We all seem to be agreed that the reason that the engine will not run unaided is that it is not firing consistently. Once this is sorted, we can worry about reducing the friction if this still causes problems.
And the most likely culprit is the ball valve. As you suggested John, I started off with a square edge and tapping the ball. This is the normal approach. But a couple of times when I sucked/blew into the tube to check the seal, I found the ball had stuck. This could be because I used a 4mm diameter hole and the ball sat too far down and wedged.
But relieving the edge with a small chamfer did not make any difference.
I did try a light spring but this did not help. Perhaps I did not pay enough attention and this might be worth another try.
I also tried an acetal ball with is lighter in case there was not enough “suck” to lift the ball high enough. This was when I was still wondering if the engine was getting enough mixture to make a big enough bang. No difference.
You and John Woods both used Jan’s design without success. So how come this design works for Jan and not for other people?
Today, I will build one to Jan’s design but with a screw adjustment through the top to vary the maximum ball height. I don’t think my machining skills will permit me to get it right otherwise.
My other thought is that since this ball valve is doing the same job as the automatic inlet valve on my hit & miss engine (Bonza), I could use a spring loaded poppet valve. I am still playing with arrangements for this to try to get one that looks a bit elegant. But I might have to forget elegant and just make one. We shall see.
El Illegitima, nil grindem carborundum!!
Jim