Atkinson frustrations

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Gordon

Well-Known Member
HMEM Supporting Member
Joined
Jan 4, 2011
Messages
1,330
Reaction score
343
This post is to vent my frustration but if someone has some insight it is certainly welcome. I am building the Atkinson Differential engine from the Gingery plans and I am not having any luck in getting to run. Gingery said that he had to turn the engine over with an electric motor and a belt to the flywheel for up to 40 hours to seat the rings before it would start dependably. I have been doing that for quite a few hours (25-30) but I am not seeing any improvement. I am sure that the problem is compression but a solution escapes me. Due to the unique design you can only feel the compression on the last portion of the compression stroke because the spark plug hole is covered until the last part of the stroke. There seems to be good suction at the spark plug hole on what would be the exhaust stroke.


At this point it kind of fires about 1/3 of the time and even that is intermittent. If fires for a while and then quits firing until something like carburetor adjustment or cool off occurs. I have tried different spark timing and different carburetors and carburetor adjustments but nothing seems to change. I am using the S&S ignition and there seems to be a good spark.


At this point I am reluctant to remove the pistons/rings for inspection because if they are actually seating the rings are going to be in a different rotation when it is reassembled which I assume would at least partially disturb the seating in process.


There seems to be a surprising lack of information and build logs on an engine which has been built by quite a few others. There are quite a few Youtube videos. There are several errors or at lest omissions on the drawings but I have not seen any reference to this by any others. I have been over by engine and checked it to the drawings many times and as far as I can tell it is made correctly.
 
Does it move the intake valve? that is just moved by the suction from the cylinder, and the spring is important. Too stiff a spring here would make it not get enough gas. Make sure there isn't too much friction on the intake valve.

Does it run assisted? given that it has gas available, does it run when connected to a electric motor?
Have you tried running it on a dab of starting fluid?

As long as you can feel the push of the compression, that should be fine. Is your timing of the spark correct? The way that the timing is explained in the book its faily easy to set it up 180 degree wrong, there is a youtube video around that explains it, but I was not able to dig it up.

The Atkinson differential is perhaps one of the worst engines to get running reliably, but it's a fun engine!
 
Intake valve seems to move well. Visible movement and I have tried stronger and weaker springs.

It kind of runs with the electric motor. I can hear it fire about 25% to 50% of the time. I cannot get it to fire when turned over by hand. I have also tried weaker and stronger exhaust valve springs. I does not really try to run on starting fluid. Just one pop. It only seems to fire with the carburetor adjustment just barely open. I have tried making a smaller orifice but that has not made much difference. I have also tried choke plates on the intake throat.

I am sure that the ignition timing is correct or at least not 180° off. I have tried running it both up to 15° BTDC and ATDC.
 
Just to cover all the basics, power stroke should be the long stroke, and compression stroke the short stroke.

What kind of fuel do you use?
It may be that it needs more time to run in, but I would expect it to fire and run for a short time after 30hr.
 
I am using Coleman fuel. I also thought that after running with an electric motor for that long it should show more of a tenancy to fire. I guess I will just try running with the electric motor for a while more.

Am I correct in my assumption that removing and reinstalling the rings would affect the seating or are the rings rotating in the grooves as it is running?
 
In my mind, the procedure for "running the rings in" is completely wrong and is likely to be the direct cause of poor compression. Best full size practice (and what I use successfully in model size) is not run the rings in at all and use combustion forces to provide correct ring pressure to create a seal. The absence of these combustion forces means the rings will wear in without creating an effective seal and will most likely never seal fully. In a full size engine, excessive idling/low load of a brand new engine results in the same thing and is known as 'glazing'. I wonder if the specified running in procedure is why this engine is known as difficult to get running?

If it was my engine, I think I'd be pulling it down, lightly honing the bore (as I'm thinking it will be highly polished from the running in procedure) and trying a fresh set of rings, with zero running-in and see if it runs then.

In my models (and this is all based on cast iron rings which I assume yours is) I lap the valves in and assemble the engine without spinning it at all. Once fuel and spark is set I try to start and the first few pops and bangs seem to increase the valve sealing and compression increases, as does the frequency of the pops and bangs. A few tweaks of the carb and maybe the timing later and it (hopefully) starts. After a few runs (maybe 10-15 minutes total running) compression is noticeably higher than what it first was, due to the rings bedding in properly. I've included 2 videos of the same engine using this procedure. The first is the very first start of it, and the second is with around 15 minutes of running on it. In the second I attempt the first hand-start of the engine around the 1:45 mark.



