The term “Speed Density” typically refers to the operation of an engine using an estimation of airflow based on pressure, temperature and some idea of the engine's volumetric efficiency (more on this below). This page outlines the basic concepts behind speed density operation to serve as a foundation for a deeper understanding of ECMTuning's implementation later.
If you're looking for more of an “ECMLink install/setup” page, then this page may be what you're looking for.
Speed density operation starts with the basic ideal gas law, which states the following relationship between gas (air) pressure, temperature and volume.
pV = nRT
p = absolute pressure of the gas (air in this case)
V = volume of the gas
n = amount of gas
R = gas constant
T = absolute temperature of the gas
For purposes of this discussion, we can ignore what R really represents and just treat it as a constant value.
In calculating fuel requirements for an engine, the amount of air (n in the above equation) is the critical piece of data we need to solve for. To solve for that, we need to know absolute pressure, absolute temperature and volume.
Pressure and temperature are easy enough to measure in our application using off the shelf sensors like the GM IAT for temperature and the 4-bar sensor we provide on our site for pressure. Other pressure sensor options are listed on our Aftermarket Sensor List. These include the GM 3-bar, GM 3.3-bar, AEM 3.5-bar and AEM 5-bar sensors. Other sensors listed on that page can be logged, but are not suitable for SD operation because they do not provide a reliable absolute pressure reading. We also offer a Speed Density bundle on our site as well, which includes the standard 4-bar pressure sensor and the IAT sensor along with necessary pigtails and bung.
For air temperature, the only supported sensor is the GM IAT sensor (AEM IAT sensor is actually the same thing). You can get part number information off the Aftermarket Sensor List page.
When you connect these sensors to your ECU, note that you must connect the GM IAT sensor to the IAT input on the ECU. There is no other supported IAT input. For the pressure sensor, you can log that on any available (and compatible) input on the ECU.
Accuracy is important, particularly for the pressure readings. The MAP sensor should be connected as intimately as practical to the intake manifold. Long, narrow hoses and tee fittings in the path to the sensor slow the response of the pressure sensor, resulting in poor airflow determination. The delayed pressure response will result in poor throttle response and poor control over air/fuel ratio whenever the intake manifold pressure changes. Since both sensors are referenced to the ECU's sensor ground circuit, it's also important that the ECU's sensor ground circuit isn't disrupted by having high-current devices attached to sensor ground.
As normally installed, MAF translators do get their ground from ECU sensor ground via the Mitsubishi MAF connector and do draw a moderate amount of current, particularly when powering the GM MAF sensor at higher airflow, so expect the pressure and temperature readings to change a little when you unplug the MAF translator. Changing the translator's ground from ECU sensor ground to chassis ground instead will eliminate this effect. Cut the MAF translator's ground wire between the translator box and the Mitsubishi MAF connector and connect the end of this wire coming from the translator to chassis ground. Leave the wire end that remains on the translators Mitsubishi MAF connector isolated and insulated.
So we can pretty easily measure pressure and temperature. But figuring out volume is where the estimation piece of speed density operation comes in.
To avoid getting into a big discussion about the physics behind it all, I'll simply state that the volume of air flowing into an engine is estimated using engine displacement and a volumetric efficiency (VE) table which is indexed by RPM and manifold absolute pressure (the two parameters that most directly influence the VE of an engine). A higher VE number results in a higher volume estimate which results in a higher airflow estimate.
Volume = displacement * VE[rpm, pressure]
If the VE table lookup for an engine running speed density is larger than the true VE of that engine at any given point, then the resulting airflow calculation will be higher than it should be and the resulting A/F mixture will be richer than intended as well (because the ECU will inject more fuel than is required because it's assuming there's more airflow going in than there really is). So you can see that getting the VE table defined accurately is one of the more critical (and variable) pieces to smooth speed density operation.
Having seen above the importance of defining an accurate VE table, it stands to reason that you'll spend most of your time dialing this in. That's not to say it's overly complicated, it's just more complicated than wiring in a few sensors to measure pressure and air temperature.
A typical SD VE table might look something like the following (click to enlarge).
It may seem intimidating at first, but it's really not as bad as it looks. We provide a number of tools and helpers to make dialing the VE table in on your own car pretty easy. And, honestly, most people may find that they can just drop the “stock” VE table we provide into their car and basically fire it up and drive around. There's no doubt some amount of dial-in will be necessary, of course. But for basic driveability, it should work fine.
As you can see from the table above, the VE table is typically indicated by RPM on one axis and absolute intake manifold pressure on the other. This particular screenshot shows absolute pressure in units of psi. This normalizes the table for a variety of pressure sensors. It doesn't matter which pressure sensor you use as long as the ECU has some way to generate an absolute pressure reading from it. You can dial the car in with a GM 3-bar and swap out later to an AEM 5-bar without making any changes at all to your VE table.
The above discussion was all about the theory behind speed density operation based solely on air properties by themselves. But the reality is that air temperature is also affected by engine temperature. As air is entering the engine, it's heated by the temperature of the engine itself. This effect is more pronounced with slower moving air (idle/cruise) and gradually reduces in impact as airflow rate increases.
To account for this observed behavior, we have defined a coolant-to-air temperature weighting table in the direct access area of ECMLink. This table allows the user to define how much relative importance coolant temperature plays over air temperature (and vice-versa) based on airflow rate. The user can choose to ignore coolant temperature all together or ignore IAT all together or interpolate between the two.
Because we know this concept may be not be mainstream, we offer the following datalogs that have *no* air temperature compensation. All the temperature compensation in both these logs is based on coolant temperature.
These were captured minutes apart from each other on the exact same car idling in the exact same spot in our parking lot. In one, air temperature varies from 172F to 130F while in the other it's only 70F the whole time.
In both cases, you see the resulting A/F ratio (measured by the LC1WB) is identical and combined fuel trim is fairly constant despite the fact that air temperature was NOT being factored in. In the hot log, you can even see where we disconnected the GMIAT sensor for a period of about 30 seconds, but absolutely NO change was observed in measured A/F ratio or idle quality. That was done to specifically illustrate that air temperature was not being factored into the airflow estimation.
If airflow temperature had been factored in, you would have seen a swing of (172 + 459) / (70 + 459) = 1.19 (19%) in mixture! Simply put, the temperature of the air had absolutely NO effect on airflow entering the engine at idle. We have collected other logs at cruise to illustrate a similar behavior.
However, at wide open throttle, the air temperature seen by the IAT sensor quickly takes over as the primary factor in determining air density. We have logs showing this effect as well with measured A/F ratio dipping to 8.3:1 with IAT at 133F compared to A/F ratio of 9.1:1 at 65F with boost around 20psi. These logs can be found below.
As before, these logs were taken minutes apart from each other on the same car. The only difference between them is air temperature. In the first, air temperature averaged 65F while in the second, it was 133F.
However, unlike the idle logs, measured A/F ratio in these actually *did* change, suggesting that air temperature in these log actually did have an impact on airflow entering the engine.