You will need a constant supply of compressed air and a leak down gauge to perform this test yourself. One face of the gauge measures the pressure of the compressed air that is being forced into the cylinder, while the other measures the percentage of this pressure that is being lost. A **** allows you to control the pressure going in so you get even results. You begin by taking out the spark plugs and bringing the piston (each cylinder at a time) to TDC (top dead center) so that all valves are closed. You will then connect the gauge to the spark plug well and the to the air hose. Apply the compressed air and modulate the **** to get a steady reading from both gauge faces. Use the same pressure on all cylinders. You will want your compressor full and stopped to achieve consistent readings for each cylinder. You will then listen to where you hear the compressed air if it is escaping. If you hear whooshing/howling at the oil filler cap, you are losing compression through the rings. If you hear it or even feel puffs at the tail pipe, it is your exhaust valves. If you can hear near the throttle body/intake manifold, it is your intake valves. If the results are inconclusive, it is probably your head gasket.
Warning: Please refer to your owner's manual to insure the correct way to prevent any ignition spark or fuel leakage prior to performing this test.

This was a copy and paste but hope it helps!

 

Maybe you can take that end piece and adapt a glow plug to fit on it, like you do to make a compression test adapter. I think you'll need to find a way (or a person) to hold the engine from spinning over when you apply the air pressure to the chamber.

 

It depends on the amount of pressure applied if I remember right. It's been a long time since I've done one of these tests.

 

Yeah... the instructions in the box and the videos I found are of little use on the 7.3L in particular. Our rings don't seal the way a gasser does (not even like many other diesels), and we obviously don't have spark plugs. The blue box in the picture shows a tool I made to adapt the tool to the glow plug hole. I've been using similar tools for years on the GP hole with other projects, and it seals quite nicely. The problem starts with the tool itself:

I've watched videos on the subject, and it shows setting the flow meter to zero with the regulator at 20 PSI (on a 100-PSI gauge). OK.... So why is there a 100 PSI gauge on a tool that is set to 20 PSI for about max? Let's shelve that for now. I set the regulator on the feed to the tool at 50 PSI to prevent supply dips - so that should not factor into this. Using the same procedure as the video, I can only set the pressure to 10 PSI to get the flow gauge to read zero in a dead-head. OK.... So I'm going to test big honkin' piston rings with 10 PSI of pressure - using a tool that has a gauge ill-suited for the purpose.

It took me hours just to perfect the tool and figure the magic formula to get repeatable results, but I can't trust what I'm seeing at a 10 PSI feed on such a big cylinder. In order for this to really work, I prefer 20 PSI at a bare naked minimum, because the cylinders "react" differently at that pressure - I get serious blow-out at the hole when I release the tool. I called a friend, and he said that because of our ring design - they don't even really seal until the pressure builds more than what this tool is allowing me to reach.

Even with all of that... the readings are not a disaster. They read far lower than a full leak, I get blowback at 10 PSI, and they are consistent across all cylinders. For a comparison, I threw the tool on AL (the white engine that was a fail), and the readings were markedly worse on that engine across all cylinders. I was told AL was an 87K-mile engine and Frank has about 180K miles. I trust the vendor that sold Frank, but it's abundantly clear to me now that the vendor for AL is a total huckster and I allowed myself to be blinded by his claims. Very expensive refresher course in "If it sounds too good to be true" 101.

 

Rich,

I think I sent this link to you before?

Engine Diagnostics: Compression Gauge Versus a Cylinder Leak Down Tester - Diagnostic and Specialty Tools & Equipment - 4WD Mechanix Magazine 'Tech and Travel' Forums

Finding TDC and on the proper stroke is the most important.

I wonder if a dial indicator would work through the glowplug hole? Maybe fit a threaded adapter to hold it in the same place (doesn't matter how "accurate" as long as it is held in a consistent position).

