No start...timing good...leak test bad...need help

[rah rah 986]

Late fall is the last time I drove my '98 boxster. Prior to that day I had owned the car for three years. I purchased it with 56k and put 30k on it. The engine has run good from day one with the exception of cold start rattle.

One fine fall day, I got in to start the car and turned the key. The car immediately fired once, but failed to turn over. Typically, the car has always fired right up. But, on a few rare occasions, it had required turning the key a second time after an instance like this one.

I turned the key again to start the car and let the car crank for a couple of seconds. I stopped immediately when i noticed it sounded as though the car was cranking with low or no compression.

Fearing a timing issue that might be worsened by further attempts to start, I left the car parked for the winter and have now started the process of trying to figure out what is wrong. I have viewed this as a learning opportunity, and an opportunity to address some larger maintainance issues along the way while hopefully fixing whatever is wrong.

So far I have pulled the engine, drained oil, checked the timing and done a leak down test. The timing marks all line up perfectly, the oil and filter are clean, and the spark plugs look great. No oil on the plugs..totally normal in appearance.

My leak down test is not near as positive. Every cylinder is leaking between 75%-90%.
When listening for where the air is escaping, it is leaking through the exhaust on every cylinder except #1, which I can hear air through the throttle body.

Can anyone provide me with some direction on what I might be up against? I am ready to tear into this thing, but, am hoping to get a clue on just how big a learning experience this might be.
Thanks, Rick

[BYprodriver]

Have you rotated the engine CLOCKWISE 360 degrees & verified the cams are rotating also? I assume you are doing the leak-down test with the crank pined @ TDC.

[rah rah 986]

Yes. I have rotated the crank numerous times in order to verify cam position. (Always clockwise). After having found a couple of good threads about checking and setting timing, I read and reread and reread...those posts to be sure that I would know what I am looking at. To be sure, I checked and rechecked all four cams and there orientation to each other at top dead center compression as well as exhaust. Thanks to "Insite" for his great write up on the timing topic.

I did the leak down test on all six cylinders with the pistons in the compression as well as exhaust positions. I pinned the crank for cylinders 1&4. For the others, I held the crank with a wrench. I performed the test on all cylinders twice getting similar numbers both times.

I appreciate your asking me questions that I think I know the answers to. While I have tinkered with cars for most of my life, and even ran a national chain exhaust shop for a few years, my technical knowledge is very limited. I am an eager "learn as you go" Porsche enthusiast.

[NewArt]

As Byprodriver suggested, I don't think you have conducted your leakdown test according to Hoyle. Make sure you are at TDC on the compression stroke for each cylinder. Otherwise you may have a valve partially open resulting in false readings. Also, it is not accurate to do this test on a cold engine.
Ah, okay, I see that you have answered back while I was typing. Still the cold engine thing though. The rings won't make a good seal.

[Jamesp]

After rebuilding my engine I blew the valve timing and tried to start the engine. I cranked and cranked with my valve timing off. When I realized what I had done, I got that sinking feeling that I just ruined a fresh rebuild. So after taking the engine back out of the car, dumping one head and finding the valves had not hit the pistons I performed a leak down test on every cylinder. Getting the valves to seal properly required me to turn the engine a few degrees counter clockwise around top dead center. There was a huge difference over couple of degrees in sealing or not sealing the valves. Can anyone tell me what the downside of rolling the engine counterclockwise is? I've never understood what breaks if that is done.

[NewArt]

If I understand correctly, the chains and tensioners are unidirectional, like a bicycle. If you go counter clockwise the timing will not sync properly with the crank position and you risk jumping a tooth. I may be wrong.

[flaps10]

You wouldn't pet a cat backwards would you?

There is quite a bit of extra chain, especially the one from the crank to the IMS. The slack is always on the side of the tensioner, while the non slack side keeps perfect (cough) unison. Turning it backwards will pull the slack side tight, and allowing the "extra" chain to now destroy the relationship of the two gears.

