Spits and sputter but no substained run

[tornitore45]

I completed my firs 4 stroke twin. Is an Upshur twin as presented in the Model Engine Builder Magazine Issue 7.

It doesw at most 3 - 4 pops on one cylinder only.
Both cylinder have low static compression but at speed I could fel a strong suction on before installing the cylinder heads. The ring gap is a bit too wide.

The ignition is operated as a "Wasted Spark" system, the spark is good.

I am using a Coleman Campstove Fuel that was bought about two years ago when I started the engine with the illusion to complete it much sooner.

The engine is designed for Gasoline. When running on Coleman the best carburator setting is at full open trottle and 3/4 turn open needle.

I have checked for leaks with soapy water around the head.

The finger detects suction at the intake ports an pressure at the exhaust.

I have a hunch that either the Coleman is too old or the carburator is too narrow for it. Possibly Gasoline will work better but I like to hear from those that have been down this road before.

 

[stevehuckss396]

If it's designed for gas maybe you should get it running on gas. After you are sure the rings are seated and the valves seated you could try to run on Coleman. Sounds like a slight modification may be needed to run Coleman but I wouldn't try it until I know what it does on gas. It may do the same on gas and then you will need to look elsewhere for the problem.

Think what I would try is try to run on gas. If that is a bust I would check ignition and cam timing.

From what you describe it does sound like it wants more fuel.

 

[Brian Rupnow]

There will be no appreciable difference in the engine settings using automotive gasoline or Coleman fuel (Naptha). The Coleman fuel might be too old, but I doubt it. I have had people tell me that small engines develop more power on automotive gas, but I haven't seen that. All I have really noticed is that automotive gasoline ran in a small engine is much stinkier than Coleman fuel. First, check your valve timing, and make sure you are rotating the engine in the correct direction when you do it. In fact, make sure you are rotating the engine in the correct direction when trying to start it!!! Next, check your ignition timing. For first start up, a good "starting point" is to have the engine fire exactly at top dead center. There are minor benefits to be gained from advancing the spark (a bit more power, but will run hotter) and some minor advantages to retarding the spark--(starts easier, runs cooler, but less power.)---but for first start ups, I recommend firing at top dead center. If you are running "atmospheric" intake valves, make sure the springs aren't too strong. Loss of compression thru poorly sealing valves has always been my main bugaboo on small i.c. engines, and sometimes it can be very difficult to diagnose. I have learned a new trick, which seems to work very well on i.c. engine start ups. If you are SURE about the valve and ignition timing, and you have visible spark at the plugs when they are pulled out and laying on the engine block, then do the following. Put a pulley on the crankshaft and hook the engine up to an electric motor with a v-belt. Turn on the gas and spark, plug in the electric motor, and let it run--The engine speed should be at about 350 to 500 RPM. If its a 1750 RPM motor, you should have approx. a 4:1 reduction thru the v-belt pulleys. Once the engine has ran and fired a few times, the compression created in the engine by ignition firing will completely seal the valves (assuming they had a "reasonably" good seal to begin with.) The "running in" will also help to seat the rings and help them seal compression in the cylinder as well. Of course, make sure that everything is well lubricated, and run about a 20:1 mix of 2 cycle oil with the gasoline. If you have lived a good clean life (as I have) the engine should begin to fire regularly and may even try to run the electric motor.--(It won't be able to because of the inverse pulley ratio). Once it is firing regularly, take off the v-belt and start the engine on its own. I was given this piece of wisdom by someone who has been building small i.c. engines for many years, and it truly does work, as I have proven to myself with my most recent engine build.---Brian

 

[Rustkolector]

I usually overlook something in my rush to hear that first pop. When nothing happens, I then settle down and go over the basics. If you know you have compression problems you may as well fix them now. I know you didn’t want to hear that. Compression will improve after the engine runs for a few hours, but only if parts were made correctly initially and the valves seal well. Do your valves seal well? I recommend using a vacuum gauge to check valve seats. Do the basic check with valve timing and ignition timing. Set ignition at 0-5 degrees BTDC to start with. Open the plug gap to .020”+. Ignition is a very common and often overlooked problem. Make sure you have adequate valve lash. Is your exhaust valve spring strong enough to keep from getting sucked open on the intake stroke? I have never had an engine start easily at wide open throttle. This is because the air velocity through the venturi is rarely high enough to give a good vacuum signal at the spray bar when the throttle is wide open. All my engines start easily with 10-20% or less open throttle and run with much less throttle opening at medium speed. Make sure the tank fuel level about 0.5” blow the carb spray bar. Prime you engine with a spritz of fuel (Coleman fuel should be fine), then close the throttle to 10% or less opening and try starting.