Rotate with crank bolt to see the highest reading, make a note of it, then put it into that position, when on the correct stroke, remove dial indicator, thread on whatever air adapter thingy you put together, and start your testing.

I think that's the best way to avoid pulling injectors, heads, etc.

What are you looking for?

Essentially when at TDC, ALL VALVES and such should be CLOSED. So you add enough air pressure to simulate combustion pressure or just a bit less.

By measuring how much/how fast the air pressure is dropping, you can gauge the "tightness" of the cylinder.

You may also be able to ascertain where the leak is (insert experienced mechanics' opinion here)

I recently saw a set of heads in the midst of having a valve job done. Or should I say "redone" as they just came back from a place that claimed to have done the heads and valves. Yep, they lost all future business... Lots of carbon buildup on the valves. And then the little seats. Unfortunately, I didn't have a place to recommend.

Lots of shops around, but WHO do you trust to have the knowledge and OCD required to properly do it?

Anyways, see if that article helps you.

oh- keep in mind that a "dry" cylinder will likely have some leakage around the rings, without the fluid seal around the piston.

 

Rich, you're on track when you notice that the results vary with different supply pressures. The final measurement of a leakdown test is not in pressure (psi), but in percent. There's a small orifice in the line between the two gauges, and as the air flows through it there's a pressure drop. You're taking the ratio of the supply pressure (pre-orifice) to the cylinder pressure. If you can maintain the supply pressure at 100psi, and your cylinder pressure sits steady at 80psi, you've got 80/100 = 80%, or 20% depending upon your point of view. Don't know if there's a standard, but it could be it's "leaked down" to 80%, or 20% has "leaked down".

If your supply pressure is 20psi, and your cylinder is 16psi, then you're at 80% again. If you know the diameter of the orifice, you can make a rather good approximation of the flowrate through the cylinder leaks, but an actual flowrate is not really relevant because, as you mentioned earlier, it's not equal to what actually occurs during operation, but the percentage should be somewhat close. You won't actually see the same percentages at different supply pressures however, because the type of air flow through the various holes will change from laminar (non-turbulent) to turbulent flow, and that change has a dramatic effect on the pressure drop through that particular hole. The type of flow is also dependent on the hole geometry.

A big takeaway from this is to add up the area of all of the "holes" in the cylinder and compare that sum to the orifice hole size. Think about it, if the two holes are the same size, then the pressure drop from supply to cylinder would be 50%, and then another 50% from cylinder to atmospheric. The typical orifice hole is around 0.040". So the sum of the areas of all the holes in your engine where the gas leaks out is less than 0.040".

It's just a more consistent way of checking internal cylinder components than spinning it with the starter. What is a "good" number for our engines? I tried a search for it online, and didn't find anything. Common gas motors are about 20% or less is acceptable, and for any engine, as cubic inches goes up, the leakdown number can also go up. 20% on a 231ci would be worse than 20% on a 460ci.

 

Good points! VOLUME is king. We're trying to use pressure to approximate volume.

The docs I read prefers to use VOLUME to establish leak down.

Here's a quote from the article I linked:



Because of the volume instead of pressure, I'd think you'd want the biggest supply orifice you can fit. You'd have to overcome the piston blowby volume and chop that off the top of your test results, with the remainder being due to valves, seats, injectors, whatever.

It seems this is likely the biggest reason you must have CONSTANT air supply, not just fill it like a tire and see how fast it bleeds off to determine the amount of leak.

It also seems to reason that there is a "minimum" supply orifice one can use to supply air.

Ideally, a dual-chambered supply tube that had one side supplying air and the other measuring pressure could be used. Maybe that is what the tool is trying to do, too?

FYI. I am trying to reason through the test and equipment, so please correct anything you see in what I post

Pressures for our trucks? Here's another quote from the article, just as an example of a DIFFERENT, gasser:



Rich- you gotta move closer if you're going to keep doing stuff like this, LOL.

 

First... getting TDC is a cakewalk with an engine "on the bench". Second... there is one **** and one outlet on this device - not a lot of control there.

