low compression number

[nsxhk]

Planning on going FI, thought maybe a compression test is in order before jumping in, here are the results of my '91 3.0L in psi:

front:
130, 130, 130

rear:
125, 120, 130

From the service manual:
Normal: ~190psi
Min: ~145psi

I'm way under normal with each cylinder. Trying to understand why?!

My internal mods:
- sos camshaft
- sos valve springs + ti retainers
- sos LMA
- all new oem seals
- sos cam seals
- oem '97 mls head gasket

Your thoughts?

Rgds,

Henry.

 

[MikeW]

I guess it is because of your camshafts.
Ramping with longer overlap reduces static compression.

 

[greenberet]

I recently had the compression checked on my 138,000 mile engine which has Comptech camshafts and heads that were were milled to increase the compression ratio to about 10.6:1.

To get as accurate results as they could, the shop that did the measurement bought a brand new MotoMeter compression tester and my engine was the first engine they used it on. The results were 4x 196 psi and 2x 199 psi. The shop said they turned the engine over a couple more times when measuring the two cylinders that showed a slightly higher compression and that that’s why not all cylinders measured completely even.

Since my shop bought a new compression tester to try to ensure accurate results, maybe the compression tester your shop used is older and out of adjustment. Getting your engine retested with a different gauge might shed light on that. And ask them to try to turn the engine over equally long when measuring each cylinder.

Regarding the camshafts, MikeW has a good point. My Comptech camshafts have modified VTEC lobes but the low-rpm lobes are stock. So when the compression is measured, the valves are being lifted by the stock low-rpm lobes and those have no overlap at all (intake and exhaust valves are not open at the same time). SoS state that the low-rpm lobes of their camshafts are modified, but I don’t know whether they have any overlap. If they do, that would bleed off some of the pressure and could also be the reason for your results. Giving SoS a call should answer that one.

Edit: it may also make a difference whether you have the SoS naturally-aspirated or forced-induction camshafts in your engine right now. In a naturally aspirated engine, more overlap will give you more top-end power but in a forced induction engine, less overlap will give you more power. That's why SoS offer different camshafts for each application. If you have the SoS naturally-aspirated camshafts, that may not be ideal for going FI.

 

[rolferde]

if you really want a good idea of how things are wearing a leak down test would give you a much better idea. you do not have to worry about cam overlap changing your readings.

 

[Big_D]

Compression tests must be made at Wide Open Throttle. I suspect yours wasn't. The main thing is that they are all within a few %.

 

[ScienceofSpeed]

hi Henry --

How is the engine idling and running?

Normal compression is around 200-210 on a good condition engine. Even with these cams, the compression shouldn't change your compression by 5-10 psi. At that low of a compression, your engine should be having a problem even idling. I suspect as others have that they left the throttle body closed or something is wrong with the gauge.

Make sure when the test is done, the throttle body is open and the fuel pump fuse is removed (to prevent the cylinders from being washed).

If the numbers are still low, you can check what the vacuum is with the car idling. Once warm, it should be around 20-22 inHg. Even this is still very low, I'd suggest doing a leak down test to determine where the leak is coming from (intake valves - heard through throttle body, exh valves - heard through exhaust, pistons - heard through crankcase / oil cap). If indeed there is high leakage, the major suspect would be cam timing from the belt installation.

-- Chris

 

[nsxhk]

I wasn't there when they did the compression test. I don't know if they kept the throttle wide open or not but I don't think they removed the fuel pump fuse. I'll try again. It seems strange that they're all that low.

Thx guys....

Henry.

 

[nsxhk]

Btw, the car drives fine, idling fine, everything working perfectly.

Just talked to the mechanic, he did do the compression test with wot but did not remove the fuel pump fuse.

Henry.

 

[Big_D]

You wouldn't happen to be in Denver would you? High altitude can make a difference. I still think your tech didn't have the throttle wide open.

 

[EdinKali]

Those look exactly like closed throttle compression numbers. If the mechanic was holding the throttle open with his foot, the throttle valve may still have been closed since it is drive by wire from 95 on if I recall correctly. The throttle may stay closed until the engine starts. Maybe someone like Larry can confirm or deny.

