Plane: 

Waiex, you've seen me bloviating about it in the Open Forum.

Engine: 

OS Max 25S -- at least that's what folks seem to call it.  OS calls it the "Max 25" without suffix.  It came to me as an RC engine, I made a venturi with a diameter of 15/64 and a 1/8" spraybar, which works out to 0.0138 square inches, compared to the theoretical best 0.0113 squares for a 25.  I have tried two glow plugs on it -- the one that it came with, and another, short reach one, that I had already.  But they're both RC plugs from 'le box du junque', so may be wildly inappropriate for this engine.

Tank:

An old Sig 2 ounce clunk tank.  Plumbed with a vent to the muffler, a clunk to the needle valve, and a fill tube that gets plugged for flight.

Problem:

I can run out the second half of the tank just fine.  Or I can run out the first half of the tank.  But in between I have to open the needle valve a whole bunch -- one or 1-1/2 turns, up from about 3-1/2 for a full tank.

Suggestions? 

I'd assume that I just did something wrong with the venturi or the spraybar, except the change is so sudden and marked -- each time I started with a full tank it stopped at the same place in the run (ditto when I started with half a tank, but one normally expects the engine to stop when the tank runs dry...).

I'd really like to not have made the venturi too big, so if there's anything I can try short of slapping the other venturi I made on there (I do have a smaller one to try) I'd like to hear about it.  For that matter, if there's a good way of ground testing the set up (do I just need to hold the stupid thing sideways and breath castor?) I'd like to hear that, too.

Next flying day is Saturday, if I don't have a solution by then I'd at least like a list of tests to run.

If that vent that shows in the top of the tank is a uniflow, it's in the wrong position.

It wasn't meant to be -- it's just a simple fill tube.  At that point I was just trying to get things done early enough last night that I could sleep, get up, get some paying work done, then go fly.

D'you think that converting it to uniflow would help?  If so, can this be done with rigid tube, and where do I want it to open inside the tank?  It's probably a good idea, as the venturi shows what seems to me to be just the right amount of sensitivity to maneuvers -- when I can get it going right on launch, it shows a very satisfying power increase on the up-lines.

Sounds pretty much like the ol' profile on muffler pressure plus vibration fuel foaming and air bubbles prob. Use clear line to the needle and see how many air bubbles are feeding in. A few are okay but a steady stream needs constant needle fiddling depending on how much fuel is left. For a profile clunk, the thing that works for me always is a dubro sintered filter pickup, uniflow open ran to the outboard side of the tank about 3/8" from the pickup even or slightly above the spraybar with no pressure and the vent blocked. The exit end of the uniflow tube should point upwards and end above the tank to prevent siphoning. Also a little foam rubber between the tank and fuse may help.  Good luck and let us know what you figure out.  

Tim,
     If you think your venturi is too big, you can use pantyhose.  Cut patches of pantyhose, and hold in place with an O ring.  The more layers over the venturi, the smaller the venturi acts.  Hope this helps!

-Grandpajake

Tim,

Have you replaced the clunk pickup fuel line inside your tank? Sounds like you may have an air leak in that line. As long as it is below the fuel level the engine runs fine, but as soon as it's exposed to air, it sucks bubbles and the engine goes lean or dies.

I like silicone fuel line, but it has a nasty habit of fracturing if scratched (much like glass). Always chamfer and deburr all the ends of brass fuel tubing, that gets silicone fuel line pushed on. I've seen fuel line with radial fractures right at the sharp ends of brass tubing, where it is almost impossible to see without removing the fuel tubing.

Brass tubing is easy to chamfer on the inner edge with an Exacto #11 blade, or a needle file. The outer edge can be sanded smooth with a fine grit foam sanding block, before installing the brass tubes in your tank stopper. Run your finger tip over the ends of the brass tubes, to be sure there is absolutely no sharp edges.

Hope this helps.

Bill

Update after some field and driveway experience:

Going to a uniflow didn't help -- at least the way I did the uniflow.  I dunno why.

I threw up my hands, took the engine out of the plane, and played with it on the test stand.  Different tank, different situation, same engine.  Couldn't crank it up.  It's making funny clunking sounds and golly, it really is down on compression.  Not good.  Grabbed my other OS 25S (goody! a spare engine to screw up!).  Ripped off the RC carb, slapped on the venturi, gave it a flip -- oh, the old one really was wore out!  That musta been the problem!  Filled the tank on the test stand, started 'er up, adjusted to sound good -- she died after half a tank, gently smoking.

@$%#

And furthermore, &*&^, ^&%$ and @$#%.

My test stand tank is set up with just an overflow and a feed.  So after I fill it, I can turn it over, and the overflow becomes a uniflow vent.  I tried that.  Fill, flip tank, start it up, adjust it to run on the ragged edge of a four cycle -- and it ran the tank out, as consistent as you please.  The only downside is that it only took four minutes, and that's from the time the motor started, not from the time that I would have launched the plane.

So I think I have several problems that I need to overcome:

One: I need to rework the tank on the plane.  I tried to put the uniflow toward the inside of the circle from the clunk, I think instead that I want it ahead of the clunk with the clunk "reaching over" to get fuel -- at least that's how the test stand tank is, and it works dandy!

Two: I need to start things a lot more rich.  I've been starting the thing on the edge of a four-cycle, and it would lean out to what I thought was a rich two-cycle but which in retrospect must have been too darn lean.  This calls for some experimentation.

Three: When I get it all running, I need to make a bigger tank, probably around 3oz.  Maybe I'll just do that first...

Four: possibly, I need to use a smaller venturi.  I'm hoping not, as that'll mean that Brett was right and an OS 25S isn't enough engine for the plane.  But I can always try the 20FP I have, if I can't find a good compromise between engine run and engine power.

Stay tuned, and thanks for the help so far.

Take it to next weekend's Salem contest.  A dozen guys will fiddle with it until it works.

I'm sure hoping that I can make it.  I'm also hoping that I'll have the engine run sorted out before I go.  But at this point I'm really only 100% up to takeoff, wingover, level laps and landing -- that should fit into 20 laps, shouldn't it?

Just an observation and my 2 cents. In the photos you posted it appears your clunk pickup does not reach the back of the tank. I think this could be a problem, though not enough to stop half way through. On my 25 fp setup the uniflow line ends just ahead of the clunk, i.e. it's soldered to it. I should also say I run in a 2 cycle (10,000+rpm) with a 9 x 4 apc. 2 ozs should be enough for 5 min. in my experience. Hope this is helpful.

Bob

Update #2:

Joy!

Put the tank & engine on the airplane, got it running, held it with the outside wing down -- and it made it through the whole tank!  It sure seemed to take more than six minutes, but time does drag when all you have to do is watch some fuel tubing for bubbles.  That's the first time that the engine has run through an entire tank on that airframe, so whatever the problem was I think I fixed it.

So there's hope for the engine setup.  Tomorrow I'll fly, and find out if it does as well in the air as it does in testing.  Hopefully I won't get too ambitious and crash before Sunday.