The video shows this to be as simple as torquing a bolt - but we all know how inept I am at even that. Dead-head air line, turn **** until leak gauge reads zero. That's it... except for don't exceed 100 PSI supply to the tool. The left gauge reads 10 PSI when the right gauge reads zero.

I went with 50 PSI feed to keep the supply constant. I'm now reading that 50 PSI feed ain't gunna cut it. I also theorize my 75 feet of supply hose can be a big factor. I'll yank the 50' of extension hose off the spool, set a higher supply pressure, and try again.

I need a diagram of this bass turd, so I can figure out how it works.

Oh... and this engine puffs up to 440 PSI on a cold dry crank with an OBS starter on my wife's single Jeep battery. See video in first post.

 

Hey Rich, sounds like you may be at TDC, but maybe on EXHAUST stroke instead of compression stroke? 90% loss sure doesn't sound right

Knowing you, I am guessing you've tried other cylinders, just to make sure you don't have one dead cylinder?

Tell me how you're going to TDC. It will be different on the other cylinders, assuming #1 at TDC is marked (is that what you mean about it being easy?)

Not trying to sound insulting or anything- making sure I am on the same page as you. Also- when is the last time there was fuel in the cylinder?

The way I'm thinking, you've got to be on the wrong stroke...if not, with those numbers, you should be able to shine a flashlight in the GP hole and see it through the side of the block

Do you still have the bottom pan off?

 

Who said I had 90% loss? I have a dead-head... Zero flow. Turn *** right until the leak gauge reads zero. At this setting, gauge on left shows 10 PSI. This is before I even wave a picture of the engine at the tool.

 

Aha! I'm not stupid or crazy (in this instance alone)! My friend White Buffalo found this for me, regarding this tool:

 

Yeah... I was wondering about that myself. My first post describes the method as you just did, using two equal pressure gauges. When you mentioned a percentage gauge, I was wholly confused. That "percentage" gauge, is a pressure gauge. Make sure that when the thing isn't connected to the engine, and it's not leaking any air, that both gauges read exactly the same.

See my earlier post, and you can use whatever psi you want, just remember the math. As someone else mentioned, higher psi lets you hear the leak better, although most of it should be venting thru the rings and you'll hear it coming out through the drain back holes in the cylinder head.

I would set your line pressure to over 125psi, and set their regulator for 100psi. That way your second gauge will read in percent. And if the regulator's crap, screw in another regulator. The only thing that's hard to come by on that rig is the manifold with the orifice hole in it. While you're running the test, the supply gauge shouldn't budge.

 

Everything shared by others makes perfect sense, it's just the tool that was jamming me up. I was getting confused with it not doing what I expected nor could make sense of. I just so happen to have a 100 PSI gauge sitting in a box... from the last time I attempted a leak-down test on Stinky. I also have the 1/4" NPT to 1/8" NPT adapter I need to replace the impish "psych" gauge on the Horrible Fright tool.

 

Here's my waiting-for-coffee-this-morning schematic drawing of the leak down tester. It's all about what-goes-in equals what-comes-out, but since we're talking about air, which is compressible, this only holds true once everything is in steady state. Once the gauges stop moving, then you've got some info you can bank.

I was reading up that other link, where there are some good points about removing the pushrods, allowing you to perform the test with the cylinders at both TDC ans BDC. If you're not going to perform a compression test - or already have - you may as well play with this too. Without the pushrods, it'll be easier to spin the motor and you can find the individual cylinder DC positions using a long straw sitting in the glow plug holes.

 

Once I realized that stupid "leak" gauge was actually a 0-10 PSI gauge marked in reverse, I envisioned exactly what you drew. I have already modified the tool (there goes the warranty), so I'm now armed with understanding and the right test equipment. The only shortage I have is time. My Chi took a big hit when I lost all that analysis time to a box of cheese in the guise of a tool.