Good Leak Down but Bad Compression after rebuild
 

After having the top end of my 3.2L Carrera rebuilt with used 3.3L Mahle cylinders and pistons, the leak down numbers are coming back OK but the compression numbers are really low (average 80 PSI per cylinder). Cylinders were re-ringed. Heads received a valve job and were twin plugged. Moved to 964/993 twin plug dizzy and added an Andial splitter. What do you suspect the problem(s) could be?

The compression test was performed multiple times by a pro and yielded the same results. Thanks for your help!

Has the motor been run in yet?

Good question. I've got about 5K miles since the rebuild.

What was the advertised CR of the 3.3 P&C set?

8.0:1. Valve reliefs were already cut into the pistons when I received them.

Try a different compression gauge?

Whats your cam timing/overlap?

It was intended to be stock. Would the car run OK if the cam timing were off? The car seems to run smoothly. Could the wrong cam timing cause the compression to be around 80 PSI in each cylinder?

The more overlap the less static compression. It can make a huge difference but that's not how you should be setting it.

Are there any other symptoms of bad cam timing? I'd imagine it that fuel economy would go to the crapper?

Overlap is built into the grind of the cam; it's the separation between the intake and exhaust lobe of any given cylinder. The timing of the cam can't be out by much or you would get some piston to valve contact. The more overlap you have the shorter the time is that both intake and exhaust valves are closed on the compression stroke which is where you build cylinder pressure. With 8:1 pistons and relieved valve pockets you could easily be at 7.5:1 or less. What cams are you using?

How is the shop determining 80psi. Do they do 5 cranks or do they just turn it until it levels off?

What should it be with 8:1 and 3.2 cams? If the CR is really 8:1, then I'm calculating it should be right about 100 psi. (14.69 x 8) - 14.69 = 102.8 psig. This is purely a static calculation. If you have significant overlap, then that may explain why you're at 80. Also, how sure are you about stated CR, i.e. did you measure?

Stock 3.2 Carrera cams

I'm not sure. The test was performed by a very reputable mechanic with 30+ years experience. He said he ran the test multiple times because the results were so negative.


I thought compression ratio had no impact on the amount of pressure that a cylinder could hold?

No measurement was performed.

Engine needs to be warm to seal, and ring end gap will cause less pressure, as will cam timing. The real thing you look for in a compression test is that all cylinders have the same compression.

By hearing your responses, it sounds like bad cam timing can only explain *some* of the low compression. Is that right? Again, we're seeing an average of 80 PSI per cylinder, and the leak down numbers look OK.

In layman's terms, leak down tests determine how much air/fuel can escape from the combustion chamber. And compression tests determine how much pressure can be held in each chamber. Is that right?

So, it sounds like the results must be explained by some combination of these 3 points:

1. False negative on the Compression Test
2. False positive on the Leak Down Test
3. Bad Cam Timing

Am I missing anything else?

What is the difference in head/combustion chamber volume between the 930 and 3.2 and could this be part of the issue?

I don't think cam timing has much of anything to do with your low compression numbers. I think the piston / head combination you have is the culprit.

I hadnt thought about that. In fact, I thought it was commonly understood that 930 cylinders/pistons were slip fit for otherwise stock 3.2L Carrera engines. Perhaps other modifications are required to fit 930 p&c's?

Valve timing is not part of the static compression calculation.
Valve timing affects dynamic compression. The higher the overlap the lower the dynamic compression.

3.3 production 930 pistons have an advertized compression of 7:1.
If they have valve pockets that number goes down. By simply changing the deck height to .060" the compression ratio goes down. .020 deck change will reduce the compression by .24:1.

Henry, do you think valve timing alone could cause the compression to drop to 80 psi? Any known issue with installing 3.3 l pistons and cylinders on a stock 3.2?

Polezei

Henry is being very diplomatic here. I will present a rather pessimistic analysis.

The compression ratio has an effect on measured compression. Isn't that obvious? Compression is how much you squeeze the air in the cylinder. With low compressiion pistons you don't squeeze the air as much, so your compression is lower.