 
Cogsy: Actually that is pretty much what I have done. I honed the cylinder and lapped the valves. I have remade the rings twice. I have been running it with the electric motor with both ignition and fuel operating. I have not run it at all just spinning free without fuel and spark. As I stated it fires part of the time but not enough to actually run on it's own. I does seem to be improving. If I try to hand start it it will keep spinning for several revolutions because the firing is at lest helping to keep it turning.
 
Also make sure that the entire ignition system is working properly, as the plug may not be firing under compression pressure. That has fooled me a couple of times in the past.

If you're spinning the engine with an electric motor, it should fire consistently even if it has no piston rings installed.

Does it emit black smoke or unburned fuel from the exhaust? If not, try enriching the mixture until it does smoke, so you know where you are in the range of fuel / air mixtures.

Try propane as a fuel (but not in your enclosed basement) to eliminate poor fuel atomization as a cause of the problem. The same carburetor should work well enough with propane to complete a quick test. Or rig up some kind of temporary "spewforth" fuel system for a quick test. Too much propane will produce a black exhaust to let you that the mixture is too rich.

Clark. Coolspring Power Museum.
 
Sorry, didn't realise that (I thought it was just free spinning). Maybe as a quick and dirty test, dump a charge of liquid fuel in the plug hole and try and start it. On a small engine it will initially be flooded but after spinning it over for a bit it should start and run quite well on that single charge for a second or so. If it runs more strongly than usual it's likely you have a fuel issue, if not then it's something else...
 
I think that I may have found at least part of the problem. Due to the design of the intake and exhaust valves there is a very short part of the stroke between exhaust and intake. Both valves are atmospheric operated, that is no mechanical opening, only exhaust pressure and intake vacuum opens the valve. I just found that prior to the actual intake at the carburetor I am getting a short burst of positive pressure at the carburetor throat prior to the actual intake. This may be exhausting the fuel back through the carburetor prior to actual intake. My theory at this point is that the fuel at the jet is exhausted and by the time the actual intake takes place there is no fuel to atomize. I am getting a good suction at the the carburetor but it may be coming too late.

Off to test my new theory. This certainly is a unique and interesting design but that is why I am doing it in the first place. After 20+ "ordinary" engines I am just looking for things that are out of the ordinary. Last couple of projects have included Howell "V" Twin and a Snow.
 
I think that I may have found at least part of the problem. Due to the design of the intake and exhaust valves there is a very short part of the stroke between exhaust and intake. Both valves are atmospheric operated, that is no mechanical opening, only exhaust pressure and intake vacuum opens the valve. I just found that prior to the actual intake at the carburetor I am getting a short burst of positive pressure at the carburetor throat prior to the actual intake. This may be exhausting the fuel back through the carburetor prior to actual intake. My theory at this point is that the fuel at the jet is exhausted and by the time the actual intake takes place there is no fuel to atomize. I am getting a good suction at the the carburetor but it may be coming too late.

Off to test my new theory. This certainly is a unique and interesting design but that is why I am doing it in the first place. After 20+ "ordinary" engines I am just looking for things that are out of the ordinary. Last couple of projects have included Howell "V" Twin and a Snow.

Gordon,
I am also building the Atkinson Differential Engine. I have experienced the same problems you described in trying to get it to run. I have run it at least 150 hours with very little success. It did run for about 7 seconds on its own. I pressurized the cylinder and the rings held 90 psi, until I moved to a different part of the cylinder, then it was leaking. Suspect my cylinder is not good enough. Ordered new cylinder material and will machine it completely, versus doing brazing of components (had problem with warpage from the heat). I found that too lose of an intake spring will cause spitting out the carb. Seems the thing was not getting enough fuel to fire consistently. This engine has been a real challenge. Have you had any success?
Thanks. Ray.
 
Ray: Not having much success. I have lapped the cylinder and have tried remaking the pistons and the rings a couple of times being careful on depth and width of the ring grooves. Just cannot get compression. Like you I can put air in the cylinder and it seems to hold but running it or turning it over by hand does not seem to generate any pressure. It seems to draw well at the carburetor but there is almost no pressure in the firing position. Strange: If I put a pressure gauge in the spark plug hole I get very low pressure turning it over in the right direction but much more pressure if I turn it over backwards. It has become one of these things that "I am going to beat this *&*%$". Not sure where to go next.
 
