Compression loss when hot

[hiddenyeti]

I’m working on a Convert and when the engine is cold, I can get 145 pounds of pressure in the cylinder. Once you start the engine and it warms up compression drops all the way down to 60. It has new cylinders pistons and rings purchased from MG cycle And the cylinder head was just rebuilt. Any ideas where to look as to why I would be losing this compression?  We’ve done a static compression test by filling the cylinder, head with air and the only place we can find air leakage is up through the drain hole from the cylinder head to the sump to me that indicates rings, but everything is brand new and it doesn’t explain why I get good compression When the engine is cold.

Any help from the experts in the forum would be appreciated.

[guzzisteve]

Lock motor at TDC and do a leak down test or BDC with loose pushrods(no rockers). Did you fit rings to spec gap & spaced properly?

[RinkRat II]

     What valve clearances are you using?

     Paul B

[Antietam Classic Cycle]

Have you retorqued the head nuts and then readjusted the valves at least once since the rebuild?

[Chuck in Indiana]

Yeah, it sounds like a valve is being held open a smidgen. (technical term)

[Vecchio Lupo]

I recently went through similar issue on a 77 Convert. Start right up, idle smooth, run good, after a bit, good and hot, won't idle and motor stalled, re start and a bit sluggish off idle, as soon as rpms drop to under 1000, stall.

I followed the 123 procedure,

1 set the points
2 set the timing
3 adjust valves

easy start, great idle, no matter how hot, it just ticks over. Turns out my exhaust valves were open a touch and losing compression once expanded. this might not be your problem, but it sure as hell was mine

[Moparnut72]

I'll bet a buck that is the answer.
kk

[hiddenyeti]

Thanks for all the replies and suggestions. We are setting the valves at .009 we have attempted to re-torque the heads. 

Seems like the general consensus is it’s valve related, could warn valve guides do what being described?

At this point, I have a set of rebuilt heads that I purchased, but I’m not sure if they replace the valve guides. Will try that and I will get back to the group. Thank you.

[guzzisteve]

Yes, worn guides will lose air. .009" is good for setting gap.

[hiddenyeti]

Update:   Still have the mystery compression loss on the left cylinder.  New cylinder, piston, and rings.  We’ve tried 3 different heads with no change in behavior.  All combinations held a static pressure of 140 - 150 for days.  Once it’s hot the compression plummets.   

Starting to think we’ll never figure this one out. 

[guzzisteve]

Sould be cyl & piston, ring gap or alignment, specially if no fault in 3 different heads. Not much else.

[Shiny Hat]

I would think a leak down test should isolate the drop.
Also try a wet compression test and leak down to see if the pressure rises which would indicate rings. Not too much oil in the cylinder you’re just looking for a change.

[pressureangle]

If valve seating has been eliminated, you only have rings left. If the rings seal cold, then leak hot, about the only thing that can cause that is that the gap is too small and the ring ends are butting, causing the ring to wave in the groove; 90% of ring seal is between the lower ring surface and the ring groove land in the piston. Tight ends cause the ring to lift away from the land floor, killing seal. Next thing that happens is the ring scuffs the cylinder wall and probably breaks in two. Pull the cylinder and verify the ring gaps.

[wirespokes]

There are only three things sealing in the compression - rings, valves and head gasket. Valve guides have nothing to do with sealing compression unless they're so bad the valve doesn't rest on its seat. So it can only be one of those three.

By the way, Pete Roper recommends .005 and .007" for round fin engines and a thousandth less for square fins. No point in hammering the tappets.

As for the compression loss - a leak-down test seems like the next logical step. If nothing else, create a fitting to hook a compressor to the spark plug hole and give it 100 psi, then listen for the leak - exhaust pipe, carb, sump, or head gasket.

[Chuck in Indiana]

As pressureangle sez.. I'd check the ring gap.

[Alfetta]

going to jump in here and show my ignorance...
not sure if the bike has iron liners or chrome or nikeyseal...

but could it be that the liner is deforming with heat ?
Just askin.. because it looks like he has gone over the other conditions and not found anything.