The fact that a 3.3 cylinder is a slip fit into the case has nothing to do with whether it is a good idea to use 7.0:1 3.3 turbo pistons in those cylinders without also turbo or super charging. Your 3.2 had a CR of 9.5:1 to start with. Even if 3.3 turbo pistons will produce 8.0:1 CR using a stock 3.2 head, you are going backward, aren't you? I don't think the extra 100 CCs will make up for a substantially lower CR.

And since you did not measure the compression ratio as you were building the motor (it is actually pretty easy if you can get ahold of a burette), you are stuck with suppositions as to what the CR is. But everything you say points to it being quite low for your motor, which is precisely in line with the compression being low. You built a nice, tight motor, though.

And that ain't all. You twin plugged the heads. That alone will lower the CR, though not by much. More to the point, I don't think there is much if any advantage to twin plugging a normally aspirated low compression motor. Twin plugging allows early small motors with high domes and high CRs to live. It allows 964/993 motors to run CRs which would have killed earlier motors. But I don't see how it will contribute a thing to a low CR motor. Maybe it would allow running pump regular instead of intermediate grade fuel?

Your options as I see them are:

1) live with a low compression engine with a slightly larger displacement than what you had beforfe. You have effectively detuned this motor. Am I right that you had these tests done at least partly because the motor wasn't as peppy as it was before you had the work done? You noticed this right away, but were hoping it would improve when the rings seated?

2) add a super or turbo charger, since you have an appropriately low CR - and the twin plugging. There might be problems, though. Didn't the turbos use stouter rods, to start with? Rod bolts are the weak point of stock 3.2s. Home brewing a blown motor is not for the faint of heart or pocketbook.

3) Take it partly apart, and replace the pistons with aftermarket pistons cut for a suitable CR. That may be your best bet if your old Ps and Cs are worn out.

4) Put your stock stuff, with new rings, back in. Sell the 3.3 stuff to someone who is converting to a turbo. Least cost alternative.

I hardly have a Mahle catalog compendium of every piston they ever made in my head or within arm's reach, but the only 3.3 piston one hears about is the turbo one made for Porsche. I can't imagine that needing valve reliefs. Don't they have a sort of reverse dome to achieve that 7.0:1? Maybe they made one for the aftermarket for normally aspirated motors as well? But then why would anyone need to cut valve reliefs? That would all be built-into the casting/forging. But no matter.

Valve timing has nothing to do with your low compression figures. The gauges and tests aren't off. Believe me, I know all about denial and wishful thinking when a motor I have built doesn't live up to what I had hoped.

Polezei

Henry is being very diplomatic here. I will present a rather pessimistic analysis.

The compression ratio has an effect on measured compression. Isn't that obvious? Compression is how much you squeeze the air in the cylinder. With low compressiion pistons you don't squeeze the air as much, so your compression is lower.

The fact that a 3.3 cylinder is a slip fit into the case has nothing to do with whether it is a good idea to use 7.0:1 3.3 turbo pistons in those cylinders without also turbo or super charging. Your 3.2 had a CR of 9.5:1 to start with. Even if 3.3 turbo pistons will produce 8.0:1 CR using a stock 3.2 head, you are going backward, aren't you? I don't think the extra 100 CCs will make up for a substantially lower CR.

Using Alfonso's formula (which is for sea level, but Norristown isn't very high, is it?), I get about 88 psi for 7.0:1. That's getting within range.

And since you did not measure the compression ratio as you were building the motor (it is actually pretty easy if you can get ahold of a burette), you are stuck with suppositions as to what the CR is. But everything you say points to it being quite low for your motor, which is precisely in line with the compression being low. A nice, tight motor, though.

And that ain't all. You twin plugged the heads. That alone will lower the CR, though not by much. More to the point, I don't think there is much if any advantage to twin plugging a normally aspirated low compression motor. Twin plugging allows early small motors with high domes and high CRs to live. It allows 964/993 motors to run CRs which would have killed earlier motors. But I don't see how it will contribute a thing to a low CR motor. Maybe it would allow running pump regular instead of intermediate grade fuel?

