Questions about my compression test results

[QPDO]

After my car threw a code for bank 1 running rich, I started to investigate. Tossed in the OBDII reader and saw that in closed loop, my rear bank fuel trim was waaaaay out at -25+/-5. I brought it to the local dealership and they let me know that my rear bank, pre-cat, O2 sensor was reading out of range and needed to be replaced, which I did. Hooked up the OBDII reader again to see that it made zero difference. Thinking back, the O2 sensor was reading out of range in order to compensate for the rich condition. So realistically, I don't think the O2 sensor was ever at fault.

Now that I've determined O2 sensors are not the issue, I delved in deeper. Thinking worst case scenario (bent valve) I preformed a compression test. The results I believe confirm there are no bent valves as the individual banks are within +/- 20% across the cylinders, (155-145-147-215-205-200) but lead me to believe that my rear bank timing might be out (the lower set of numbers). I am not as mechanically inclined as some of you folks... so any insight before tossing the car back at the dealership for another dartboard diagnostic and $300, would be greatly appreciated.

The car is a 1998 NA2, with comptech headers, AP-X resonated test pipes, and AP-X stainless exhaust.

Thanks,

QPDO

 

[BATMANs]

So when I first brought my NSX to have the SOS SC installed, the shop did a leak down test and said that my 100k~ miles engine was uber healthy. I forgot what the actual compression numbers were, but he said that I had less than a 2%~ leakage and it was even across the board.

When my engine kaboomed, leakage ranged from 6%-50%~.

So based on that I'm going to conjecture that with your numbers being all over the place your piston rings are shot.

Ur not forced induction are u?

If not, were you low on oil and take a fast turn (momentary oil startvation)?

 

[QPDO]

Thanks for the speedy response. I am naturally aspirated, and I'm always on top of my oil levels (check every second fill-up). I guess the next step is to do a leak down you figure?

 

[BATMANs]

ur compression test should be ur leak down.

The numbers u posted looks really bad.

If ur "lucky" it's just a head gasket.

Did u over rev the engine like down shift to 1st at freeway speeds or something close to that?

 

[QPDO]

ur compression test should be ur leak down.

The numbers u posted looks really bad.

If ur "lucky" it's just a head gasket.

Did u over rev the engine like down shift to 1st at freeway speeds or something close to that?

I did a typical dry compression test this past weekend with those results, which is easier and quicker than a leak down so that's why I went that route. If the head gasket was shot, would it not show a complete loss of pressure? Or a variance in pressure between cylinders? The engine was over rev'd on a downshift... but not nearly to the extent of 1st on a freeway, though the over rev was the main reason I performed the compression test. Being that the compression test showed less than 7% loss between the rear bank cylinders, my minimal mechanical knowledge leads me to believe my valves are fine.

A leak down test would confirm if my piston rings are in good order, but would piston rings being shot cause a run rich situation?

 

[BATMANs]

head gasket leaks can show all sorts of symptoms. The ones u described are typical of what I've seen in other cars. head gaskets are one of the weak links of our engines.

Tell me more about the over revving.

If u over revved u could have down any number of things:

1. float a valve
2. valve and piston kissed
3. lift the head

i don't think u can do a leak down and determine the exact cause of leakage between piston and valves, just what cylinder is leaking, which u seem to have that info already.

 

[QPDO]

The over rev happened from a third gear to second gear shift while decelerating, at which point the car likely saw 10,000rpms.

Now please correct me if I'm wrong, as I am learning more and more as I delve into the diagnostics, but to answer your points:

1. wouldn't this have shown on a compression test?
2. if the valve kissed the piston, would the valve not have bent and again would this not show on a compression test?
3. this could very well be true

The leakdown test would give me a percentage leak per cylinder, and if I:

A: open the radiator cap and see bubbling, I'd know you have a leak into the coolant system
B: open the oil cap/remove the dipstick, and I'd hear/bubbles you know you have ring failure
C: remove air box and open throttle body to hear for air passing the intake valve (which should be if the compression test shows similar numbers across the board)
D: listen/feel at the end of my exhaust for air passing my exhaust valves (which *should* have already been confirmed via the compression test)

 

[warrenw]

Rear bank timing? one tooth away on exhaust? 1 tooth will not allow valves to hit pistons. I don't know what the compression would exactly show as you are experiencing, but the entire rear bank looks to be low. Missing by one tooth on the exhaust is an easy mistake to happen to the most experienced mechanics. Other than removing the front covers, I don't know if there is an easier way to check.
Did your engine feel as strong after the timing belt change? I think you would have noticed that is a 3 cylinder if timing is off. We had this happen to a former local car at WestSide a couple of years ago.

