slipstream444

I finally got around to running a leak down test on my S2000.
I haven't done a compression test as of yet, but the leak-down test gave me a good idea how things are doing in the engine. My car is about to turn over 40,000 hard driven miles.
I couldn't find the limits in the Honda Service manual - but my numbers were better than I expected. I used a Mac Tools leak-down tester.
The only leak down I could detect, was past the rings (sound came from the oil pan - which is normal. Noise from the intake would indicate an intake valve leak, noise from the exhaust - an exhaust valve leak, and bubbles in the radiator would indicate a head gasket leak). The test was run with a warm motor. I also ran a cold (dry) test (prior to the warm test) for comparison. The results were consistent, although the cold test showed a bigger difference in the #3 cylinder. The cold results are not used to measure the actual leak down in the cylinder, but can be used to detect a cold engine head gasket leak. I didn't run the cold test to look for a head gasket leak - only out of curiosity more than anything.

The numbers (engine warm) are as follows:
1) 1.2%
2) 1.2%
3) 1.5%
4) 1.4%

All the cylinders showed less than 2% leak down and were within 1% of each other - which has to be well within factory specs.

These numbers are consistent with other leak down tests I've run on previous Honda's of mine. Honda has always made a tight engine from the factory, but I attribute at least some of the results to the way I break in my engines, and the way I maintain my cars.
I don't follow all of Honda's factory break-in procedure(s).
I've used essentially the same break-in procedure for all of my new cars.
With my S2000, I basically drove it relatively 'gently' for the first 750-850 miles (following the manuals recommendation to avoid RPMs above 5,000 as much as possible), and then slowly increased the maximum RPM over time until I was working the full RPM range (up to the redline) by 1500 miles (or so). I think it's especially important not to lug the motor during the break-in as well. I switched my car to a semi-synthetic at around 1500 miles, and then drove the car quite hard (frequent redline runs) during the following 1500 miles (up to about 3000 miles). At 3000 miles I switched to full synthetic Amsoil, and ran the first synthetic fill for only about 1000 miles or so (continuing to drive the car quite hard). I then changed the oil again at around 4000 miles, and then began a normal oil change regiment.
I've never had a Honda with a leak down greater than 3%, but frankly expected to see around a 3-4% leakdown in my S2000 based on how hard I've driven the car over its 40,000 mile life. This includes some not so swift driving to include mechanical over-revs of 10,200 RPM as well as a couple 9400-9500 RPM over-revs. I've had my intake retainers replaced as a precaution due to the past 10k+ mechanical over-rev. All said - I'm pretty content with the results.
Please post any leak down numbers for comparison.

CoralDoc

Those leak down results are excellent!

I don't have results for my S2000, but the 124k mile engine in my Miata had the following results:
#1 32%
#2 15%
#3 18%
#4 6%

Pretty horrible, eh? The engine ran fine, but was down on power. I used that as an excuse for being slow on the track. I've since replaced the engine.

slipstream444

Thanks Doc. I was actually surprised by the results.
I initially thought I was somehow reading the guage wrong when I ran the first test cold. Cylinders 1 & 2 read 2% cold, #3 read 3%, and #4 was around 2.5%. I was expecting higher numbers when I ran the warm test, but the averages all were below 2% when it was all said and done.
The above listed numbers were averages for the warm test. I tested each of the cylinders from front to back, and then repeated the process two other times.

The numbers you posted for your Miata kind of look like dry (cold) numbers. An older engine may show cold numbers around 15-20%, but show less than 10% leak-down once warmed up. The number one cylinder was pretty high either way though. A good excuse to throw a new mill in there though!

CoralDoc

Those results were obtained on a warm engine. I had driven all the way up to Daytona Beach (about 245 miles) for corner weighting and tuning at a well-respected Spec Miata shop. The technician was pretty surprised by those numbers too. We managed to squeeze 95 hp out of that 1.6 liter engine, and in comparison, the new 1.8 liter replacement (with only 23k miles on it) puts down about 125 hp.

Enjoy your car - it sure looks like you're doing something right .

Slows2k

10% or less is considered acceptable for a street engine.

BlownAP

jyeung528

i have 6 to 7% on all 4 cylinders.

i also use amsoil 10w30.

CoralDoc

How many miles?

slipstream444

Those numbers also look like cold or dry numbers. The engine should be warm and not have sit for too long of a time (oil completely drained back down).

Additionally, how many miles were on your car before you changed over to Amsoil?

The break-in procedure I use (have basically used for on every car I bought new) I believe has a big impact on the results as well.
The first car I used that basic break-in procedure was a '89 Swift GTi, that I bought new. I used Castrol GTX (changed at 750 miles) for the first few thousand miles, then switched to a semi-synthetic until Castrol Syntec was widely available )at about 4,500 miles). I used Syntec for the rest of the time I owned the car. It had ~3% leak down at 110,000 miles. My '95 Civic VX had less than 1% leakdown at ~25,000, ~55,000, and ~85,000 miles. I used Castrol Syntec in that car as well. I should have never sold that car! My '99 Jeep Grand Cherokee V8 (4.7l) had around 3% leakdown at 40,000 miles. I used Amsoil 0W30 and 5W30 in that vehicle. I had a mechanic do the leakdown tests on those cars.

Only the Swift GTi was driven nearly as hard as my S2000. I used to beat the hell out of that poor little car! It was a fun little car.

jyeung528

break-in:

i stayed under 6k rpm and when i reached 600 miles i vtec'd and redlined.

changed oil at 7K using castrol gtx. switched to amoil at 10K miles.

been using amsoil ever since. over-revved and switched out retainers to ap1 (ap2 retainers were not known to be stronger at this time...as even billman didn't know the difference when this happened).

a warm leakdown showed 6% leakage at 1, 2, and 4...however 3 had 65% leakage due to bent valves.

head was repaired...and last month, at 95K miles, leakdown was 6-7%.

I do not know if it was cold or warm, but being that it resembles the numbers when leakdown was done warm, i will assume this was warm test also.

ben (supertech) did the leakdown test.