DIY Leak Down test experience and How To

Joined
21 August 2011
Messages
150
Location
Northern Virginia
Thought I would share the steps I took to do a leak down test. Below the procedures is a short explanation of my experience doing the test that led me to do this write up and maybe alleviate the same issues I ran into.
<O:p</O:p
Total time about 45-60min with air tools<O:p</O:p
Difficulty- Easy (slightly harder than changing the spark plugs)<O:p</O:p
Cost –one time investment of 59.99 for the OTC brand tester on Amazon (Prerequisite – Air compressor that can supply more than 50PSI)<O:p</O:p
Tools – Leak down tester (OTC, Matco, or Snap-on stay with a name brand), compressor (50PSI or greater), 10mm socket, air ratchet, 5” 3/8 extensions, 18mm socket, various ½ extensions total of about 16” or so, breaker bar, Spark plug socket, long screw driver.<O:p</O:p
These are the steps I took. <O:p</O:p
1. Warm up the engine (10-15 minute drive).<O:p</O:p
2. Put car on jack stands.<O:p</O:p
3. Remove right rear tire so you can rotate engine form the balancer bolt using a 18mm socket, ½” extensions and breaker bar.<O:p</O:p
4. Remove coil covers and coils.
5. Remove the spark plugs one at a time for the cylinder you are testing. (Make sure that hole that spark plug sits in is clear of water and/or debris and if not clean out with compressed air or other method before removing the plug.) <O:p</O:p
6. Insert long screw driver into spark plug hole and rotate engine clockwise to the TDC for the cylinder.<O:p</O:p
7. Remove screw driver and screw in 14mm rubber extension hose that comes with the leak down tool.<O:p</O:p
8. Attach air compressor hose to leak down tool and adjust leak down tool per instructions then attach extension hose to read leak down number. <O:p</O:p
9. Record the number (if low you may need to rotate the engine and find TDC again then retest).<O:p</O:p
10. Replace the sparkplug and go to the next cylinder.<O:p</O:p
If you have a percentage loss of greater than 10% or you are curious like I was you can try to locate where the leak is coming from. You can do this by using a long rubber hose and putting it up to your ear and the putting the other side on the exhaust header on the port for the cylinder you are testing, then the intake, and then just inside the oil fill cap. I did not need the hose I was able to tell that when I took the oil fill cap off the air was coming from there as the pitch of the escaping air changed. If you cannot determine where it is coming from you can try removing the adjacent cylinders spark plug to see if it is leaking through the head gasket. Not sure if this does or can occur on the NSX but worth looking into. You also want to keep an eye on, or put the rubber hose your were listening with up to, the coolant overflow as a blown head gasket could be leaking right into there.
<O:p</O:p<O:p</O:p
These Instructions above are based on my experience outlined here:<O:p</O:p
Decided to check the health of my engine and thought I would share my experience completing a leak down test. First the car is 92 with 143k miles on it. No problems but thinking about turbo charging in the near future. I had previously down a leak down test on a 69 mustang when I was in high school but not since. I went to Harbor Freight tools and bought there leak down tester. <O:p</O:p
I warmed the car up with about a 10 minute drive. Then pulled into the garage and let it idle why I jacked it up, removed the passenger rear wheel and the coil covers. I then turned the car off and removed the coils and then the first spark plug. Next I put a long screw driver into the spark plug hole and rotated the engine (clockwise using a 18mm socket on the balancer nut and breaker bar) to TDC (highest point the screw driver goes). I then screwed in the 14mm hose that attaches to the leak down tool. I then attached the leak down tool to my compressor and zeroed the second gauge by adjusting the regulator on the leak down tool (while not attached to the engine) this took about 15 PSI on the inlet side of the first gauge. My compressor regulator was set to 100PSI. Next I hooked up the adapter coming from the engine and checked my leak down percentage. This is where I began to get concerned. 1<SUP>st</SUP> cylinder was right near 35%, 2<SUP>nd</SUP> 25%, 3<SUP>rd</SUP> 30%, etc. Now the gauge says that 0-40% is good but I read that if you have leak down number above anywhere between 15-30% you should rebuild. I exchanged some PM’s with Larry B and he thought those percentage indicated a well worn engine and that he typically sees 5-10% on a good low mileage engine. <O:p</O:p
Well I could not understand why my engine was so bad. The previous owners had religiously changed the oil and did regular maintenance. Based on the receipts they took the car in with minor issues and did all the recommended service found during the visit. I have all the oil change receipts with changes every 1500-3000 miles on average. I started to look online and read some concerns about low pressure leak down test that are less than 50PSI not being accurate and also issues with leak down testers that give a percentage instead of a PSI on the second gauge. Ok, but how could it be off more than a few percent. I ordered the OTC brand leak down tester off of amazon for 59.99 and redid the test.<O:p</O:p
Prep is the same for both leak down tester and only differs on how you "adjust" the tool before connecting to the rubber hose extension coming for the spark plug hole. Once ready you turn the regulator on your compressor to supply about 110PSI then the regulator on the leak down tester to 100PSI. Then connect the hose (already screwed into your sparkplug hole) to the leak down gauge. If you are very near TDC the pressure from the leak down tester will not push the cylinder down and then you subtract the second gauge number form the first gauge to get your leak down percent. If you have problems with the pressure of the tester pushing the engine over you can adjust the regulator on the leak down tool to a lower number and slowy increase it. If you play around with it you will see how it works. The inside of the case for the tool even has a table that shows the differrent leak down percentage based on the input pressure on the first dial and the reading on the second so that you do not have to use 100 PSI if you do not wish to do so. My first cylinder was at 9%, 2<SUP>nd</SUP> 3%, 3<SUP>rd</SUP> 4%, 4<SUP>th</SUP> 6%, 5<SUP>th</SUP> 2%, and 6<SUP>th</SUP> 3%.

