Classic MB Streetfighters
Let's Talk Performance Tuning the M104 Engine
More specifically, the 3.6L AMG engine. A good place to start is Ignition Timing. Note the stock 3.6L Wide Open Throttle (WOT) ignition timing map. A few things jump out. First, low single digit timing values below 2000 RPM and never more than 12degrees below 5000RPM. The engine can and absolutely will make more power with several more degrees of timing advance throughout this range. The torque gains are surprising with the best gains in the 3000RPM range. Why did Mercedes map this with such a tame and odd shaped curve, with a trough at 3000-3500 RPM? I can only speculate that was done to allow a more linear power delivery instead of all out peak performance. The next thing to jump out is the large spike in timing values over 5000RPM, adding another 12 degress by 6000RPM. In actual running, the values will not hit those numbers outside of very high octane fuel and REALLY cold air temps. The HFM utilizes 2 knock sensors and will pull individual cylinder timing in 0.35 degree increments. The actual timing will only be what the knock limit of the fuel and air temps allow. It is quite common to be several degrees below the mapped value due to knock retard. This is also why its hard to gain peak horsepower with these engines. The knock sensing is VERY good and protects the engine diligently.
In order to get past the tendency to knock, fuel has to be added. WOT fueling can be mapped to increase fuel. The additional fuel does very little to add power but it allows more timing before knock onset which is where the power gains come from. The 104's stock injectors do not have a lot of headroom but the AFR can be tuned from mid to low 14s down to low 13s. There is room to go slightly lower even, into the high 12s but this did not show any gains on the dyno. It does help with peace of mind with the relatively poor premium grade fuel available locally for the test mule.
About Airflow....
​
The intake manifold switchover point and camshaft advance / retard point are both tunable. When running the 3.6 engine with the 3.2 module, some of the best gains are from optimizing the switchpoints. We found a 39hp gain at 4400RPM from optimizing the camshaft switchpoint. We ran the car on the dyno with the cam always advanced and another run with it always retarded. Comparing the graphs showed a point where the power crossed over implying an optimal switchpoint. This was almost exactly where the switchpoint was originally mapped for the 3.2 engine. It was surprising to find that this led to a significant dip in power at the switchpoint. By further delaying the switchpoint several hundred RPM, the dip was tuned out. The intake had almost the same behavior, a dip at the switchpoint that was remedied by switching notably later. If you have a 3.6 swap running on the 3.2 ECU, significant power is being left on the table just from the cam and intake switchpoints.
Lastly, its worth noting that on the 3.6 engine, a stock 3.2L airbox with a new quality paper air filter only cost 3hp versus an open inverted airbox lid. Don't spend time or money trying to re-engineer this piece, its damned near invisible to airflow as is.
About EGR...
​
EGR feeds exhaust gases back into the intake under certain conditions. This spent gas serves to lower the combustion temperature in the cylinder. This helps emissions, notably NoX emissions. If your car needs to meet emissions requirements, EGR can be left functional with virtually no adverse effects to WOT power. It is not active during WOT as mapped. Why delete it then? Well in part throttle operation EGR is allowing hot exhaust gas into the intake. In an all out effort, that intake heating may be incrementally reducing knock resistance, hence tuners disabling it altogether. This should not be done on emission compliant, street driven vehicles.