 

[nsxhk]

my car is a '91 3.0L. I think I'm going to go to another shop and redo the test.

Henry.

 

[nsxhk]

Re-did the compression test from another shop, the results are much better....

235, 240, 230, 240,240,240psi

Strangely enough, now they're a bit too high?!

Maybe the mechanic cranked it too long? The starter motor did overheat and seized for a little while.

Well, guess I'm ready for FI

Henry.

 

[pgilliam1]

Sounds like you're also ready for a new starter motor. Good Luck with it.
Happy Motoring!

 

[nsxhk]

Already on order

 

[Big_D]

Re-did the compression test from another shop, the results are much better....

235, 240, 230, 240,240,240psi

Strangely enough, now they're a bit too high?!

Maybe the mechanic cranked it too long? The starter motor did overheat and seized for a little while.

Well, guess I'm ready for FI

Henry.

Dang, that is a lot of cranking.

Perhaps the gauge may not be all that accurate, who knows when it was last calibrated.

Again, the main thing is that all the cylinder readings are close.

 

[nsxhk]

The compression gauge is brand spanking new. The mechanic cranked for about 10 seconds each cylinder.

 

[greenberet]

Normal compression is around 200-210 on a good condition engine.

Re-did the compression test from another shop, the results are much better....

235, 240, 230, 240,240,240psi

Strangely enough, now they're a bit too high?!

I stumbled across the section of the Service Manual that states what compression a brand new NSX engine should show: at least 142 psi (10 kg/cm2) on each cylinder with a maximum variation of 28 psi (2.0 kg/cm2) between cylinders. In addition, until 2000, the Service Manual stated what the nominal compression of the engine is - the compression readings the engine would theoretically produce if there were absolutely no leakage past the piston rings, intake, or exhaust valves: 199 psi (14.0 kg/cm2).

If an NSX engine shows a compression above 199 psi, it kind of makes you wonder how accurate these compression testers are even when new. Oil collecting on top of the pistons or carbon deposits could potentially decrease the effective combustion chamber volume and thereby increase the effective compression ratio. However, 3 cc’s would have to collect on top of a piston to bump the compression to 210 psi and 10 cc’s would be required to bump the compression to 240 psi – if there were absolutely no leakage past the piston rings, intake, or exhaust valves. Since there is some leakage, more than 10 cc’s of deposits would be required on each piston to generate those compression figures. That’s a lot of deposits. Engine builders should know whether that’s realistic or not, but from my uninformed standpoint, it rather calls the accuracy of the compression testers into question.

In case anyone’s interested how the nominal compression is calculated: a stock NA1 has a stroke of 78 mm and a bore of 90 mm so the piston displaces 496 cc’s during its stroke. Since the engine has a compression ratio of 10.2:1, the combustion chamber above the piston has a volume of 54 cc’s. However, to make things more complicated, the intake valves are still open when the piston has already started its upward stroke. According to the Service Manual, the intake valves close 25° after bottom dead center, so the intake valves are not closed for the entire 78 mm of the piston’s upward stroke. They're only closed for 75.2 mm so the air compressed by the piston has a volume of 479 cc’s. 479 cc’s of net displacement and a 54 cc combustion chamber give an effective compression ratio of 9.9:1. At sea level, the ambient air pressure is 14.7 psi. However, as the air is compressed in the engine, it gets hotter. As it gets hotter, the pressure increases by a factor of about 1.37. With that data, we can now calculate the engine’s nominal compression ratio. The effective compression ratio (9.9) times the ambient air pressure (14.7) times the pressure increase due to the air getting hotter (1.37) = 199 psi (don’t round off any decimal points anywhere any you’ll get to 198.86 psi). That’s the compression the engine would theoretically produce at sea level if there were absolutely no leakage past the piston rings, etc. Since the tests are often not carried out at sea level and since there is leakage, I agree with you nsxhk - compression readings of 240 psi or even 210 seem really high for a stock NSX engine.