If the tensioners were rigid it wouldnt matter much, but they are a combination of a not particularly ballsy spring and a piston trapping oil behind a check valve (I'm simplifying for purpose of the description) These tensioner are already a source of trouble (look up chain rattle on start up)

This pic shows the chain I'm referring to. The zip tie is near the point where the tensioner pushes



Even if turning it backwards wasn't enough to cause valves to touch pistons it would make it very hard to get consistent timing. You would have to spin it the correct direction several times in order to get it to settle in.

Once that zip tie was cut i put sharpie arrows on both ends of the engine to keep me from doing something dumber than usual.

[Jamesp]

Flaps10 Thanks for the picture and reply. I can not answer your question because I don't discuss cats on this forum. :chicken: Interesting point on reversing the tension in the timing chains. Do you think the force into the chain tensioner would increase somewhat if rotated backwards? I don't think that would actually break the tensioner if it did, and I think you are right, the valves would not hit the pistons.

[flaps10]

James I'm sort of winging here but the force generated by the tensioner just has to be enough to take up the slack. When they aren't pumped up with oil you can easily compress them against your bench with one hand.

With a static engine if you turn the crank backwards it would be the valve springs against the tensioner spring (plus quite a bit of chain drag). The cams would hold the IMS still until the slack side was pulled tight.

By comparison, my Triumph motorcycles are inline triples - sort of like half a "real" motor.
One cam chain, one tensioner which is a long but quite beefy spring. 9750 rpm as much as i like (and i do). Off topic but when you take a cane to the triumph triple it sounds similar to a nasty flat six. And stone axe reliable i might add.

[rah rah 986]

Is there anyone who can add to this thread that can shed some light on what might be causing the symptoms that I described in my original post? Your help is needed, and will be appreciated.

[JFP in PA]

Originally Posted by rah rah 986

Is there anyone who can add to this thread that can shed some light on what might be causing the symptoms that I described in my original post? Your help is needed, and will be appreciated.

Assuming the cam timing is where it is supposed to be, take a look at your crank position sensor, they can cause an intermittent non coding no start condition where the engine just spins over.

[flaps10]

Originally Posted by JFP in PA

Assuming the cam timing is where it is supposed to be, take a look at your crank position sensor, they can cause an intermittent non coding no start condition where the engine just spins over.

I was going to suggest the same thing. I would think an OBD2 reader might uncover that, but if you have the engine laying on the ground that is out.

I wonder if there is a way to bench test the sensor. Perhaps a resistance check with a multimeter while you whiz a piece of metal past the end. It's just a Hall sensor isn't it?

If the engine were in the car you could just swap it with a known good one, but again you'll have to reassemble the whole beast before that is an option.

Either way I'm tuned in for the fix. Best wishes to the OP.

edit: just re read JFP's post about it not pulling a code. That puts us back at bench testing or replacing with a known good sensor.

[JFP in PA]

Originally Posted by flaps10

I was going to suggest the same thing. I would think an OBD2 reader might uncover that, but if you have the engine laying on the ground that is out.

I wonder if there is a way to bench test the sensor. Perhaps a resistance check with a multimeter while you whiz a piece of metal past the end. It's just a Hall sensor isn't it?

If the engine were in the car you could just swap it with a known good one, but again you'll have to reassemble the whole beast before that is an option.

Either way I'm tuned in for the fix. Best wishes to the OP.

edit: just re read JFP's post about it not pulling a code. That puts us back at bench testing or replacing with a known good sensor.

The correct way to test a CPS is with an oscilloscope trace to show the pulse generation as the shutter teeth pass the sensor (which is a Hall Effect sensor).



The problem with diagnosing the CPS is that it can fail without throwing a code as the DME just thinks the engine is not turning and shuts off the fuel and ignition. Bench testing flakey units with a multimeter can be misleading; they can look fine, but still not deliver the pulse width required for the car to run properly. They are also heat sensitive, working when cold but failing when hot.

[Steve Tinker]

I'm a little confused here (as normal)...surely a failed CPS won't make the engine sound like it has no compression when cranking.
The OP (rah rah) says in his opening question that the car sounded as though it was "cranking with low or no compression" which says to me that the starter motor was turning over the engine @ 600 rpm instead of 350 rpm) - just like if there were no spark plugs screwed in the heads.