Ray: Not having much success. I have lapped the cylinder and have tried remaking the pistons and the rings a couple of times being careful on depth and width of the ring grooves. Just cannot get compression. Like you I can put air in the cylinder and it seems to hold but running it or turning it over by hand does not seem to generate any pressure. It seems to draw well at the carburetor but there is almost no pressure in the firing position. Strange: If I put a pressure gauge in the spark plug hole I get very low pressure turning it over in the right direction but much more pressure if I turn it over backwards. It has become one of these things that "I am going to beat this *&*%$". Not sure where to go next.
Gordon,
I have made two sets of rings in trying to get this to run, the rings are sealing at certain points in the cylinder (holding 90 psi).
I too noticed low pressure at the spark plug hole when turning clockwise, but the pressure more than doubled turning counter clockwise. Not sure why. I even ran it backwards with the electric motor (just for the heck of it), but no luck in it running. To develop the power needed, seems the pressure needs to be higher on the compression stroke. Maybe a new cylinder, rings, and pistons will do it (I have to eliminate any leaks).
To get strong pops, I had to squirt heavy oil in the carb, which indicates bad cylinder and ring fit. Since rings seal at certain points in the cylinder, it indicates to me that the rings are good, but the cylinder may have some waviness in it.
Best run I had was with a 1 1/2 inch link on the right and 1 7/16 on the left (tried all kinds of different link configurations and cylinder positions). A choke plate on the carb seem to help some with controlling the air intake.
Hope to get new material in a few days to start on a new cylinder. This time, I am going with air cooled instead of water cooled. My material has an I.D. of 1.125, so no boring operation needed, just honing. The wall thickness is enough to allow machining of the bosses. Will not touch the cylinder with a torch (warpage caused me a lot of problems with the cylinder the first time).
Been a fun and interesting project, though frustrating at times. I will keep in touch as I progress.
Thanks Gordon. Ray
 
Gordon,
I have made two sets of rings in trying to get this to run, the rings are sealing at certain points in the cylinder (holding 90 psi).
I too noticed low pressure at the spark plug hole when turning clockwise, but the pressure more than doubled turning counter clockwise. Not sure why. I even ran it backwards with the electric motor (just for the heck of it), but no luck in it running. To develop the power needed, seems the pressure needs to be higher on the compression stroke. Maybe a new cylinder, rings, and pistons will do it (I have to eliminate any leaks).
To get strong pops, I had to squirt heavy oil in the carb, which indicates bad cylinder and ring fit. Since rings seal at certain points in the cylinder, it indicates to me that the rings are good, but the cylinder may have some waviness in it.
Best run I had was with a 1 1/2 inch link on the right and 1 7/16 on the left (tried all kinds of different link configurations and cylinder positions). A choke plate on the carb seem to help some with controlling the air intake.
Hope to get new material in a few days to start on a new cylinder. This time, I am going with air cooled instead of water cooled. My material has an I.D. of 1.125, so no boring operation needed, just honing. The wall thickness is enough to allow machining of the bosses. Will not touch the cylinder with a torch (warpage caused me a lot of problems with the cylinder the first time).
Been a fun and interesting project, though frustrating at times. I will keep in touch as I progress.
Thanks Gordon. Ray
Correction Gordon, it was 1 1/2 inch link on the left and 1 7/16 on the right.
 
Correction Gordon, it was 1 1/2 inch link on the left and 1 7/16 on the right.
Gordon,
One theory I have is that the steep angle of the piston rod, when the piston is at the end of the cylinder, may be putting a side load on the piston, causing leakage across the rings. When it was holding pressure, the right piston was far into the cylinder. The majority of the leakage is on the right piston in my engine.
Ray
 
How are you checking pressure at various positions? The only place to add air pressure is through the spark plug hole and there is only a short distance near the firing position where you can pressurize the chamber. Once the left piston starts the return stroke the spark plug hole is covered and there is no way to pressurize the chamber.

I also have played around with various lengths of connecting rods and linkage bars. At this time my cylinder is mounted quite low. I have made the engine from flat stock, not castings and the front plate looks like Swiss cheese from all the times I thought that I had the right location for the piston in relation to the spark, intake and exhaust ports only to find that it was not right and I ended up moving it yet again.
 
How are you checking pressure at various positions? The only place to add air pressure is through the spark plug hole and there is only a short distance near the firing position where you can pressurize the chamber. Once the left piston starts the return stroke the spark plug hole is covered and there is no way to pressurize the chamber.

