The Upside Down Engine

Part Three


After thinking on it for a couple of years, I've decided to opt for a good running engine without the fancy slide valves.


21 April 2016:
Well, after a couple of years and a couple of other projects have been done, I've decided to get back to this one.  As I said, poppet valves are the next modification.  Slide valve engines, as I have subsequently learned, are always very low compression, developing around 40 PSI of compression pressure so as to not cause the slide valves to leak when the engine fires.  This greatly impacts the performance of the engines, makes them a maintenance headache and is a major reason why slide valves were abandoned early-on.

What I have in mind is to remove all of the slide valve components and replace the pressure plate with a cylinder head that has caged poppet valves running off of the eccentrics.  I've come-up with what I think is a workable method of converting the slow motion of the eccentrics to the fast motion of a cam while retaining proper duration and lift.  We'll see how it works.  I will probably have to raise the compression ratio but will hold-off on this until I have the engine running.  I think it will run best at with about 80-100 PSI of compression.  

Here's what I've removed.
I've boxed-up the take-off parts in case I decide to re-visit the slide valves.  I may retain the propane mixer, depending on how I feel about gaseous fuel when I get to that point.
Cleaning-up the scrap for the head.                                                       Roughing-out the head.    
The head is to be made out of steel, 1-1/2" thick, sized to fit in place of the slide valve pressure plate.  It's good that the piece of steel scrap I had on the shelf was about 0.050" thicker than 1-1/2" because it took all of that to clean it up at size.  Even then, on the opposite of the "business" side of the head, it still shows some rust pits.
    The head, sized.                                                                         The head, partially done.
As you can see in the photo above, there are some irregularities in the surface finish of the head blank but, I can't measure them and they occurred because the cutter must have picked up some embedded rust particles.  In any case, they are unimportant to the sealing of the head with the base plate of the engine.  As you can see in the right-hand photo above, I have machined a 0.010" deep circular ring around each of the ports to the combustion chamber.  I will make a couple of gaskets out of 0.018" high-strength gasket material and will cement them in place.  Because of their small surface area, I think they will seal fine even with only the four 1/4-20 studs for hold-down.

Also, as you can see, the ports connect with the pockets for the valve cages.  Next-up is some more CAD work on the cages and then I will start on them.  I leave making the valves and their actuators until after I've thought about it a bit more.
22 April 2016:
The whole day was spent working on the valve cages and I'm not done yet.  There's more to them that meets the eye.  They have to be accurately fit to the head and all the internal dimensions need to be very close.

The valve cages partially done.
23 April 2016:
Some more got done today.  The valve cages are done except for the ports in their sides.  That will wait until I have them mounted in the head.  At that time, I'll machine the ports all the way through the head and the walls of the cages and drill and tap the flange screw holes.  Doing it that way insures that the ports will line-up.

The valves are also done although I haven't lapped them.  I did apply Sharpie marker to the faces and seats and gave them a twirl in their seats.  Initially, they look like they'll seat perfectly.  I await Mister Murphy........

Valve cages and valves.
Just so you know, I made the valves in my usual way.  The heads are gray cast iron.  The stems are 0.125" drill rod.  The heads are drilled with a #31 drill (0.120").  The ends of the stems where they go through the heads are turned just enough for a light press fit in the heads.  Then, the tops are peined over to secure the head to the stem.  The guides (integral with the cages) were drilled #31 and reamed with a 0.125" reamer.  The fit was pretty snug so I used some very fine lapping compound to give a little clearance.  They're still a little stiff but I can always lap them some more if they have a tendency to stick.

The valve springs I'm going to use are out of a couple of empty hand lotion pumps.  They are fairly stiff for the size of the valves so they ought to close fine.

As you can see, the cages are made in two pieces.  The cages are a light slip fit in the head and the mounting flanges are a
0.0015" interference press fit onto the cage.  I also used Loktite for insurance but am not really sure that's going to be sturdy enough.  If they slip, I can always pull 'em out and tack weld them together underneath, where it doesn't show.

I'd try to explain how the valves will work but, if I do, I'll probably botch it and leave you confused  I suppose you'll probably figure it out from the photos as I go along.
24 April 2016:
I got the intake and exhaust ports drilled into the head.  

Drilling intake port.
As I said before, it was done with the appropriate valve cage installed so the ports would line-up.  I used the Dremel to clean up the ports in the cages, feathering edges, etc., to help smooth the flow.

Then, the valve keepers were made and the valves lapped.