At least the engine is making noise. That’s a good start.
Jeff

 

[Brian Rupnow]

Rustkolector--Very good point about checking the valve lash. I see that you and I differ on where to set the ignition for start up on a new engine. I wonder what other engine builders do?---Brian

 

[steam58]

i had use coleman fuel that was over 10 year old with no problem in my single cyl. upshur. I did have problem with the rings.Then i switch it to an o-ring and i still had problem and ignition at first had home made points. Then i switch it to a s&s cdi system and it start to run right

 

[stevehuckss396]

Rustkolector--Very good point about checking the valve lash. I see that you and I differ on where to set the ignition for start up on a new engine. I wonder what other engine builders do?---Brian

I build multi-cylinder engines so this might not apply to the hit-miss world but I started tuning my V4 and V8 both at 25 Degrees BTDC. The V4 loves advance and is now at 38 degrees. The V8 not so much and is now at 32 BTDC. There is no difference at low RPM but you can really tell when trying to get the engine reved up. Crisper when lots of advance. They still idle well under 1000.

 

[tornitore45]

All good points. One thing ruled out is the coleman age.

This is my first IC engine.
Both valves are lifted, no atmospheric intake.
Valve seal. Not knowing any better I did the following:
Insert valve in seat
Plug valve passage with a bit of paper napkin bto prevent grit from reaching the stem-in-guide area.
Press a semi flexible tubing on the stem
Dab a slurry of 800 carbide grit and oil on the seat.
Swivel back and fort by hand and then drive the tubing with a drill at slow speed.
Inspect and interval and notice a dull ring developping around the valve and the aluminum seat, no shiny surface, the dull ring on the valve head is about as wide as the seat bevel around 0.030".

Tested the valve by blowing from the port but the pressure was enough to lift against the spring. Tried with lung pressure, one was bubbling the soap and needed a bit more lapping. As far as I can tell they do not leak.

Run the engine with drill, without push rods, no presure of vacuum felt ot the ports.

The valve gap is adjusted by feeling the pushrod rattle a bit when the cam lobe is away.

I could not follow the instruction on valve phasing so by trial I have them set to "look" right. I rotate the flywheel and see the the exaust opening going up, a bit of overlap with the intake going down, intake close going back up. The few pops suggest they are rougly right.

At this point my greatest concern is compression because one cylinder is poor and the other does not fire. After that it seems like is starving but the question is whether is starving for air volume or mixture.

Thanks to all, it cleared my mind to try a few things.

 

[rklopp]

You may not be the only one experiencing difficulty getting an Upshur twin to run. Go to Florida Association of Model Engineers' website and search on "Upshur Twin". I have six Upshur twins about 80% complete, but progress has stalled while I have fun building other engines.

 

[tornitore45]

RKLOPP I read the FAoME entries and is not encouraging.
Thanks for the pointer.

 

[Rustkolector]

Mauro,
I still recommend you tackle the low compression issue before looking for other problems. It is very important to an easy starting and good running engine. I read the FAME site posts on the Upshur Twin as rklopp suggested because I was very interested in building this engine a while back. It does appear there may be a design issue. What is most interesting to me is that this engine is known to run very well, but for only for 45 seconds or so, before stopping. Even Hamilton Upshur’s engine did that according to Dale D. With the center mounted carb, and long intake runner, it could be loss of manifold heat due to fuel vaporization. Jerry Howell had this problem with his “Bill” engine which has about a 2.5” long intake manifold. The engine would run a short time and then quit when using liquid fuels. He solved the problem with his Bill engine by switching to propane. It runs fine on propane, but wouldn’t run long on any liquid fuel. I built this engine and it did just as Jerry said it would. It only likes propane. I believe he might have designed his V-twin with two carbs expecting the same problem. Just a thought for future reference.
Jeff

 

[tornitore45]

Iraised the tank, the small amount of fuel was about 1" below the spraybar.
Instead of 3-4 pops I got a dozen or so at much restricted throttle as is suposed to start still from one cylinder.

I removed the cylinder assembly leaving the head mounted, built a rubber plug connected to my compressor and pressurized with the head under water.

A very gentle bubble stream form at the valve ports but not the head or spark plug. Funny the non firing cylinder appears to have lower leak but is the one with the slightly longer intake. The intake is a different design: bent copper pipes flanged at the carb, need to make sure no vacuum leaks at the head entrance. I expect a minimal advantage versus the evaporating cooling since the copper pipe can conduce more heat from the ambient. I wander if a vapor carburator a la Jan Ridders may help.