While leak down tests are all well & good + more accurate than a simple compression test, as a couple of people has mentioned you have to get it right otherwise you can have some misleading results as with his comment "And EVERY cylinder is down 75-90%".

At this stage with engine out (which I think was a premature move) removing the spark plugs, opening the throttle body and conducting a simple compression test - even with a cold engine - will give some indication of the condition of the valve seating and piston compression on each cylinder.
I mean, this could be just a couple of cylinders that have collapsed hydraulic valve lash adjusters causing the starter motor to turn over faster than usual.....and yes the CPS can be the rout cause of the original problem too!

[JFP in PA]

Originally Posted by Steve Tinker

I'm a little confused here (as normal)...surely a failed CPS won't make the engine sound like it has no compression when cranking.
The OP (rah rah) says in his opening question that the car sounded as though it was "cranking with low or no compression" which says to me that the starter motor was turning over the engine @ 600 rpm instead of 350 rpm) - just like if there were no spark plugs screwed in the heads.

While leak down tests are all well & good + more accurate than a simple compression test, as a couple of people has mentioned you have to get it right otherwise you can have some misleading results as with his comment "And EVERY cylinder is down 75-90%".

At this stage with engine out (which I think was a premature move) removing the spark plugs, opening the throttle body and conducting a simple compression test - even with a cold engine - will give some indication of the condition of the valve seating and piston compression on each cylinder.
I mean, this could be just a couple of cylinders that have collapsed hydraulic valve lash adjusters causing the starter motor to turn over faster than usual.....and yes the CPS can be the rout cause of the original problem too!

Depends upon how he defines "no compression" appearance while cranking. We have had several similar customer comments that were on cars with bad CPS and which aimlessly spun over very rapidly, but made no attempt to start.

If he really believes he has no compression, a quick compression check would either bear that out or prove it wrong. But based upon the number of times we are told the car seem not to have compression, only to find a bad CPS, I still think it is worth a look.

I am also somewhat circumspect about the leak down results; unless you have done one of these a few times, is very easy to miss TDC and get weird results, and because of the way these alloy engine's change, I would prefer to do a leak down on a warm engine.

[Jake Raby]

Sounds like you have sticky lifters... These can cause a total loss of compression like this.

[Gelbster]

"Sticky Lifters" - is his because over the long period of disuse they drained most of the oil out of many of the lifters?
If so, no amount of churning with the starter motor would re-fill the lifters . So the lifters must be replaced ? Quite a job and a significant expense. Well worth doing some more diagnostics first to verify simpler issues are not causing the problem.

[rah rah 986]

Thanks for all of the feedback. I am at a bit of a standstill, and not sure what to do next. My goal when pulling the engine was to rule out or fix any conditions that might lead to catastrophic engine failure. My greatest source of concern being the cranking of the engine that sounded as though there was little internal resistance. My first thought was that I had better check the timing before doing anything else. When attempting to remove the cam plugs, I knocked one into the the cavity behind the valve cover. Retrieving the plastic plug seemed near impossible without removing the cover. Being that we were well into winter, I decided this would be a good opportunity to pull the engine; check the timing; hopefully rule out any major internal problem; and then do some more in depth maintainance procedures such as RMS; check IMS; replace clutch; replace cam to cam chain pads; replace other tensioner ramps as possible; and maybe replace the lifters.

As noted earlier, timing is perfect. Leak down...awful. A few posts above suggest the leak down may be inaccurate for a few possible reasons. Jake suggests that sticky lifters can sometimes cause poor results. In order to rule out the lifters, I removed all of the lifters. My assumption is that the valves should all be completely closed. Also, without the lifters in place, I thought the piston being at exact TDC would be less critical in assuring the valves are all closed. (If I am way off in my thinking....just learning). With all of the lifters out, I have retested, and am still getting near total leakage. It is very audible through the exhaust.

The other suggestion above is that the results are skewed because I am either performing the test incorrectly, or that the results are skewed because the engine is cold. I am open to either possibility; because I can't argue with any certainty otherwise. I will say, I can remove the test hose from the spark plug hole, and control the readout on the gauge simply by applying pressure with my thumb to the end of the hose. This would tell me that indeed, the gauge is accurately reading that the cylinders are not holding any pressure.

