The biggest difference is the drain-back design and how the baffle connects to the stepped 'ledge' on the left (drivers) side of the sump.
The STMPO baffle is welded all along this 'step' which better prevents oil from sloshing up on the step and away from the oil pickup. Also the STMPO drainback holes are more like slots than large dimpled holes. Because its smaller, the slots do a better job at slowing down how fast the oil fills that 'stepped' cavity and thus keeps the oil where it needs to be - in the sump where the oil pickup is. Having too big of drain-back holes as well as without welding the baffle to the 'step' itself will result in the oil filling the cavity faster and less efficient at keeping the oil in the sump than the STMPO baffle.
While the small drain back slots are good at preventing the oil from filling the stepped cavity and keeping the oil in the sump, it DOES NOT affect the drain-back efficiency of the baffle. Oil drains back into the sump to feed the pickup far faster than it lets oil out of the sump during cornering.
The main problem with the NSX's pan is this 'step' and open sump with no baffling in it. When cornering hard around right hand corners, oil sloshes up and fills this higher void which can eventually cause the oil pickup to starve and suck air -which leads to low oil pressure and a blown motor. Preventing the oil from sloshing up and filling this void is the key purpose of the baffle.
Through FXMD's extensive track testing: under long duration corners (5 second+) under full throttle, generating a sustained 1.0G+ can cause this oil starvation. Imagine holding the oil pan 90* sideways and all the oil sloshing away from the pickup. (You can even do this with your oil pan off and filled with water or oil as a demonstration).
We will have oil pressure data and the effectiveness of the STMPO baffle in the worst possible scenarios (street, R-compound, slicks, NA/Turbo, Professional driver, etc...) up in the near future. Stay tuned for future FXMD/STMPO collaborations and testing.
Billy