The trottle-needle combination is extremely touchi.

 

[canadianhorsepower]

The trottle-needle combination is extremely touchi.

do you have the plans or drawing of your carb??
Ide like to see them

 

[tornitore45]

I will scan the carb details and see if I can figure how to post pictures.

 

[tornitore45]

Scanning and posting would violate the copyright of Model Engine Builder.

I rather describe it, the twin's carb is sligtly different from the single.

Here is the twin

Imagine a Tee pice with a 1/4" hole in the horizontal feeding the 2 cilinders.

Entering that hole perpendicularly is a 3/18" hole 1" long this is the Venturi part
The open part opens conically at a 24 degrees included angle merging with the 3/18" hole at a depth of 0.293".
A bit of trig gives the ventury opening to 5/16"

1/2" from the Venturi opening there is a radial barrel (trottle) with a 3/16" hole in line and completing the initial 1" long hole.
When the barrel rotate 90 degree it completely closes the Venturi.

The sprayer has a 0.070" hole and reach the Venturi bore center, a conic needle as slender as I could make it with the lathe plug the spray hole.
Do not temember the needle angle.

If a picture is worth a 1000 words I think I just found a compression algorithm.

 

[mu38&Bg#]

Most certainly a scan of the top assembly of the carb would fall under fair use. Scanning and posting all the detailed parts to produce it would be a problem. I don't have MEB issue 7, but in 8 I saw a top assembly cross section and it's a typical rotary barrel carb. How well this particular design operates throughout the throttle range I don't know.

The "touchiness" of the needle adjustment is directly related to the taper. A finer taper will result in less sensitivity. There is also a bit of an issue with the needle not centering in the metering orifice, but it's hard to say if that might be a problem here. From my experiences with glow RC engines on gasoline the optimum needle setting can be a 5-10° range on the needle. But they will usually start and run quite rich if everything is in order.

Compression is important. Low or leaky compression results in lower temperatures in the cylinder. The means less of the fuel evaporates into a combustible mixture. Add poor atomization in general and this could lead to poor starting.

Greg

 

[Lawijt]

Or try to run the engine with modelengine fuell. This is Methanol with synthetic & castor oil. Also it have some nitro in it. The Nitro give a amount of oxygene with the combustion.
So you can give the engine more fuell. Also the octane amount is very high.
Try it & let us know.

Barry

 

[Art K]

Mauro,
Referring to post 12 and the bubbles coming from the valves. I built Upshur's vertical single and was concerned with the guides pushing out so I put a step on the valve guide so it couldn't push out. They were not very true, I pulled the valve out marked the head with a sharpie and put the valve back in and spun it with my fingers. I then tapped with a plastic handled screwdriver away from the shiny spot. I kept going in this manner till I got a shiny line all the way round, it took an hour and a half. Once I got that situated the valves never leaked. You may not be able to adjust the guides this way but you can use it to find out where the valves are leaking.

 

[tornitore45]

All good pointers to try out as soos as we stop setting new high temperature records in TX.

Thanks all

 

[tornitore45]

I am elated to report that yesterday I got it to run very well, with plenty of power at full throttle and smooth idle, albeit very sensitive to throttle.

Problem 1) I installed the Rimfire spark plugs just as it came gapped without a second thought. After countless fiddling with the mixture and pulling the cord it dawned on me that the gap may be too small to provide the required energy.
After raising the gap to 0.040" the sputtering time increased.
Problem 2) After much head scratching I notice in the corner of my vision a sparking on the PC board.
Explanation: As the engine revved up the dynamic compression increased (less time to leak past valves and rings), that raised the spark voltage to the point a second path became preferable.

After improving the creepage distance on all secondary circuit the engine really sped up to sustained running.
Then it blew out the two 90* pipes on the exhaust, they were simple elbows about 1.25" per side friction set into the head. As a result speed increased substantially. Did not realized how much breathing obstruction they presented.

While on one side I was happy that it finally worked, I was humbled that a guy that spent 50 year in power electronic could do such a shoddy work on the electronic ignition assembly.

Here is a link to the movie, you can FF to 1/2 to skip the failed starts due to a previous stopping by chocking, which kind of flooded it.
https://drive.google.com/file/d/0BzuYizhKL8gacGJJblRsS2lDVGM/edit

You will notice occasional sparks jumping from the right plug to the push rod, using the booths fixed the problem.

I hope this gives other builders the motivation to finish their Upshur.

Off to the next engine we go, but need to build a wood display base box to hide the electronic ignition.