Valve keepers, springs and one installed in the cage.
It was a real pain to get the keepers set on the valve stems due to the location of said parts.  After the proper amount of swearing, the job was done.  The cages were installed in the head and the head installed on the engine.

Head on engine.
Note that I kept the studs that were used on the slide valve version in case I want to go back to it.

After getting it all bolted down, I turned the engine expecting compression.  No such luck.  Because the valves seemed to lap in so easily, I thought I would be in luck.  Not so!  Both valves leak so they have to be removed and worked-over again.

I know it looks weird but the valve mechanism will make sense once you see my cunning plan made into hardware.
28 April 2016:
The last few days have been taken up with some fiddly stuff.  The cam followers need to be a very nice slip fit in the cages and this took a lot of time to get right.
     The cam follower parts.                                                           The cam followers in the engine.
The rollers are actually ball bearings out of a defunct hard drive.  The spacers are brass 8-32 washers that have been bored out to 1/4".  The O.D.'s of the spacers had to have 0.030" taken off so the rollers contact the cams instesd of the spacers.  I figure the bearing pins are low-stress components so I made them out of leaded steel.  One end is threaded 1/4-28 and that is how they are secured.

The next thing is the linear cams and the cam guides.  After I get these parts made, you will be able to see how it all works.  After that, all that is left is the pullrods and timing.
29 April 2016:
The cams are done as well as the linkage to the eccentrics.
The cams with the pullrod yokes installed.  One cam with rod and eccentric yoke attached.
I think you will see that I made a slight boo-boo on the cams.  Note the "notch" just as the rise starts.  When I eyeballed the vertical location of the cutter, I did it wrong.  It was too low and I didn't see the goof until after the cut was done.  They both ended-up like that because I machined them while laid together.   It will make no practical difference because the bearing that is used as the follower roller is 1/4" in diameter and it doesn't fall into the "canyon".  

Cams connected to eccentrics, ready for their guides.
I've allowed plenth of adjustment in the pullrods to allow a lot of latitude in cam timing and duration.  Interestingly, a quick check of the timing without changing the slide valve settings on the eccentrics gives something very close to workable.  I'm sure there's some kind of relationship between the slide valves and poppet valves so it's not really surprising it's close.

Tomorrow, I will make the cam guides and their mountings.  I may even get it all together so I can accurately tram the valves.
30 April 2016:
Well, I didn't get as much done today as I thought I would.  The slider blocks for the cams took a lot of time and I only got one done.  There was a lot of fiddling around and I had to rotate the clevises on the cam ends of the pull rods to allow another axis for motion to account for the motion of the eccentrics.  

Cam and cam guide parts.
When I fitted it up to the engine, I found that I had to reverse the guide.  You can see the difference between the photo above and the two below.  When I put it all together with the attendant fitting and filing, I found out that the cam slope is too steep and the cam followers bind.
Exhaust valve closed.                                                                     Exhaust valve open.
Tomorrow, I will stick the cams back into the mill and increase the angle from 20 degrees off the vertical plane of the cam motion to 40 degrees to see if I can get the follower to move without binding.  This will necessarily decrease lift and slow the opening of the valves but, since I originally had the lift set to 0.250", much more than I calculated I'd need, I will have about 0.125" lift which should be adequate.  After all, this isn't a racing engine!
1 May 2016:
The project seems to be back on track.  I changed the "cam" angle from 20 degrees to 45 degrees and the valves now work smoothly.  I then proceeded to machine the parts for the intake valve cam guide and install it.

The new cam angle (bottom).
Both valves installed and working.
I didn't get real fancy with a degree tape on a flywheel but simply adjusted the openings and duration until the exhaust opens at a few degrees before BDC and closes at TDC and the intake opens at TDC and closes at BDC.  I spun the crank and it does have compression, although it's kinda puny due to a dry cylinder.

I then proceeded to make the intake and exhaust flanges.

Intake and exhaust flanges and propane mixer.
If everything goes well tomorrow, I may attempt to smoke-up the shop but I can never tell when a simple thing becomes a rats nest of problems. 
2 May 2016:
Today was a good dary here at Hoyt-Clagwell & Company.  First off, I got the intake and exhaust flanges finished, got the exhaust pipe done and made the carburetor adapter for between the throttle body of the carburetor and the flange.
Carburetor parts ready for assembly.                                                       On the engine, governor hooked-up.
Once the exhaust and intake plumbing was done, I figured out how to connect the governor and decided to see if it would run.  Hooked up the propane and ignition, oiled everything up and cranked.  And cranked.  And cranked.  Checked the ignition timing and it was right on TDC.  Cranked some more and worked up a sweat.  My only reward was a few feeble pops out of the carburetor.

