[Rick Parkington]

Blimey, that'a a bit odd! I used to look after a 1937 Model 14 which I guess was similar. Few years ago now but luckily I did make some notes about the lubrication which seemed to cause a fair amount of grief, the sump tending to overfill and cause smoking. If the slotted cap you mention is the one on the front of the timing cover then I think that is the ball valve John mentions. It seems to be there as an anti wet sumping valve like the telltale on the earlier models - but in both cases, being between pump and big end, it cannot prevent oil passing from the feed to the return side of the pump via the driving shaft and heading backwards down the scavenge to fill the sump. I ran loads of tests on this bike which I quite enjoyed reading again - but I haven't recorded whether I ever completely solved it and now I can't remember.  If you are the current owner of RG8149 and it smokes I can only say I tried!
As with the larger models the ball valve spring is lifted by pump pressure and opens the path to the big end and rockers.
According to my notes the ball opens to open a drilling leading to the crankshaft quill bush housing in the timing cover via a gauze filter. In the bush a 1/ 16" hole diverts oil into the quill bush to feed the big end while the rest continues up a vertical drilling to the rocker feed pipe. This system seems different to the earlier models but in both cases the Sunbeam rocker feed system is a bit tricky because the roller big end offers little resistance to oil flow, so it's easier for oil to head straight for the big end than to fight gravity all the way up to the rockers. On the earlier bikes (1932 etc with the tank gauge) it appears that the rocker feed opens before the big end and it's essential that there is enough back pressure (as shown on the gauge) to ensure the telltale is forced to open further and open the crank supply.
If my notes are correct it looks like the 250 relies on the small crank feed hole to create the back pressure to feed the rockers. My notes also say that it is not a pressure system injected into the rocker shafts (which of course causes quite a bit of back pressure) but sprays oil into the rocker box where it runs into drillings, feeding the rockers before draining back down the pushrod tubes. There was an adjustable needle valve on our 250 rocker feed which was completely knackered but I had to fix and re-machine it because otherwise the rocker oiling was excessive.
     None of this is much help with your query but I thought it might be worth recording for posterity!
I wonder of the key on the pump gearshaft is missing or broken so that although it works at low speed the increased back pressure at high revs causes it to slip? Seems a bit of a feeble suggestion but it's hard to see why else it would be a problem. Feed is usually more reliable than return owing to the assistance of gravity; the usual problem that afflicts it being crankcase pressure blowing oil back up the pipe but that was more of a problem on 1920s total loss bikes where the pipe went straight into the case and supply was just a drip feed.
The pump is contained in the case under that little cover with 4 screws. The cover is rebated snugly into the cover so it's a bit of a faff to get out, but I have written in the notes that the outside gear pair are the feed ones so it should be fairly easy to gently lock the gears and turn the motor to see if the slip or not.
Sorry to waffle on but hope this is some help!
Cheers Rick 

[singleminded]

i just had a look at the parts list for 1938..There is a non-return ball valve listed, if this is in the supply pipe could it be blocking when the revs rise?
would it be worth trying to by-pass it to test if it is that before stripping the engine?  John

[PaulW]

I bought this running bike last week and, as I always do with a new bike and before riding it, I checked it out on my bike lift: tyres, spokes, girder fork spindles, wheel nuts, mag timing & point gap, valve clearances, slow running adjustment etc etc...and an oil change which involved cleaning all three gauze filters (oil tank, sump plug and pressure relief valve) and removing the tank to rinse it out.

Have re-filled the oil tank with straight 40 monograde and primed the pipe from tank-to-crankcase, I started the bike and everything seemed fine.  Nice tickover and oil pressure gauge reading of about 30psi and a healthy oil return to the tank.  Jumped on the bike and had done about 500 meters before the gauge dropped to zero and oil return to the tank stopped.

I coasted back home and bled the oil pump of air by undoing the slotted cap on the side of the timing chest cover while the engine was running ... once oil poured out, I replace the cap and tried again...but same thing happens again.

After the second time. I took off all the pipework from tank-to-crankcase and checked that it would hold a vacuum.  It did, so I'm confident that the pump is not pulling air anywhere between tank and crankcase.  There were also no obstructions etc.

I've now narrowed the problem down to revs.  If I bleed the pump of air, then the bike will tickover with a healthy oil pressure and a strong return to the pump.  However if I rev the bike for any length of time (i.e. 5-10 seconds), then the gauge will fall back to zero and there is no oil return to the tank.  There is no recovery from this; going back to tickover does not result in a return of pressure.  I would have to bleed the pump again.

So I'm at a loss....and as I bought the bike from a dealer, I'm also somewhat disappointed.  I have no idea if it was like this when I bought it on Tuesday because I didn't ride it before doing the oil change.

Anybody got any ideas?   I have not yet removed the the timing cover to look at the oil pump (this is a '38 model with the internal oil pump).

Many thanks for any suggestions you might have.

Paul