So how the heck are the numbers so different? Well from my research low pressure testing can wildly vary, the orifice size between the two gauges on the tester may not be optimal, orifice should be sized for the size on the engine (bore) so if you do the test make sure you use a leak down tester that has a PSI gauge on either side, can test at 50PSI or above, is not made by harbor freight, is made by OTC, MATCO, Snap-On, etc (appears same manufacturer builds the OTC and MATCO units).<O:p</O:p
 
Thanks for this write up. It helped in doing my own leak down. I also used the OTC kit.

I followed your steps with a few exceptions.
-Used a 19mm socket to rotate the motor from what I believe is the balancer (pulley directly in front of the oil filter)
-Didn't heat up the car since this leak down was to check for engine damage from overheating. Didn't want to risk more damage.
Cyl results were: 4%, 11%, 4% ... 7%, 8%, 26%

I checked the last one 3 times, with the same result. Looks like the overheat damage that area.

Thanks again.

Mike B
 
Thanks for this write up. It helped in doing my own leak down. I also used the OTC kit.

I followed your steps with a few exceptions.
-Used a 19mm socket to rotate the motor from what I believe is the balancer (pulley directly in front of the oil filter)
-Didn't heat up the car since this leak down was to check for engine damage from overheating. Didn't want to risk more damage.
Cyl results were: 4%, 11%, 4% ... 7%, 8%, 26%

I checked the last one 3 times, with the same result. Looks like the overheat damage that area.

Thanks again.

Mike B

what were the circumstances surrounding your car overheating?
 
When I learned how to do these type of compression tests it was on piston engine airplanes. This is a two person job.

This is how we did it.
1. We would remove all of the top spark plugs (there is a set on the bottom too).
2. Install the air hose w/spark plug hole adapter in the spark plug hole.
3. Stick your finger over the end of the adapter.
3. Rotate the propeller until the air stops coming out of the hole.(This will be TDC on the compression stroke)
4. Attach the gauges.
5. Hold onto the propeller.
6. Slowly increase the air pressure up to the desired pressure, usually 80 psi.
7. Rotate the prop back and forth until you find exactly TDC.
8. Read the pressure on the residual pressure gauge.

Brad
 
Last edited:
what were the circumstances surrounding your car overheating?

Driving through Malibu canyons. On a downhill section, coolant hose burst. Didn't realize it had burst until about 5 minutes later when I can to a stop light, and noticed a little burp in the idle. Noticed my OE water temp gauge was peaked. I GoPro all my driving in the NSX, so I was able to review the footage and see when the temps began to rise and when I caught it.

I'm going to do another leakdown and see if I can find where the leak is. I'll post my findings.

Since I just purchased this car and haven't driven it after the incident, I really don't know how the overheat affected it. I know it starts and runs, but for all I know, this cylinder could have already had a 26% leak. Either way, the consensus seems to be not to risk driving the car to prevent further damage.

Mike B
 
I know this thread is a year or more old, but found it useful and thought I would add to it rather than starting a new one after I just had this done (not using the HF gauge, I think he used OTC). Some questions/observations:

1) Aren't you missing the step to basically stand on the brake with car in gear? This prevents the piston moving, no?

2) Why does it have to be at TDC? And if it isn't, what does that mean?

Mechanic pulled the big screw-in fuse and had me quick crank half a turn until he got to where he wanted, presumably at or close to TDC? Or does this approach mean you need to go with #1 because of not being at precisely at TDC?

About a year ago I had a compression test done which showed: front L-R: 176-168-175 and rear L-R: 172-165-150. Right rear plug/coil has been fouled at least twice by water leaks.

My leak down test done: 3%-6%-2%-3%-1%-2%.

My main goal here was to determine the health of my 152k high boost SC motor. Comments on my numbers?

Hopefully LarryB will chime in!
 
The reason to be at exactly TDC, is that the connecting rod is straight and the leakdown pressure will not move the piston down;)...... In addition, consistency in the readings may not be there if the pistons are in all different positions within the cylinder.

BTW, typically I use 100psi for a leakdown, and using the TDC method, I have never had the lock the engine in position.

HTH,
LarryB
 
Last edited:
Thanks Larry, was hoping you would chime in! Any ideas on how far off the consistency would/could actually be given it's air pressure and the cylinders are "in theory" even all the way down. We were using 100psi. Not to hijack the thread any further but do you see a lot of cars with right rear plug/coil fouling?

Thx!

andre
 
I do not have a good feeling about the amount of inconsistency. Rear Plug/Coil issues are typically related to the coil cover gasket, water getting in there. Too much heavy water during a wash;)....
 
Last edited:
Heavy water? Like loaded with Uranium or something? :eek: I know of another local member who had the same issue, I should source a new gasket I guess....keep the heavy metals out. Thx Larry!
 
By heavy, I really meant high volume of water, like aiming the hose in a way that the water stream enters the engine bay:)
 
By heavy, I really meant high volume of water, like aiming the hose in a way that the water stream enters the engine bay:)

Lost track of who's joking with who at this point....:tongue:
 
Back
Top