Thinking that the intake valve needed a little more lapping, I took the intake cam and guide apart, removed the cage and re-lapped the valve.  When I put it back together, I found that the cam was binding a little so I made the appropriate adjustment.  Cranked some more with the same result.  In addition, suddenly the piston blowby got really bad.  I stuck my finger down in the cylinder and found that the oil was really dirty and thin so I wiped it out and oiled the heck out of it.  A little more cranking and the blowby suddenly stopped.  I think a ring stuck and cleaning out and adding more oil unstuck it.

I cranked some more and got the same result, just weak pops out of the carburetor.

I sat back and "thunk" on it some.  The "AA-HAH" moment finally arrived and another timing check showed that the valves were trammed correctly and the ignition oas set to TDC but the ignition was at TDC on the overlap stroke!  That wasn't hard to fix.  I just loosened the setscrews on the eccentrics (one at a time) and, while holding the eccentric, turned the engine exactly one turn then re-tightened the setscrews.  A slight tweak of the valve timing and it was time to try again.

This time, my cranking was rewarded with smoke!  A little tweak of timing and mixture and it settled down and ran fine.  Had I been in a Victorian frame of mind, I would have stood, raised my hand with forefinger extended and shouted, EUREKA! or EXCELSIOR! but, being the staid individual I am, I just grunted a happy Humpf!

There are a few tweaks I'd like to do but am not in a hurry.  One thing that would help would be to increase the compression.  With the slide valves, the engine had to have compression below 50PSI to keep the valves from leaking.  Increasing the compression would involve either making a new piston with a longer compression distance or making a longer connecting rod.  I sort of tend toward making a longer rod but that's not a priority right now.

Also, I should modify the ignition timer so I can get more advance.  Right now, with it advanced to the stop, it's only firing a little before TDC.  A few more degrees of lead would help but, right now, I ain't complainin'.

 In the next few days, I will take it outside and let it run for a few hours to finish seating the rings.  With the slide valves, it would only run until it got warm, then the valves would either bind-up or leak so I never got over about 20 minutes of running before it would quit.  With the poppet valves, that shouldn't be a problem.

I may also make a YouTube video of it running so the valve action can be better understood.
5 May 2016:
Yesterday, I ran the engine for almost two hours with no issues aside from a pesky occasional misfire.  That will get sorted out in due time.

Chugging right along.
Here's the latest video of the engine running.
Now, I can play with it for a while.  Mixture adjustments will correct the misfire for a while but it still misfires about twice a minute.  I think it could be made worse by the governor, which is hunting a bit.  At the present time, I've got it set to the most sensitive setting on the throttle arm so it hunts slightly to vary the load which should help seat the piston rings.  I think the throttle changes are upsetting the mixture.

25 September 2017:
It's been a while since I've done anything with this engine so, I thought I'd do a little upgrading.

First of all, I bought plans for a demand regulator to replace the one I designed and which doesn't completely turn off the gas when the engine stops.  After building the regulator and putting a fuel mixture adjustment on it, the engine ran better but still had a couple of things that bothered me.

One of the "botherations" was the fact that it still had the very low compression that was necessitated by the now replaced slide valves.  At 3.8:1, the engine liked to run but didn't make any appreciable power.  I removed the piston and made a 1/8" plate that bolted to the head of the piston.  This raises the compression ratio to somewhere in the region of 6:1.  It's taken about four hours of running to get the rings close to seating again.  I know, I should have just bought new rings but what can I say?  I'm cheap.

Then, when the compression came up as the rings seated, the engine became easier to start but the governor showed it's shortcomings.  Beside not being very sensitive, it developed a couple of tight spots that wreaked havoc on engine speed regulation.

I've come up with a modification to the governor, using some of the old parts.  Instead of the four weights with the short links to the governor rod, I now have a 45 degree wedge that contacts similar wedges on two weights.  This arrangement will require a tension spring on the governor rocker to allow the throttle to close.
Parts of the revised governor.
I am going with only two weights because the governor has much less friction to overcome in the throttle linkage than in the slide valve version.  I think this simpler governor will work better than the original arrangement.  If it's not sensitive enough, I can always add two more weights.  I'll finish it up in the next couple of days and see if it's any better.
27 September 2017:
The linkage betweeen the governor rod and the throttle arm is done.

