and codes (see Figure 2 below). So I started the following...

1: Swapped bank 1 and 2 Cam Position Sensors - Problem remained
2: Swapped plugs and coils from bank 1 to 2 - Problem remained
3: Tried reving the car up higher in the revs to see if varocam oil seal issue - still missfire around 3500-4000.

One interesting thing to note is if I clear the codes the car will start and sound normal for a few seconds, then slowly it devolves and continues to get worse. Tomorrow I am going to check over some of the wiring to the cam sensor and generally poke around the top side of the engine. This weekend ill do a compression test as well, and maybe shove the boroscope down into the cylinders, but mine cant look back very well, so its hard to see valves.

But im starting to run out of ideas. I would like to inspect the timing chain, but I cant see a way to do so aside from removing the valve cover, maybe oil pump removal? But I would be shocked if the chain was off, seeing as how it made it home from the store and to 4000 rpms without catastrophe. I also inspected the pistons while swapping plugs and coils around, no markings or scratches. But... these cars are always full of surprises.

Any ideas or help is appreciated.



Figure 1: Codes driving home from the store





Figure 2: Codes after a short start and idle

Petza914

See if if runs any differently with the MAF disconnected. Seems like the misfire codes that were specific to a cylinder were all from the same bank, which may be a clue. Could be O2 sensors for that bank. Look for a vacuum leak somewhere on that side.

After racing up the mountain one day with the car running perfectly, I parked it for golf, went to leave and it wouldn't even idle. Turns out one of those cheap plastic vacuum check valves had broken in half, I'm assuming it was on the verge of failure, and sitting in the hot engine bay after the spirited run finished it off. Replaced the little valve and all was good. Look for an issue on the intake manifold for that bank too. If the MAF measures X amount of air coming in, but there's a leak downstream from the MAF where it's sucking in a lot of unmetered air, it could do this. Might also do a 10 minute battery disconnect to reset the fuel trims and see what happens.

That's all I've got for you. Let us know what you find.

speedin

Ok... so I checked all round the vacuum points and sensors. Nothing obvious. Tried adding some octane booster as well to rule out bad gas.

I dropped the borescope down the cylinders and everything looks normal enough.

Ran a compression test and I think ive found a clue. Bank 2 was consistently lower than Bank 1. This points me towards something in the timing and varocam. The compression results are as follows

BANK 1 ///// BANK 2
3 - 228 ///// 6 - 195
2 - 216 ///// 5 - 198
1 - 220 ///// 4 - 186

Can the variocam be off without throwing a code?
Does anyone know a way to take a look at the timing/chain for inspection? All I can find is pulling the valvecover using that special cam holding tool. Even then they say there is a risk of snapping a cam.

Macster

My tech told me that misfires on one bank can be due to something wrong with the VarioCam Plus system. What can happen is if this doesn't work right the DME will change the fueling of the cylinders of that bank in order to get the reading from the #2 O2 sensor it expects. In some cases the fueling can go too far afield and msifires can result.

(This came up as my Boxster developed a very rough idle with rather big RPMs swings. No CEL but a pending code pointed to the VarioCam system. I checked fuel trims and one bank the affected bank had wildly swinging fuel trims. When I mentioned this to the tech he then told me what was going on. The DME was trying different fuel trims to try to get the expected O2 sensor reading from that bank.)

I can't guarantee this but I would expect in this case there would be a least a pending error code pointing to a problem with the VarioCam Plus system.

The VarioCam Plus system has at least two ways to fail. One is the solenoid (or actuator) goes bad and intake valve timing doesn't change when it should or doesn't change enough. With the proper Porsche Diagnostics computer one might be able to monitor cam timing vs. cam advance signal and see some difference.

Another way is if the low/high lift feature doesn't work right. If one or more intake valves do not properly engage low lift mode these valves open the full lift amount when they should be in low lift mode. This lets in too much air and the cylinders can misfire.

The way to detect this -- based on a section of the factory 996 Turbo manual that covers misfires arising from low/high lift problems -- involves monitoring the O2 sensors and the low lift active state during a road test. If this is found to be the case then all the intake lifters of the affected bank are replaced. This is because the problem is attributed to "dirt" and it is better to replace all vs. having to go back in again maybe more than once as other lifters fail over time.

speedin

Macster, thank you for pointing me in the direction of the VarioCam system. I started swapping its sensors bank to bank tonight and believe I have found the culprit.

When I swapped over the "solenoid for variable cam adjuster" 996-105-303-06 the misfires and code traveled onto the other bank with it. Now doing some research, and for future reference, I read in multiple places to replace these in pairs. Apparently if you just swap one the load becomes higher on the remaining old solenoid and quickly fails.

So I have two of these on order and will update this after install, hopefully with good news.

speedin

Solved : It was those variocam solenoids. Two new solenoids later and everything seems to be working again.

Thanks again for all the help.