What sort of psi am I looking for in a compression test on an M21?
I feel that power is down and want to start investigating.

Hi Matt..'They're all like that, Sir'..
If you are of average weight and you get the engine onto the compression stroke the kickstart should hold your wieght, perhaps just turning the engine gradually through compression. If you (or someone else) has done a few miles since the engine was last apart, chances are it is time for a regrind of the exhaust valve seat. A 'leak down' test is the thing to do..I think Hans has posted something on the procedure on the Technical section of the website...Ian

email (option): ian@wright52.plus.com

The two words "Power and M20" does not compute!

Around 60 psi is about right, normally that figure would be low on most bikes, but on the M20 it seems to be about an average figure, to check where it is leaking you really need a differential compression tester which allows a given amount of air in the cylinder and it's leakage is measured, you can make a "makeshift" one by welding an air line fitting on an old spark plug body and allow about 30psi into the cylinder and feeling/listening for leaks at the carb/exhaust/and engine breather for leaks to pinpoint where there are any failings

email (option): davmax@ntlworld.com

Yes Matt, there is a description of a DIY leakdown tester in the Technical Info page of the website

email (option): viaconsu # planet dot nl

Thanks guys, forgot about Hans' leakdown tester, will try to russel up some parts for it.
Sorry Douglas, should have said it's ability to gain speed (notice I didn't say accelerate) is reduced.

You should only really need to make a spark plug with an airline adapter if you already have access to a compresssion tester, basically you'd only have to see if the comression is about the right reading, and if it isn't you'd just put the adapted spark plug in there just to see where it is leaking from. Here's a hint, use your moistened lips at the exhaust/carb or breather to feel for the leak as they are more sensitive to a slight loss of air (as you would finding a puncture) Hope this helps
Dave

email (option): davmax@ntlworld.com

Thanks for the suggestion Dave.
I'm going to check the compression tomorrow, I'll have to wait 'till Tuesday after easter when I'm back at work to make the adapter.

Reading the leak tester proceedure, shouldn't there be a time factor involved somewhere? Leaving aside the question of whether a leak changes the carburation, a leak affects engine power only on the compression and power strokes and for the greatest part of each of those strokes the pressure inside the cylinder is nowhere near the 60 or 80 psi maximum that it reaches "eventually" (although on the power stroke it clearly far exceeds those numbers for the first part of the stroke, just after ignition). One stroke of a four stroke engine at 4,000 rpm takes less than 4 milliseconds. Is it realistic, in testing, to put on the pressure and then allow the leak to work for several seconds while a reading is taken? Am I reading the proceedure correctly?

email (option): jonny.rudge@verizon.net

email (option): teladelujo@msn.com

I must admit I never do a leak down test if I suspect compression is down...I just take the top end apart and check everything...you have to do that anyway once the leak test has confirmed what you probably already know...Ian

email (option): ian@wright52.plus.com

It really is such a simple engine to remove the head that I'd be tempted to do a valve grind first, ive even cut head gaskets out of old hot water tanks when funds were low to remove cost from the job :-)

It's actually more of a "rate of flow" test... you apply an air supply at a specified pressure to the cylinder with the piston held at TDC, and see what the percentage rate of loss/flow is through the valves and around the rings.

It is a much better diagnostic tool that a compression test alone, as it allows any points of leakage to be identified specifically, and the overall "health" of the cylinder to be determined.

JDE[/quote]

Exactly my point: in this type of testing you are getting the "rate of flow" with a relatively constant gas pressure and allowing the leak to work for a longer period. But the actual pressure in the cylinder at work varies from a negative number at intake to a number much higher than 5 atmospheres at ignition, several times per second. Just doesn't seem that the testing is done under the same circumstances as when the engine operates. Finessing this test to the point of saying, if you get 96% you're OK but at 92% you need a major repair, I don't know about such conclusion. I should think that such figures would be much more reliable when testing with a compression tester (i.e., while the engine is rotating), even if this doesn't give you a specific place for the leak. Once you have determined that there is a compression problem or leak, then you can pinpoint the leak by using any of the leak testing systems, such as the colored liquids used to find leaks in automotive air conditioning systems.

email (option): jonny.rudge@verizon.net

Ian Wright

I must admit I never do a leak down test if I suspect compression is down...I just take the top end apart and check everything...you have to do that anyway once the leak test has confirmed what you probably already know...Ian

email (option): teladelujo@msn.com

Just out of interest, the maximum permissable difference between air going in and the leak should be no more than 10%, personally I too never do a leakdown test on any singles, usually you know it has a problem with either the valves/piston/head gasket so it is quicker to just start by taking the head off (after checking the tappets) and working down to taking the barrel off. It seems a little bit like buying a £1500 scanner to read a trouble code on a £200 moped!

email (option): davmax@ntlworld.com

Hey! Don't tell everyone... I got a non starting deisel Lister Dumper truck cheap once because the tappets had no clearance & needed 10mins work.

It's always a good idea to check the tappets and valve lifter adjustments first, especially as BSA's tappet setting method was different to most motorcycles

email (option): davmax@ntlworld.com

Not just different ... Impossible on my M24 there is no way to check other than experienced guess!

(I stayed clear of the word educated through fear of reprisals)

Well, I checked compression today and surprisingly it came to 80-90 psi. Checked tappets and valve lifter, all fine. Wondering if it's just my imagination now, however there was a discussion here late last year about the mixing chamber union nuts on the carb , I'd discovered that I had the mid size fitted and changed it to the short one. Now My 276 carb is an amal copy Made here In Australia in the early '50s due to the lack of supply from the UK, it has a bronze or brass body but uses all amal components. This carb body may be skewing the mixture, I'll refit the mid size union nut and see if things improve.
What I'd like to understand is why does altering the mixing chamber union nut and therefore altering the volume around the main and needle jet make such a difference, since fuel is being constantly supplied from the float chamber?

The atmosphere must be heavier "Dahnundah" than the rest of the world. Atmospheric pressure being about 14½ psi, with a compression ratio of about 5 you would get less than 73 psi at 100% efficiency. Doesn't look like your 80-90 reading can be correct, if your combustion chamber has not been altered.

The union nut doesn't alter the volume of air around the jet or the needle, it controls the level of the fuel in the float bowl relative to the height of the jets in the mixing chamber. With the throttle wide open you would get the same compression reading regardless of which union nut you have. Just that with the "wrong" union nut you would get a very lean mixture or a very rich mixture (or, in the second case, flooding in the mixing chamber). More likely, with the wrong union nut she'd be difficult or impossible to start or run.

email (option): jonny.rudge@verizon.net