I would think that with a ring gap issue, or an egging liner, that there would be an increase of blowby.
and that should be easy to detect or compare to the "cold" condition.
is there not a case vent directly forward of the flywheel, between the cylinders ?

[John A]

There are only three things sealing in the compression - rings, valves and head gasket. Valve guides have nothing to do with sealing compression unless they're so bad the valve doesn't rest on its seat. So it can only be one of those three.

By the way, Pete Roper recommends .005 and .007" for round fin engines and a thousandth less for square fins. No point in hammering the tappets.

As for the compression loss - a leak-down test seems like the next logical step. If nothing else, create a fitting to hook a compressor to the spark plug hole and give it 100 psi, then listen for the leak - exhaust pipe, carb, sump, or head gasket.

Right. The cam and lifter like less clearance so they don’t pound themselves to death. The valve and seat like more clearance, it allows more time on the seat when the valve has the best heat path to get rid of heat. To find where it leaks , pressurize the cylinder at tdc when it’s warmed up and listen at the sump for an air leak. Listen at the valves and you’ll pinpoint the leaks. You don’t need a differential pressure tester for that

[Chuck in Indiana]

Have you checked valve clearance when hot?

[wirespokes]

With .009" clearance I can't imagine it all going away after one heat cycle.

Just to make sure I understand - you pulled the left side complete - cylinder, head and piston - and replaced with spares? And still had the compression loss?

Or did you replace one thing at a time?

[pehayes]

Just covering all bases.   I think your cold and hot compression readings both suck.   Please describe your technique for testing compression.
Both spark plugs removed?
Throttle twisted wide open and held there by a bungee or an aide?
Crank for 3 to 5 cycles when testing each side.
Very common error that people neglect to open the throttles and then report very low compression readings.

BTW, what did you mean by "... we have attempted to re-torque the heads."  You either succeeded or  you didn't.  Like Yoda says, 'no try'.  If you did or do re-torque the heads, be absolutely certain that you readjust the valves AFTER the head torque.  The head torque compresses gaskets and quickly changes whatever valve settings you had.

Patrick Hayes
Fremont CA

[hiddenyeti]

I just wanted to follow up.  Turns out it was indeed the ring gap. 

The last time we took it apart (we were going to give it one last shot) we decided to check the rings as suggested by some of you who replied, even though we were sure we gapped them correctly. This time, after removing the rings, we checked the gaps again.  Two things were different this time when we checked.  We laid the cylinder on its side when we tested and used a feeler gauge removed from the set.  Sure enough, the gaps were too tight.  We had gapped them originally with the gauge still attached to the set and with the cylinder standing up. 

We regapped a new set of rings, with the solo feeler and the cylinder on its side, and now we’re holding pressure.  Whether it was the different way we held the cylinder and the solo tool, or we were just more careful this time, it’s working as expected. 

Go figure.   

[Chuck in Indiana]

You are supposed to remove the rings, and insert them one by one into the cylinder using the piston to keep it square when checking. It's not unusual to have to file the ends to get them within spec. You were lucky if it didn't score the cylinder. 

[matt franklin]

Thanks for the follow-up.  Too often, guys will report issues here, get a ton of advice, and then, we hear nothing...

[SIR REAL ED]

You are supposed to remove the rings, and insert them one by one into the cylinder using the piston to keep it square when checking. It's not unusual to have to file the ends to get them within spec. You were lucky if it didn't score the cylinder. 

 

[SIR REAL ED]

...... we were just more careful this time, it’s working as expected. 

Go figure.

Most of us humans are more careful the second time. 

Especially when the second time is trying to figure out what we did wrong the first time.

I have just enough dyslexia that I find measuring 4 times before being confident is a good thing.  I try to look at fixing my mistakes as tuition.

Glad it all worked out.

I'm curious, what were the measurements you got, first time and second time?