There was heavy pitting on 5th gear.  Metal debris had been flung all around in the engine and it took a toll on the piston and cylinder.  The engine still ran OK, but it was nowhere near as crisp and powerful as it was when I started the test.

I installed my spare engine to continue testing on the flat-top pistons.  The spare had a 94mm flat-top piston, and a cylinder head that flowed a bit less than the one I was running on the 97mm engine.  It wasn’t as powerful as the 97mm flat-top engine, but it still had a lot of steam.  I started a new 10,000-mile long-term test.

The smaller engine ran well throughout the long-term test.  I confined the WOT stuff to 1st thru 4th gear.  It seemed like a prudent move since I wasn’t exactly sure why 5th gear failed on the big-bore.  I kept a close eye on the magnetic drain plug and no chunks of metal showed up.  I did notice a gradual decline in compression as time went on, but the engine ran perfectly.

At the 10,000-mile mark it was time for a teardown.  I wanted to see exactly how things held up.  To my disappointment, 4th gear was severely pitted.  

Reducing load is the obvious choice, but I don’t wanna reduce the load.  Increasing the load is the object of my affection.

Improving the material is cost prohibitive, so that one is on the back burner.

Increasing the load bearing surface is a viable option.  The old 4-speed gears are wider than the newer 5-speed gears.  My two failures were on 5-speeds.  I am currently running a 4-speed to continue testing.  Unfortunately, I’m collecting metal chips on the magnet again.  Although I can’t see any gear pitting, it’s obvious something is failing.

Option 1 – Bigger Pump

As it turns out, the LS650 uses a gerotor pump that is like the SV650.  The SV650 has three more camshafts, one more rod & piston, and one more gear.  It needs more oil.  Wouldn’t it be nice if the SV650 oil pump could be adapted to the LS650.

This is a stock LS650 oil pump.

The SV pump does not incorporate a bellmouth suction strainer, but it just happens to utilize a rotating assembly that is exactly the same outside diameter as the LS.  The two pump shafts are the same diameter too.

Here you see the SV rotor next to the LS rotor.  Except for the thickness, they are pretty much the same.  The SV rotor is .512” thick while the LS rotor is .394” thick, a 30% increase.  That should result in a 30% increase in volume if you ignore slippage.  Note that the SV rotor I purchased on eBay is kinda beat up.  Oh well, it’s an experiment.

If you are very, very, very careful and get the setup just right, you can machine the LS pump housing deeper to accept the SV rotating assembly.  You have to get the face and bore running dead-nuts to pull this off, but if I can do it anyone can.

I managed a solid 20% increase in flow on the dragbikemike test bench.  Not perfect, but I’ll take what I can get.  Looks like the modified pump might be a significant improvement.  It will bolt right into the LS650 case, use the stock shaft and drive gears, and for sure pump more oil.  It requires complete engine disassembly and I have no idea how the stock system will handle the increased flow.  More to come.

What if we could switch those gears?  Put the 35-tooth gear on the clutch, and the 30-tooth gear on the pump.  Then instead of a speed reducer, we would have an overdrive system.  One revolution of the clutch will turn the pump 1.17 revolutions.  You could install the hi-speed gearset without tearing apart the engine.  Seems like a good deal to me.  Less work to install, 36% more volume.  What’s not to like about that?