PPI Compression & Leakdown Question

Had a PPI incl. compression and leakdown done on a car I am looking at. It's a high mileage N/A 996 - 3.6L block with 3.4L heads. The engine is a recently (< 5k miles) rebuilt engine from a competent shop with a fair bit of experience building race engines. Given the recent rebuild, I expected good results. The results were reported as : Cyl Compr Leak 1 185 1.5% 2 185 1.5% 3 180 2.0% 4 175 3.5% 5 185 1.5% 6 180 2.0% Two questions. First of all the cylinder to cylinder variation looks reasonable - and since the leakdown seems to correlate to the compression, probably legitimate variation rather than just "noise". Correct ? Anything to worry about on that ? The compression pressures seem to be low vs. what I have read in the forums for typical test numbers (240-250) for a N/A 996. But, frankly, having difficult finding any specs on 996's other than CR and you can't really compute one from the other from what I understand. The mechanic who did the test (again, a reputable shop) says they normally just look for cylinder to cylinder differences and don't put any stock in the "actual value" ... In his comments, he stated the engine started and ran excellent. Only marginal component was a spark plug tube leak on cyl #1. Any comments on the test results ? I'd be paying a bit of a premium due to the rebuilt engine so it better be good ...

Those numbers look very good. You have a heathy engine there. The important test is the leakdown and those numbers are nice and show a well sealed cylinders. Your mechanic is right.

Thanks Imre - as an aside, am really enjoying your "brought home ..." thread. Wish I had those skills and time myself.

Quotes source post: 10125363

Well, there may be something I'm not aware (heck there's tons of things I'm not aware of) but I would consider a 3.6 engine with 3.4l engine heads a Frankenstein engine and unless further research turned up this is the hot hot hot setup shun the car because of this. The 3.4l heads probably have smaller chambers which may have the compression ratio up (or down) from stock. The 3.4l heads' smaller combustion chambers can have edges that stick out from the block deck and thus get hot and lead to pre-ignition and detonation. The smaller combustion chambers may not flow as good as the 3.6l head chambers. Both the cylinder to cylinder numbers and the overall numbers want to be within spec. For cylinder to cylinder IIRC (for engines in general) all like to be within 10% of each other. As for the actual pressures, even if all cylinders read the same if the actual pressures are down (or up) from what is normal this is cause for concern. Ask the shop what the over all pressure should be. Unless my math/reasoning is flawed, the overall numbers seem high. The 3.4l engine compression ratio is 11.3:1. Assuming 1 atmosphere of pressure (14.6960psi) this works out to 166psi. A 185 psi number then to me appears high. If it is high the engine may require higher octane gasoline than the 93 octane to avoid the DME from dialing back the timing. There is the risk that the DME may not be able to dial back the timing enough. There is the concern too about how this engine will run on say hot days. Furthermore be sure you take the car in a nice long drive. Cover 30 miles and under a variety of driving conditions to give the best compression tester -- the DME -- and chance to identify/flag any misfires which is much more significant than a possible passing grade from a compression test. A complication is it is cold where you are and the engine may run just fine in cold weather but when warm may not run all that well. Unless you can get confirmation from some reputable engine builder that doesn't have a fiduciary interest in the car being sold that gives his blessing to the use of 3.4l heads on a 3.6l block I'd give the car a miss regardless of how good the compression test results are.

The 3.4 VarioCam heads/system is quite different than the 3.6 VarioCamPlus heads/system. and the DME for each is different. What year is this car? What DME doe it have? There are just too many variables here to really address the issue of the use of teh 3.4 heads on a 3.6 engine. And why would someone do this? Perhaps what this is, is a 3.4 engine that has been re-sleeved/bored to 3.6.

The car is a '99 C2 (3.4) - the engine was money shifted and died. The DME is for the OE 3.4L with variocam (not variocam+) so the current owner didn't just put a 3.6L engine in. I'm actually not 100% sure on the motivation for using a 3.6L block and not going for just a 3.4L rebuild - presumably he wanted some torque improvement at low end. The replacement engine was built based on 3.6 block with 3.4 heads by a shop who has done it previously with good results in a race environment. The build was essentially inspired by a similar build by somone else on this forum. My plan for this car is pure track use (DE).

Quotes source post: 10125513

If the engine builder has done this 3.6l block/3.4l head combination before with good results in a race environment and you are confident in the results and are confident the engine will be everything you hoped for in a race/track environment then my concerns are moot. Enjoy the new car and its 3.6l/3.4l engine.

Thanks Macster ... I certainly hope it turns out to be reliable as well. I did talk to the shop who built it and exchanged emails with guy whose build inspired the similar build, etc. but I am not a mechanic - barely know enough to get in trouble. Based on those discussions, however, I'd made the leap that the concept was OK ... but I was a little confused by these test numbers. My number crunching was like yours but then of course the charge heats up as it is compressed so the pressure ratio can be higher - a typical pressure ratio including the effect of heating could be more like CR ^ 1.4 (eg. PR of 29.8:1 for a CR of 11.3:1) from what I found on a few Wiki pages. That however doesn't account for valve timing, etc. which could lower it when cranking via the starter motor due to timing overlap and so - on. There was a tech note on Pelican which specifically mentioned that not really any point in looking at the raw numbers - just the variation. And then I did find a few posts where people said their compression readings were 240-250 ... which just seemed really high to me ... Again - thanks - all input is valuble.