Link in place.
The engine runs well and the governor is much more responsive.  Now, the Law of Unintended Consequences comes into play.  With the fuel mixture right, the throttle butterfly doesn't fit it's bore well enough for the engine to idle down unless the spark is retarded.  Maybe tomorrow, the intake will come off again and I will spend some quality time with it.
29 september 2017:
Yesterday, while I was trying out different springs on the governor, I got a combination that caused the engine to run away.  I got it stopped before it really took off but it went into "shaky" mode.  Once it stopped, there was a big leak in the cooling tank. A piece had rusted out.

Minor problem with cooling tank.
(after cleanup)
The cooling tank was made of a couple of big coffee cans soldered together.  Even using rust preventative antifreeze, it only took about three years for it to die.  I thought of making another tank but, since I was disgusted with it, I quit for the day.
      Radiator installed, front view.                                                           Radiator installed, rear view.                                                      Cooling check while engine is running.
Overnight, I remembered that I had originally made a radiator for Engine Number Four.  After I re-powered The Algore Edition Hybrid Hoyt Clagwell with the engine, I used the larger radiator on the tractor and hung the smaller radiator up.  Shazam!  The new cooling system for The Upside Down Engine!

It was easy to mount, just had to make a couple of standoffs go get it to the correct height.  I did consider mounting it the other way 'round and belting the little cooling fan to the flywheel rim but decided to see if it would cool without it.

I ran the engine at between 500 and 600 RPM with the timing retarded a bit and the temperature didn't rise above about 185 degrees.  After advancing the timing back to where it was, the temperature slowly fell to 176 degrees F (80 C) before quittin' time, so there's no big need to add the fan.

It will run hotter outside of my air conditioned shop so, tomorrow, I will roll it out into the driveway and let it run to see how warm it will get.  Even if it boils a little, there's no harm as long as the water doesn't boil away.

Just for the heck of it, since the copper radiator is insulated from the steel/iron engine, With about 140 degree tap water as the coolant, I measured the voltage between the engine and the radiator.  The radiator is about 0.22 volts positive.  Now, the question is whether I should connect them together or leave them apart.  My first thought is to leave them insulated, otherwise, the two will act like a battery with the poles shorted together.
8 October 2017:
After running the engine for a while, I got to where I really didn't like the stink of the blowby.  When I machined the liner, the bore turned out to be around 1.556".  There was no really close Imperial size ring available but a metric 1.5mm X 39mm chainsaw ring was close (only a little smaller than the bore) and I used a couple of them.  They didn't seem to be cutting the mustard, I got in touch with Dave Reed ([email protected]) who fixed me up with three 3/32" X 1-9/16" cast iron rings that are on the way here.  In the meantime, I worked on the piston.

First, I re-visited the compression increasing disk.  What I originally used was a 1.5" diameter 0.125" thick drop-off from a waterjet cutting operation.  It was a little smaller than the piston diameter and I figured that, since it did improve performance and I had the piston out anyway, I would replace it with another drop-off.  This one is 0.164" thick which will increase the compression a little more.

New compression increaser spacer.
After the spacer was bolted and Loktited to the piston, it was turned to the piston diameter.  The smaller diameter area around the rings was made wider to accomodate a third ring.
Feeler gauge pack for 0.094" width of ring grooves.
The new rings may be in tomorrow.  I am leaving the piston in the lathe until I can verify that the rings fit.
10 October 2017:
The rings came in today so I immediately went out to the shop and ascertained that they fit the new grooves okay.  Then I gapped them to around 0.008".  Then, I put them on the piston, hung the rod and promptly broke one of them while inserting the piston in the bore.  DRAT!   It hung-up on the edge of the chamfer I machined on the "bottom" of the bore and snapped.  I wasn't even gronking on it when it broke.  Gee!  Those things break easily.

Dadburn it!
Anyway, I simply put it together with two rings (the same number that were originally on the piston) and got it running.  I've only got about an hour on it but it does run differently and seems to make more power, as measured by my shop towel prony brake.

Not to bother Dave for one more ring, I will wait until I have an order for more.  Until that time, the engine stays as-is.  Maybe I'll forget about the missing ring and be happy with it.
26 January 2018:
I've been tweaking on the engine off and on for a while.  To get the governor to quit hunting, I've made an idle throttle stop which helps somewhat.
Throttle idle stop.
Also, I've ordered new piston rings to get rid of the annoying "click" sound of the blowby.  I've run it for about four hours total at differing speeds and loads and there is still some leakage so I will add the third ring and hope Mister Klutz doesn't strike again.      
BOY, This is fun!

Comments? [email protected]