Which leak down test kit?

[anthonym]

K R500 , any recommendations and advice?

Anthony

[revilla]

I can't really recommend a particular model, I can only offer some thoughts based on what I found out when looking at them ...

The calibration of these is completely arbitrary. Knowing that you have "5% leakage" means absolutely nothing in itself as it depends entirely on the design of the tester. Unless you know what your engine reads on that particular tester when healthy you don't know what the reading is telling you. Big variations between cylinders are however telling you there's a problem.

They basically work by comparing the size of the leakage paths through the rings and valves with a "standard orifice" which is a small hole placed between the two pressure gauges. It's almost like an electrical potential divider circuit. The air flow (current) everywhere through the circuit from the compressor, through the standard orifice and then through the leakage paths is constant, so the pressure drops (voltage drops) are (roughly, it's not quite so simple with fluid flows) proportional to the resistances to flow at the various points. So in very rough terms if the leaks in your cylinders are about the same as the standard hole, you'll lose 50% of the pressure across the hole and 50% across the cylinder and call this a 50% leakage.

The problem is that there isn't really an absolute standard for the design of the standard orifice or the working pressure and things such as the diameter of the hole, the length of the hole and even the entry and exit angles affect the flow and therefore the readings significantly. So an engine that leaks 5% on a tester with a larger hole might leak 20% on a tester with a smaller hole as you're reading 5% or 20% of a different reference. On its own it's just a number.

The cheapish ones you see on eBay are useless. They only run at around 25psi pressure which isn't realistic, in that it doesn't put decent pressure behind the rings to push them against the bore walls to seal. When I tried using them I just got wildly inconsistent results. Winding the engine round a bit and back to the same point would give different readings each time. They are usually scaled with some random "set point" and then the second gauge just reads good, bad etc. so you don't know what actual pressure you are testing at. I would only use one where both guages are marked with proper pressure scales, psi usually or millibars etc. and you want a working pressure of 80-100psi, probably variable. A word of warning though, be careful using these pressures; as soon as you turn the engine even a few degrees from top dead centre with that pressure in the cylinder it will kick round aggressively and if you're there with a breaker bar on the crank pulley you can struggle to hold it and it can do some real damage! I tend to use a low pressure around 20psi to locate top dead centre and only once the engine is settled there do I wind the pressure up to 80psi to test it, then wind down again before moving the engine.

There are two relatively well established standards for the designs of these and it's worth getting one that conforms to one or the other as at least then you have a reasonable idea of the calibration.

One is "80 at 80" which means that if you connect it to a test leakage consisting of a hole of 80 thousands of an inch (0.2mm) diameter it will read 80% leakage. This is used a lot by American drag racers. It might not be entirely appropriate for our smaller engines.

The other is the FAA (American Federal Aviation Authority) standard for testing light aircraft engines. This is very well specified and there is a defined orifice for use with engines around our sizes. From memory it's something like 1mm (0.040") diameter, 6.35mm long (1/4") with 60° entry and exit angles, running at 80psi. At least with this you have a rough idea that your readings conform to some known scale and make some sort of sense in an absolute way, although it's still best to be comparing readings between cylinders or looking for changes over time from a known good reading.

If I were to buy one I would probably get one intended for light aircraft use, although you might need appropriate adapters for the right spark plug threads.

In the end I made my own. Basically just a pressure regulator followed by two pressure gauges with a standard orifice between them which I made very carefully to the FAA standard. I pulled the cheap eBay one apart and used the inlet and outlet pipes so I got all the connectors and spark plug adapters ready made, but made all my own guts for it. I has always given absolutely consistent readings and these have made sense in the light of what I've subsequently found in terms of engine problems. So for example a recent valve seat problem showed up as around 70% (56/80) leakage on a newly built engine:

One resolved this dropped to about 5% (4/80):

And on my freshly run in engine (you wouldn't normally do the test on an engine that hasn't been run in as you would expect some leakage from rings etc. but I was specifically suspicious of valve problems) it read pretty much zero leakage with 80/80 on the two gauges.

For a working pressure of 80psi you'll want a decent compressor that will produce 100psi or so (standard 7 bar) but it doesn't need to deliver much flow unless you've got real problems.

For working out where any leakage is going, best tool is your ears! You'll hear the hiss of escaping air in the inlets (open the throttles), exhaust or crankcase (take the oil filler off and listen in the hole). A cardboard tube can help. For this you want the compressor to have a decent capacity so that it spends reasonable lengths of time up to pressure and not running as you won't hear anything over the din a running compressor makes.

Hope this is useful!

Andrew

[SM25T]

A little light reading !!

[elie boone]

Indeed a good read but bush mechanic style is to convert a spark plug i.e. remove the porcelain and weld a connector to it, put it in 4th or 6th gear with the hand brake on and apply min.10 bar when the piston is at TDC then listen.

[oilyhands]

Most good quality testers allow you to calibrate them with a variable orifice, the tow most important factors for the result are variances between cylinders and the volume and position of any hissing. My Sealey tester has seen many years stirling service, I operate it at around 8 bar. Do make sure the engine is locked when testing at TDC and that you are on the firing stroke.

Oily