The new pump might help a bit (if your old one is significantly worn), but the root of our problem is that the non-return valve (a check valve) buried in the RH crankcase half, at the end of the oil pump to timing bush oil gallery (drilling) ain't doing it's job! And unfortunately, there is not much we can do with it as is.
There is also a leakage path across the bearings in the pump, and out through the drive shaft (behind the drive adaptor). Recently SRM have added a shaft seal to their pumps (revised design), and I have also machined a couple of the original BSA pumps and added a seal. I don't believe this is a significant issue in causing wet sumping, anyway. If you buy a new SRM pump, it should have a shaft seal (nice).
If the integrity of the check valve was good (no damage to the seating face in the aluminium case, or to the hardened ball, or any shit between the ball and face), then leakage across the gear pump can't get past it. Unfortunately the long and extremely light spring often doesn't hold the ball square against the seat, and I suspect the spring is too light to hold the ball closed against the weight of the oil. therefore, the check valve doesn't (or unreliably) work, and the rate of wet sumping is relative to the clearances across the pump, and the viscosity (thickness) of the the oil. Sometimes you will shut the engine down, and it might seal; most times it doesn't (and our bikes all wet sump at a slowish rate); ie the rate the oil can leak through the pump, across the ineffective check valve, and between the crankshaft journal bearing and the inner bore of the bush AND sometimes the crankcase and the outer diameter of the timing side bearing.
Only the last few weeks I have been checking this leakage issue, on one of my bikes. The clearance between the crankshaft journal and the bush is within tolerance; only 2,000 miles on the bottom end, and the oil is CLEAN. The crankcase had 0.012" machined out of the bore for the timing bush before the hole would clean up (concentricity); yep the OD on the crank case was out of round! I discovered this when I removed the oil pump and blew air into the bearing oil delivery drilling (the same one which the check valve resides in). Unexpectedly, I got sprayed with oil which got blown out of points between the bush OD, and the crankcase.
So, I check the crankcases for concentricity in the timing bush bore, often they are out of round (high mileages or hard use I suppose is the reason). Then machine them until they clean up (are round), then machine a new oversize bush, and shrink in. Then final machining of bushing ID, to suit crank shaft.
Anyway, back to the oil leakage issue. I removed the sump plate, and observed the rate of leakage onto the bike lift over several days.
Then, removed the oil pump and squirted clean oil into the pump delivery port (across the check valve), and prodded the check valve open. I noted that the very light spring is so long that the ball often deflects sideways, and as a hydraulics engineer, if this arrangement was in a system I was involved with, I would give it the Order of The Royal Rhubarb. I estimate the existing spring as not more than 1 psi force; a 5 psi spring could be used here without any detriment to the lubricating system (the check valve is upstream of the bearings), and it would keep the ball securely against the seat when the engine was stopped.
OK, so re-fit the oil pump after attempting to assess and perhaps clean the check valve, and wait a few days. The results; no change - same size puddle under the bike, and some oil also from the timing side, under the oil pump.
Next step; fit blanking plate with solid gasket in place of the oil pump. This blocked the oil supply from the tank, and revealed the rate at which oil will leak down between the crankshaft journal and the timing bush, and the big end bearings.
Result after a few days: same rate of leakage, but without the oil pump blocking the view, I can see the leakage into the timing cover is oozing from between the crankshaft journal, and the bushing ID.
I would expect the same rate of leakage (drain down) on the flywheel side of the timing bush. Eventually, after about 4 days of wiping oil up from underneath the bike, it ceased all together. This is because the crankshaft galleries, and the drillings in the case, had finally emptied. This exercise satisfied me that the bottom end clearances were OK; no major rate of leakage.
So in summary, if the built-in non return valve was effective, the engine wouldn't wet sump, regardless of the condition of the bottom end, or the leakage across the oil pump.
A weakness with the existing non return valve, is that the ball seats on the edge of the hole drilled in the (soft) aluminium. Any mark / blemish on this face, and the oil will leak through. therefore whenever apart, the ball should lightly be tapped into the hole, to re-seat. Be careful; hit it too hard, and the seat could be damaged! The A65 arrangement is better, because the seat is the hole in the base of the oil pump, and can be inspected easily, and re-worked. A hardened (replaceable) seat would be prefer
It is worth considering a replacement spring for the existing arrangement, but it has to compress enough to allow the ball to travel past the oil hole in the side of the crankcase (oil way to the timing bush), otherwise the flow could be blocked to some degree. I.e, the closed length of the heavier spring (which will be made of thicker wire) must equal or exceed that critical measurement - not hard to check during installation, but needs to be investigated.
What to do:
A50 / A65 engines; BSA repositioned the check valve so that it sits underneath the oil pump and is therefore readily accessible. Until recently I have had no experience with these engines, but now have an A50, so will have a closer look at their arrangement.
Do the unit twins wet sump as badly as the pre-unit? I would have thought not.
Has anyone modified A10/ A7 engines to copy the unit twins system?
I am starting to think up a way to solve this problem, and am putting together an A7 bottom end, so watch this space!
Richard
