I posted this over in the Oil Pump Solenoid thread but thought that others who are discouraged from reading 2 million posts might find this interesting.

OK, here are some interesting observations. My car was stationary and at idle when I took these readings.

The first is tranny temp at startup. TCU reads 27°C (80.6°F), display shows 88°F. Difference of about 7°F or about 9%.




The next is coolant temp. ECU shows 79.5°C (175°F), display shows 192°F. Difference of about 17°F or about 10%.




The last is oil temp measured with two different touchless (IR) thermometers. The thermometer readings were of the oil pan at or near the drain plug. Thermometers show 186.4°F and 179°F, for an average of 182.7°F. The display shows 183°F. Difference is 0.3°F or about 0.2%!





So, according to my empirical observations, the oil temperature display is far more accurate than the two systems that have actual sensors!

Hmmm, what shall we do?

 

I think the data source is essential to trust results. The IC display is nearly useless for our purpose. We need trustworthy data to step this up.

​​​​​Master Surya has the best setup to collect engine live data. That's one of his field of expertise.

We have already visited the whole topic of heat instrumentation. This is what got me to pull the plug and experiment!

Now we are interested to see how oil removes engine heat given enough RPM to provide spray pressure above 25~30psi.

The question we want to qualify is:
- "What RPM is necessary to spray cool pistons given oil conditions?"

We already know viscosity is a function of oil type + oil mileage + oil temp.

> Interesting live data graphs:
oil viscosity vs. Engine RPM + (Oil Pressure) + Oil Temp + Coolant Temp

I believe we're going to see spray RPM is linked to oil pressure that's linked to oil temp thats linked to viscosity - Meaning here our only Temp control is oil viscosity if we don't want to hold RPM manually.

When oil is not sprayed well at driving RPM then a dry-lube oil package is needed to lubricate dry cylinders that are then leaking compressions.

 

Besides being curious, I was really wanting to see how closely the displayed temps match the measured temps. There was doubt cast on the accuracy of the oil temperature displayed because it was calculated rather than measured. My observations show that the oil temperature displayed is more accurate than the coolant or transmission fluid displays, at least while at idle. Next will be to take measurements while driving -- just got to figure out how to get under the car when it is moving.

 

Banks gauge seem perfect for logging experimental data.

It collects data PID across CAN bus that I've shown to be bottlenecked from factory. I would not drive longterm with anything plugged in my DLC. Collect data then unplug.


Surya has pretty much logged everything necessary already. That what enough for me to pull the plug.

 

The details we are dealing with can be easily lost in confusion. We have different engine types, different ambient temps, different driving style, different tranny, different favorite oils and are concerned with similar issues.

We are dedicated Mercedes patrons that don't fully enjoy dry-lubing burning hot cylinders up with random squirting. ​​​​​​

We need to spray cool pistons at driving RPM to seal cylinders that balance contribution and restore the engine response from 900 to 3000.RPM.

That means if we are going to compare improvements, we need one or more BASELINES per chassis and per oil to compare to, right?

Its not that we are going to find something we don't know... some of us are already experimenting MOD-2.x. We want to measure "with vs. without".

Here we are only DEALING WITH ENGINE HEAT REMOVAL, not the super power/torque related to GDI timings. That should be somewhat easy to capture eventhough MB has deleted many useful sensors.

We want to keep this simple but solid without getting lost in much data overdose. We just need good meaningful data...
What constitute good results: we know "garbage in, garbage out", yes?

Many numbers with different meanings. We don't want to farm data for no reason... we need testing to be PRACTICALLY USEFUL!

I already got my own answer to control heat and GDI. I call it MOD-2.1 to boost viscosity as it gets lost with heat/age.

I grew up with digital truth: 0/1, yes/no... the world physics is made of shades of grey, so is good visvosity to spray at 1500.RPM.

Simple setups can cool at 3500, 2750 or 500. RPM.... I like 1500.RPM not 2500 or 2750.

I don't think ECU likes leaky dry rings at driving RPM.

Perhaps some of us overlook that MOD-1.0 only spray before 3000. RPM ie. "pulling the plug" gets not much improvement at all. A little extra protection to always be cooled a little sooner.
MOD-2.0 drops that to 2500 for a while.

That's what all the data digging is going to drive back to

"Knowledge is organized data!"