Your options as I see them are:

1) live with a low compression engine with a slightly larger displacement than what you had beforfe. You have effectively detuned this motor. Am I right that you had these tests done at least partly because the motor wasn't as peppy as it was before you had the work done? You noticed this right away, but were hoping it would improve when the rings seated?

2) add a super or turbo charger, since you have an appropriately low CR - and the twin plugging. There might be problems, though. Didn't the turbos use stouter rods, to start with? Rod bolts are the weak point of stock 3.2s. Home brewing a blown motor is not for the faint of heart or pocketbook.

3) Take it partly apart, and replace the pistons with aftermarket pistons cut for a suitable CR. That may be your best bet if your old Ps and Cs are worn out.

4) Put your stock stuff, with new rings, back in. Sell the 3.3 stuff to someone who is converting to a turbo. Least cost alternative.

I hardly have a Mahle catalog compendium of every piston they ever made in my head or within arm's reach, but the only 3.3 piston one hears about is the turbo one made for Porsche. I can't imagine that needing valve reliefs. Don't they have a sort of reverse dome to achieve that 7.0:1? Maybe they made one for the aftermarket for normally aspirated motors as well? But then why would anyone need to cut valve reliefs? That would all be built-into the casting/forging. But no matter.

Valve timing has nothing to do with your low compression figures. The gauges and tests aren't off. Believe me, I know all about denial and wishful thinking when a motor I have built doesn't live up to what I had hoped.

Rookie Mistake
 

I was chatting with a fellow Pelican last night regarding my engine woes, and he asked me this question:

"When you installed the 97mm turbo pistons and cylinders, you measured the deck height, right?"

I'm embarrassed to say that not only did I not measure the deck height, but I didn't even understand at the time that that is standard operating procedure!

So, it's likely that there's nothing wrong with any of my parts. But I must tear down the engine to fix it. And while I'm in there, I'm going to swap P&Cs with the intention of building a NA engine.

Thanks guys!

Polizei,

FWIW I ran a quick simulation on a 3.2 with 95mm cylinders and 9.5 compression 7000 RPM rev limit, came out 220 HP about right, changed the bore to 97 mm with 7.0 compression 201 HP.

In this case displacement was trumped by compression.

I had CP build pistons to match my otherwise good 97mm cylinders, the verdict on how it all works out is not yet in.

Quick question- I'm familiar with measuring deck height on V8 engines; I would describe it as how close to the top of the block the piston rises. Zero deck height would be that the piston is flush to the block at TDC.

And of course, regardless of deck height, valve reliefs and the volume of the combustion chamber would play a huge role in the compression ratio.

So while this is important to check, what is the spec for a Porsche 3.3, and how would you change it if it didn't check out?

QSX - for measuring compression ratios, deck height is used as you say if you are using the usual multiple measurements method. I found that a trial assembly and measureing how much fluid the chamber would hold at TDC removed the need to do that measuremnt.

I think that for 911 engine building the figure most checked is piston to head clearance. Too little and things can hit. Too much, and you don't have a suitable quench zone around the circumference. I like to use hollow core solder in preference to clay to measure this. But at least one of the highly experienced engine builders who contributes here doesn't trust this and always uses careful measurements. I don't quite trust measuring small distances down to a sloped surface, but if I were a machinist who does this for a living no doubt I would.

I'd have to hunt up what recommendations are for this, as I tend to forget details like that once I have applied them. As in my profession (law), it is usually enough to know something must be checked, and to go check it rather then relying on memory for precise details unless you deal with it all the time. Do a search. How you get your clearance and the CR you want can be done various ways.

A super trick motor would have the exact same CR for each cylinder. But, like having the exact same air flow in and out, that last ounce of extra power and balance is not usually done because you have to futz with the piston domes or the combustion chamber shapes, and it tends to lead to a least common denominator because removing material is more practical than adding it.