 

[QPDO]

Hello Warren! Ya, I'm not exactly sure what the issue could be... which is why I kind of thought the low numbers in the rear bank could be a timing thing. But then again, due to my limited mechanical knowledge, that may not be a cause whatsoever. My engine felt, sounded, acted just as strong after I replaced the belt last summer... so I'm left scratching my head, and asking more knowledgeable people on Prime.

 

[BATMANs]

Years ago my friend accidentally down shifted at freeway speeds where he hit 10k RPMs on his E36 M3.

The piston and valves kissed on some of the cylinders. His leak/compression test was similar to urs. All over the place.

Based on his experience and what happened to ur NSX I'm guessing u need to replace ur piston and valves.

 

[goldNSX]

The best scenario: your engine simply skipped a tooth on the TB. ->check
If the timing is still ok your bad scenario is what BATMAN said.

I'm not sure but I hope for you it's a good sign that only the rear cylinders are low on compression.

 

[BATMANs]

Warren if u have to rebuild ur engine I have experience in this area. U probably seen my DIY FI engine build thread.

 

[d_prada1]

After my car threw a code for bank 1 running rich, I started to investigate. Tossed in the OBDII reader and saw that in closed loop, my rear bank fuel trim was waaaaay out at -25+/-5. I brought it to the local dealership and they let me know that my rear bank, pre-cat, O2 sensor was reading out of range and needed to be replaced, which I did. Hooked up the OBDII reader again to see that it made zero difference. Thinking back, the O2 sensor was reading out of range in order to compensate for the rich condition. So realistically, I don't think the O2 sensor was ever at fault.

Now that I've determined O2 sensors are not the issue, I delved in deeper. Thinking worst case scenario (bent valve) I preformed a compression test. The results I believe confirm there are no bent valves as the individual banks are within +/- 20% across the cylinders, (155-145-147-215-205-200) but lead me to believe that my rear bank timing might be out (the lower set of numbers). I am not as mechanically inclined as some of you folks... so any insight before tossing the car back at the dealership for another dartboard diagnostic and $300, would be greatly appreciated.

The car is a 1998 NA2, with comptech headers, AP-X resonated test pipes, and AP-X stainless exhaust.

Thanks,

QPDO

Hey! Sorry to hear about your troubles but let me chime in with some of my experiences. Your compression numbers do not look like you have a bent valve, or a messed up head gasket. Not being mechanically inclined you deduced a very logical explaination. All your numbers in the rear bank are within a small percentage so it would have to be something affecting all cyclinders evenly. My first thought was also timing.

This actually happened to me on a J32A3. I guess the tensioner was at fault but after a 4th to 3rd gear drop over rev (accidental while spirited driving) my engine felt sloppy. I was gonna start to pull the engine out after my compression test was done and found low number all across the front bank. In my process of pulling the engine I thought "let me check the timing". Turns out the front bank was 2 teeth off!! I figured the over revved caused the tensioner to slack and jump a couple of teeth. I put it back together and BAM! Easy money everything was good to go.

 

[BATMANs]

now that would be a better ending.....

 

[d_prada1]

Years ago my friend accidentally down shifted at freeway speeds where he hit 10k RPMs on his E36 M3.

The piston and valves kissed on some of the cylinders. His leak/compression test was similar to urs. All over the place.

Based on his experience and what happened to ur NSX I'm guessing u need to replace ur piston and valves.

This could also be the case. But in most of my over rev experinces i usually only tap on one cyclinder maybe two max. I used to build a lot of K20s and K24s. revving these things to 9.5 was kinda sensitive to over rev problems missing a gear.

I would check timing first. If not then yes, you might have tapped the valves. Worst case scenario you can pull the head off and throw new valves and seats in. Most likely the exhaust valves will be the only bent ones. Being that we have forged pistons I doubt you did any damaged to them if the car is still running. But you can see for yourself after the head is pulled.

 

[BATMANs]

.......Being that we have forged pistons I doubt you did any damaged to them if the car is still running.......

When did the NSX have forged pistons?

They certainly don't look like forged pistons when I pull the engine apart from detonation.

They look and feel like cast pistons.

 

[d_prada1]

When did the NSX have forged pistons?

