mark_schnell

Here's an interesting video on the subject, from one of the channels I follow:

Mike Murphy

Intersting indeed. The LSPI study would be interesting to read up on. Back in the good old days of Chevy/Ford V8 carburetted engines, we would call this Ping or predetonation and it would be prevented by keeping the carbon buildups and compression ratios low. And no lugging.

What I didn't realize is that this LSPI indicates more of a graph than a specific RPM point. So using 10-15% throttle at lower RPM might be fine (the way an automatic does things) and then if you give it more than 20-30% throttle, a downshift occurs. Give it over 50%, another downshift (kickdown in the old days). Give it 100% and it's yet one final downshift (ever notice how many BMWs have a switch at the bottom of the accelerator pedal that only activates at WOT?). So, of course, engineers designed these cars to be operated this way to avoid these problems. Maybe we should drive the way an automatic transmission car operates with the exception of the fuel efficiency settings. Put an automatic in Sport mode, and you instantly realize higher RPMs across the board on many vehicles.

I tell my my wife to put her Q5 in Sport mode, which keeps her 2.0 Turbo in the 2000-3000 RPM range, especially since that car will run the RPMs well into the 1100-1200 range all day long.

dprantl

Yes, I have seen that guy's videos although the one I saw I think he was driving in a car while talking about lugging.

The first part about the engine working less efficiently and creating more heat, then suddenly making a leap to the engine overheating (!) is a bunch of hogwash.

Then he talks about pinging/knocking, which, once again, is very easy to tune out at any RPM/load level. On a stock car this simply will not happen if the factory knows how to tune engines, and I'm pretty sure most do. You can greatly advance ignition timing at lower RPMs to get more power, but once you detect knock, you just pull it back a safe margin in the tune. Also modern engines have knock sensors that detect pinging and retard ignition timing - this is not 100% safe for the engine, but again, it should not even get to that point. Back in the day with fixed/vacuum advance distributors and no knock sensors, this would definitely be a problem. I spent literally hundreds of hours real-time tuning and testing my old 928GT with a SharkTuner2 and looked at knock counts till I was blue in the face. Once I was happy with the tune, I could do 20 WOT pulls starting as low as 1,000RPM and would never register a single knock; simply keep the AFR in a safe range and adjust the ignition timing accordingly.

Then there is LSPI. This is a real problem with certain turbo engines, this is true. I'm not sure it applies to a 3.6 liter twin-turbo engine though, which is almost twice the displacement of the engines that have typically suffered the issue. It sure didn't seem like an issue on my old 2.3 liter Saab 9000 Turbo with a small modification to boost to ~24 psi from the stock 15. It got a full stock 15psi at 2,000 RPM and on a cold day up to 24 psi by ~2,500 RPM. It made it all the way to just under 200k miles before I sold it for $500 due to just too much interior stuff falling off/breaking and the undercarriage rusting (it was a New England car) badly. The engine still ran perfect.

Dan

996tnz

Interesting discussion.

I'm normally watching the road, not the rev counter, and it doesn't get below freezing here, but I take it particularly easy the first half-mile or so (also pleases the neighbours). I then stilll aim to keep her between about 1900 and 3000 or so until she has warmed up fully. I just look for the coolant thermostat to have hit the top section, ie over 80C/180F - thats also pretty much when the car considers itself warm enough to let the driver check its oil level.

I try to avoid lugging her at any time, but especially when cold, so normally change down early for any uphill sections.

Should probably check her owners manual again for Porsche's word on it I suppose.