Can this really be simply because the engine is cold? It would seem that even a cold engine would hold some pressure. it is probably worth repeating that the poor leak down numbers of 90%+ is at almost every cylinder.

I also should comment on what I have seen as I have been involved in the dissassembly of the engine. To my untrained eye, everything looks pristine. Everything under the valve cover is extremely clean. Nothing that I can see with having gone as far ar removing the cams and lifters looks out of order.

Since the high leak down numbers are at every cylinder, I need to ask; based upon what I have reported here, is there any problem that comes to mind that would affect every cylinder in this way?

At this point I don't know what to do next. If the best advise that I get is to ignore the test results, move forward, and reassemble the car; then, that is what I will do. I would be thrilled to find out that I have no significant internal engine problems. I only hope that my problem is as minor as a crankshaft position sensor.

If anyone can tell me with any certainty that these leak down results are insignificant, then I will ignore them. (What is difficult to ignore is that I can hear the leak through the exhaust....the leak is real)

Once again, thank you all for you help so far.

[JFP in PA]

Originally Posted by rah rah 986

Thanks for all of the feedback. I am at a bit of a standstill, and not sure what to do next. My goal when pulling the engine was to rule out or fix any conditions that might lead to catastrophic engine failure. My greatest source of concern being the cranking of the engine that sounded as though there was little internal resistance. My first thought was that I had better check the timing before doing anything else. When attempting to remove the cam plugs, I knocked one into the the cavity behind the valve cover. Retrieving the plastic plug seemed near impossible without removing the cover. Being that we were well into winter, I decided this would be a good opportunity to pull the engine; check the timing; hopefully rule out any major internal problem; and then do some more in depth maintainance procedures such as RMS; check IMS; replace clutch; replace cam to cam chain pads; replace other tensioner ramps as possible; and maybe replace the lifters.

As noted earlier, timing is perfect. Leak down...awful. A few posts above suggest the leak down may be inaccurate for a few possible reasons. Jake suggests that sticky lifters can sometimes cause poor results. In order to rule out the lifters, I removed all of the lifters. My assumption is that the valves should all be completely closed. Also, without the lifters in place, I thought the piston being at exact TDC would be less critical in assuring the valves are all closed. (If I am way off in my thinking....just learning). With all of the lifters out, I have retested, and am still getting near total leakage. It is very audible through the exhaust.

The other suggestion above is that the results are skewed because I am either performing the test incorrectly, or that the results are skewed because the engine is cold. I am open to either possibility; because I can't argue with any certainty otherwise. I will say, I can remove the test hose from the spark plug hole, and control the readout on the gauge simply by applying pressure with my thumb to the end of the hose. This would tell me that indeed, the gauge is accurately reading that the cylinders are not holding any pressure.

Can this really be simply because the engine is cold? It would seem that even a cold engine would hold some pressure. it is probably worth repeating that the poor leak down numbers of 90%+ is at almost every cylinder.

I also should comment on what I have seen as I have been involved in the dissassembly of the engine. To my untrained eye, everything looks pristine. Everything under the valve cover is extremely clean. Nothing that I can see with having gone as far ar removing the cams and lifters looks out of order.

Since the high leak down numbers are at every cylinder, I need to ask; based upon what I have reported here, is there any problem that comes to mind that would affect every cylinder in this way?

At this point I don't know what to do next. If the best advise that I get is to ignore the test results, move forward, and reassemble the car; then, that is what I will do. I would be thrilled to find out that I have no significant internal engine problems. I only hope that my problem is as minor as a crankshaft position sensor.

If anyone can tell me with any certainty that these leak down results are insignificant, then I will ignore them. (What is difficult to ignore is that I can hear the leak through the exhaust....the leak is real)

Once again, thank you all for you help so far.

If you have the cam followers (lifters) out of the engine and a leak down test still blows air out the exhaust, you have an internal failure of some sort (valve, cracked head, etc.), and you next step should not be to reassemble everything, but rather to pull the heads, starting with the one where the air cam out the exhaust, looking for damage.

[Jamesp]

A cold engine can give poor leak down results, but you are experiencing near complete failure. I'd agree with JFP - next step is dump the head.