Doing my first leak down test, is the pressure from the compressor supposed to stay on? When I turn it off the pressure on the tester regulator and the cylinder drops from 100 to 0 in half a minute. Is that normal? I’m thinking it is because none of the YouTube vids I saw mentioned shutting off the air supply. Second question: is it normal for a piston to move when I very slowly increase the regulator output to just 20 psi? #4 cylinder took 100 psi without moving but #1 moved at 20 psi and 2 moved at 35 psi.
All three cause air leakage only at the timing chain, which is the crankcase? The valve cover is off for finding TDC. Meaning rings are bad? Pressure is only 2 psi less at cylinder compared to regulator gauge. Don’t see a % reading on the cylinder gauge like all the ones on the vids have
many thanks

 

I would recommend watching these two videos. Pretty detailed info on leak down testing.

 

Spent 23 minutes watching the Timmy one, after watching 8 others, none of them answered my questions. I think after I get a cylinder on compression TDC I’ll put the tranny in gear to prevent the cylinder from moving. I’m still curious though, whether 20 or 30 psi moving the piston indicates a problem.
As to whether or not constant pressure from the shop compressor should be necessary I’m assuming it is until someone with more experience says otherwise.

 

Yes, you need constant shop air pressure. What a leak down tester is really measuring is flow: How many mL/sec flow through your engine at [30psi]. If you're engine was tighter than ______ (which it could never be), you'd get 0 flow.

Should 30psi move a piston? I would hope so! To do a leak down test you need the piston to stay put at TDC. If the piston is set exactly, there is no torque either way. But because setting the piston isn't exact, you need to hold the piston still somehow. As long as you get it really close, the piston might not be able to overcome the friction, but since it's dangerous to have the engine whipping around you'll want to put it in 5th gear or something.

Your leak down tester doesn't read (directly) in flow, but in "percent." What is a good percent? Well, I don't know. It depends on how well your tester is calibrated (not very well). So the number you get is really only useful for comparison. Measure your engine, replace the rings and head gasket, measure again.

So what good is a leak down test? Where it really pays off is by your listening. Is their bubbling in the radiator? Head gasket. Can you hear air in the throttle body? Intake valve. Exhaust? Valve again.

 

"Cylinder Leak-Down Testing
The ALLDATA Tech-Assist Team
...1. Warm up the engine, so the rings have a chance to seal.
2. Pull the spark plug out of the cylinder you want to test.
3. Turn the engine over by hand until the piston in that cylinder is at top dead center (TDC).
NOTE: Often, when air pressure applied to the cylinder, the engine will turn over. Try to have the piston as close to TDC as you can so the crank, rod, and piston are as vertical as possible. Leaving the other spark plugs in the engine will help provide some resistance to turning. WARNING: Remove the breaker bar and socket from the crank nut before putting air to the cylinder. If the cylinder you’re testing is not perfectly on TDC, the engine will turn over quickly and the breaker bar can strike you or damage under hood components.

...Conclusions: The leakdown tester is a great diagnostic tool to identify real cylinder pressure difficulties. But don’t be alarmed if you get 15 to 18 percent leakdown on all cylinders. If the cylinders perform within a couple of percentage points of each other, look elsewhere for any perceived problems." https://www.alldata.com/sites/defaul...ing_011019.pdf

 

Very good info, thanks, One plug out at a time makes sense. The three I checked had just a 2 psi drop between the two gauges which is quite good.
I think you mean it’s measuring the pressure of the flow not the flow. Flow wouldn’t measure in psi but cfm or some other volume per time.
Yes I wasn’t going for perfect tdc just for both valves closed ( valve cover off). Which was also contributing to the piston moving. Hard to turn the crank from above on the 2.4

 

This video actually answers your questions. I 've just watched it again to make sure.It might seems less polished than some, but he does a good job with the demo. But Scope103 answered them as well.
Officially, as mentioned in the video i linked, a leak over 20% is bad. That said, we preferred a number less than 10%, and got nervous with readings from 15-19%.
We tested a new engine once out of curiosity. It was 2%
Much like a compression test, you'll want a variance 10% or less from your highest to your lowest.

 

Yeah, you'd think. The gauge is measuring pressure drop through a (supposedly) calibrated orifice. Which is a measure of flow (and is why you have to keep the shop air running). If the orifice was actually calibrated, the pressure drop (sometimes displayed as % pressure drop) would be directly proportional to flow, and "a leak over 20% is bad" would be meaningful. If you use the same tester over and over (as I imagine jimkola did) then you'll learn the "good/bad" percentages. But for me (only using it on 2-3 vehicles), it's only really useful when I have an earlier test to compare to.

 

No matter what the gauge reads, you'll want to see all four cylinders fairly close to each other. if you have a significant variance then you know you have issues.
And being able to detect for possible air leaking/pressure loss at the four major spots is the huge benefit.

 

Thanks both of you, as I understand it you are warning against relying on exact measurements and looking for relative readings, which is all good but there are different kinds of test assemblies. Mine I think is the simplest because it doesn’t seem to have the orifice you mentioned or the % loss reading, like I’ve seen on some vids. Instead it’s got a table where you derive the % loss from reading the two pressure gauges. Which on my kit are identical (0-100 psi). The simplicity reduces instrument error, seems to me. I have a OTC. The regulator keeps the incoming pressure at an extremely stable reading, and the cylinder’s pressure reading also stays extremely stable, the leak being a constant size, making reading the two very easy. No needle bounce. On the three cylinders I tested I had a 2% drop. I’ll test the 4th if I have any issues.
Its easy to figure the % loss if the pressure in is 50 or 100 psi of course, without the table

 

So what's between the two pressure gauges? If it was just a straight open passage, the gauges would always read the same. Your down-stream gauge reads lower (by some percent), which is due to the pressure drop through the orifice. The pressure drop is proportional to the flow through it (and on into the engine). If there was no flow, then the two gauges would read the same, no matter how small the orifice.

So you're good. Your leak down tester works like all the rest, just with the table-lookup step. The advantage of your arrangement is that the two 0-100psi gauges are inexpensive to replace, and you don't need to replace the face if you do.

 

Thanks, it still seems too simple but I guess some engineer calibrated it right.