I also have played around with various lengths of connecting rods and linkage bars. At this time my cylinder is mounted quite low. I have made the engine from flat stock, not castings and the front plate looks like Swiss cheese from all the times I thought that I had the right location for the piston in relation to the spark, intake and exhaust ports only to find that it was not right and I ended up moving it yet again.
Gordon,
I pressurized the cylinder at the spark plug port (while holding the flywheel) to test ring seating on the left end of the cylinder. For the right end of the cylinder, I used an air hose with a rubber tipped nozzle at the intake port (while holding my finger over the exhaust holes in the muffler) (note: piston ends must be open at intake port). Saw leakage by the right piston rings to the outside. Good method to check valve seating as well. FYI, you can also use a piece of air hose and put it to your ear to check for leakage while running (poor man's stethoscope).
I raised the cylinder today about 1/16 of an inch to reduce piston rod angle (put slotted holes in cylinder mount plate for adjustability, have some slow water leaks but will fix that later after I get the engine running). I used 1 1/2 link on left and 1 7/16 link on right. Definitely improved the running of the engine with the electric motor, but still did not run on its own power. It did kick over a few times when I shut the electric motor down, also kicked over a few times when cranking by hand. I am going to run it in some more tomorrow in this configuration to see if it improves. I did not run a pressure test today to see if raising the cylinder improved the leakage, but I suspect it did help some by the way it was running. Still waiting on new cylinder material in case I need to make a new one. Ray
 
I have been checking leakage by spraying soap solution around the pistons and turning it over by hand. I get bubbles mostly on the left cylinder. Relying on my ears is not a good test. My ears do not work well in even ideal conditions. Too many years working in noisy shops without hearing protection. Even with hearing aids they are below normal range.

In my research on piston rings it seems like it is necessary to have pressure behind the rings forcing them out. I wonder if at low or no speed the rings do not actually seal properly. I have learned a lot about piston rings recently. Mostly that some folks claim the only way to get them to seal is to make them to very close tolerances and carefully heat treat them and others seem to get satisfactory rings with very crude methods.

I assume that you have mounted the cylinder at the recommended position and then adjusted port clearances with various connecting rod lengths. I have been moving the cylinder around trying to find the magic spot.
 
I have been checking leakage by spraying soap solution around the pistons and turning it over by hand. I get bubbles mostly on the left cylinder. Relying on my ears is not a good test. My ears do not work well in even ideal conditions. Too many years working in noisy shops without hearing protection. Even with hearing aids they are below normal range.

In my research on piston rings it seems like it is necessary to have pressure behind the rings forcing them out. I wonder if at low or no speed the rings do not actually seal properly. I have learned a lot about piston rings recently. Mostly that some folks claim the only way to get them to seal is to make them to very close tolerances and carefully heat treat them and others seem to get satisfactory rings with very crude methods.

I assume that you have mounted the cylinder at the recommended position and then adjusted port clearances with various connecting rod lengths. I have been moving the cylinder around trying to find the magic spot.
Gordon,
My piston rings could be suspect, but they do seal well in certain locations of the cylinder when applying pressure. I can wiggle the flywheel and hear them seat.
Had a good day today, for the most part. With the cylinder raised 1/16 of an inch (used a 9/32 drill bit as a height spacer from the base plate to the bottom of the cylinder attach plate (2x3 inch plate), changing to 1 7/16 links on both sides, and advancing the spark, I am getting consistent and strong pops (best run-in to date). Still not enough to run on its own, but getting close. Had an issue with the set screw coming loose (the threads in the aluminum oscillating arm wore out) on the right oscillating arm to piston rod, allowing the pin to come out and contact the cylinder. No damage though (got lucky). I fixed it by drilling a 1/16 hole through the piston rod ear and pin, then using safety wire to secure it (worked well and I don't ever have to worry about that pin coming out again). Will come up with something more professional looking once I get the engine running.
Tomorrow, I plan on raising the cylinder some more to reduce the piston rod angle (hopefully it will help with piston ring sealing, and gain more power). I will need to slot my cylinder mounting holes some more to accomplish this. I will raise it to the point where the two pistons just about touch using the two 1 7/16 links. Reducing the piston rod angle is really helping.
I have used the same connecting piston rods throughout the project, but used various lengths of link arms to adjust the port clearances, as well as moving the cylinder around. For the rings, I built a fixture to put the rings in and placed them in my foundry for heat treating (heated fixture and rings to barely red hot, then allowed them to air cool). I did a final machining (after heat treating) of the rings using a fixture similar to the one in the book. My rings are made from cast iron. My rings have about 1 to 2 thousandths side wall clearance in the piston groove. I may be too deep on the groove depth (about 0.012" in addition to the ring thickness). I hand filed the gap to just barely go into the cylinder when compressed. The first set of rings I made was a learning curve. The second set of rings I made was used initially, but I felt the gap was too much, so I made a third set I am currently using. The piston rings should be tight enough to seal with hand starting. I have tried varying the electric motor speed to see if it helped anything with the running of the engine, but did not seem to make any difference. Ray
 

Latest posts

Back
Top