They certainly don't look like forged pistons when I pull the engine apart from detonation.

They look and feel like cast pistons.

I have never taken apart an nsx engine. I just read they had titanium rods and forged pistons. I can very well be wrong!
I just became an owner last week so in terms of the nsx im still new. Sorry for the NEWB post =) MAYBE the post i read was what HE had :apologetic:

 

[BATMANs]

Forged Ti con rods - yes.

My build:
http://www.nsxprime.com/forum/showthread.php/154927-Semi-DIY-Mild-Engine-Build-for-FI/page4

 

[d_prada1]

Forged Ti con rods - yes.

My build:
http://www.nsxprime.com/forum/showthread.php/154927-Semi-DIY-Mild-Engine-Build-for-FI/page4

Sweet Build!

Do you know if there is any OEM stroker build. I tried searching but couldnt find anything. I had an S2000 where i used a k24 crank from a tsx, h22 pistons and h22 eagle rods with the rods machines to accept the k24 crank journal widths.. It was an oem stroker that made gobs of torque with a 99mm stroke.

Reason i went with oem h22 pistons was because im an all motor type of guy and the s2000 blocks (f20c) are FRM sleeves and so were the H22s. So those pistons were ideal fitted with eagle rods.


Dont want to pull this thread too far off track so PM me if you know of anything!

 

[BATMANs]

Yes, plenty of strokers.

 

[Old Guy]

I vote for a slip on the timing belt or something else causing a shift in valve timing primarily because I am having a hard time with the idea that the valves floated on all the cylinders on the rear bank and none of the valves floated on the cylinders on the front bank. The first, and cheapest thing would be to check the valve timing marks. If they are out (intakes closing late?), then you likely have your answer. Might want to consider replacing the timing belt if it did jump a notch or two during your mis shift. You can assess the condition of the belt when you are checking the timing. If the timing is OK then consider the leak down test to confirm the worst. The leak down test requires special test equipment (which is not cheap) and a reasonable sized air compressor. It also benefits from someone who has experience with doing the test as there is some 'art' involved in interpreting the test results. Rather than return to the wizards at your dealership, you might want to check out a competent engine builder to do the test (Edmonton must have at least a couple)

 

[QPDO]

Thanks for all the responses! I called the dealership who installed the timing belt and they offered to check the timing for free next week. I guess I'll go from there. Being that the timing belt is virtually brand new (7500kms) I'll hope and pray that it's all good. That said, I'm not sure what's all involved in adjusting the belt on my cam gears so it might involve all the work for timing belt replacement again. I'm hoping the results confirm my early guesses and Old Man/dprada_1's interpretations of my information. It would sure save me some cashola!

Old Guy, I do have numerous mechanic friends with the leak down gear that have offered their help... I'll turn to them should the dealership tell me any bad news.

Cheers!

 

[d_prada1]

To adjust timing you shouldn't have to take the belt off. Just the inspection cover, loosen the tensioner, re position the cam gears, and tighten up the tensioner again. But like I said I am no nsx expert. Just off most other cars I have done. I'm not sure if the valve cover houses the gears or you can access them without removing them.

 

[JD Cross]

I'd like to learn something here.

Your readings were (155-145-147-215-205-200).

The NSX shop manual says the results of a compression test should have the following values:
Normal 199 psi
Minimum 142 psi
Max variance 28 psi

Does this mean the variance between cylinders regardless of which bank?
Would this mean your readings are over the maximum allowed?

 

[Old Guy]

I'd like to learn something here.

Your readings were (155-145-147-215-205-200).

The NSX shop manual says the results of a compression test should have the following values:
Normal 199 psi
Minimum 142 psi
Max variance 28 psi

Does this mean the variance between cylinders regardless of which bank?
Would this mean your readings are over the maximum allowed?

First question - yes. Second question - most people don't get too upset about being a little higher than normal. Some carbon build up can raise the test results. If they were uniformly high, the compression tester could be a little bit out of calibration. If the compression tester is a bourdon style gauge which is calibrated at sea level, it will also tend to read high at Edmonton's altitude. Finally, altitude affects compression test results (beyond the effect on the compression testing gauge). I can't remember whether Honda specifies sea level or some other altitude in the shop manual. However, I expect that is not at play here unless Honda specified the compression at the top of Mount Fuji! The important thing is that the variation between the cylinders is within the limits. If the average of the numbers is closer to the 'normal' value rather than the 'minimum' value, that would also give me more warm and fuzzy feelings!