Somebody correct me if I'm wrong here. Any tables will be specific to 96530006.

First, load has nothing to do with injector scaling.

Raw load is based on MAF pulses between CAS pulses and is multiplied by the 1-D MAF Size value.

The baro voltage is scaled and then raw load is multiplied by the scaled value to get baro compensated load.

The IAT voltage is converted into a temperature, that temperature is then looked up in a table (333e) to get a scaling coefficient. The baro compensated load is then multiplied by that coefficient to get the baro+IAT compensated load.

These loads are only used for table look ups though, and are not part of the IPW calcs. Compensation for that is done separately.

There is an IAT based IPW correction coefficient table (36DC) that is separate from the load correction that is used to correct the IPW for IATs. The scaled baro mentioned above is used to correct the AFR for baro changes.

 

But many people use Injector Scaling to compensate for a MAF that isn't calibrated dead on, so it indirectly effects Load. Mostly AFRs and Trims are used to calibrate, and these can be effected by MAF or Injector Scalings.

 

And that's my thinking as well. I don't claim to know Mitsu's specific mass air logic, but I don't see how/why injector scaling would be involved in the load calc. In a nutshell, load is air mass per revolution, which is proportional to VE and torque output.

The question remains - how is it can 2-byte load be lower for a setup that generates much greater torque?

 

There can be up to ~10-15% difference in the different load variables. I think most people are logging the same variable though, so this probably isn't contributing, or at least not contributing much. JBs response on airflow through the MAF is key. I suspect that other contributing factors are tune and efficiency of the turbo. Remember that load is a measure of airflow per revolution. HP reaching the wheels is depends on how well that air is utilized in combustion and what fraction of the HP is wasted on spinning the turbo.

 

it would be easy to put the wrong car details in vdr and accidentally skew the numbers, was that double checked?

 

Yes, actually, we can take VDR out of the equation. For example, my still-being-tuned 93 octane map has logged a peak of 433 2-byte load at ~32psi from an HTA3582. The AWD car shows 420 load, and is dyno tested at 500+ ft/lbs at 39psi with an FP Black and E85. Even without VDR, I'm quite confident I'm not anywhere near the torque figure of the AWD car.

Again, I realize non-mechanical factors like ignition timing come into play, but the load figures just seem too far off the expected pattern.

 

The uncompensated load calc from MAF signal has no factors that are user adjustable. As other people said, baro and air temp-compensated loads are calced. Baro+airtemp compensated load can be 10-15% lower than uncompensated load in warm air especially if an exposed element filter is being used and the engine is sucking in lots of hot engine bay air. Seems to me that this along with some efficiency factors could be enough to explain your comparison.

 

If that's the case, then it's good news to me. I've got plenty of airflow. I just need to get out, put my 'tooner' hat on tightly, and go find the detonation threshold ...

 

another data point is that my dynojet numbers at 38psi were 539tq through the stock ecu/mas but I don't have a honeycomb anymore. Removing the screen dropped my load about 20 IIRC. My load on those pulls were hitting 452.

FWIW DLL showed 518tq a few hours early on the street

 

I noticed that the "screw" on the bottom of the MAF will allow more air to bypass the sensors in the MAF. Thus, the AirFlow Hz will go lower.

I tested this when I was having issues with installing an HKS Suction intake.

I have since left the "screw" set to have the inside flush and all the way open. I noticed that my load values are not as high as those with similar mods.

 

Do you still believe that there is something happening with the load calc that people haven't already mentioned?

 

No, by default I understand load as a calculation of those variables that relate to mechanical efficiency. None of those should include user manipulated variables. I just didn't want make assumptions about Mitsubishi's logic.

In my case, it's probably just that since I am presently tuning an octane limited fuel, I seriously doubt I can approach MBT at 30 psi. Where that is evident, the mechanical efficiency is there, but the combustion efficiency isn't due to the octane restriction. That would lead to a high mechanical load being indicated, but torque being moderate on account of the fuel limitations, so it makes sense when one thinks about it that way. FWIW, Drifto and I both observed much greater indicated 2-byte load since installing the same CNC ported heads, so that is significant within itself.

 

WOW! I remember being blown away by your peak load numbers and that is why I **** my pants the other night when I hit 467.5. My car makes nowhere near what your does though.

 

Another thing to point out, the load calcs look to be based on the ideal gas law.

Nothing in life is ideal.

Two cars could have the exact same inlet air mass, and make the same power, but if the inlet conditions are different and the ECUs approximations to account for those differences are not accurate, you will see different corrected loads.

Also, EVERYTHING has manufacturing tolerances. For all we know, the MAF could be calibrated at 500Hz to +/-5% and yet at 2000Hz, there could be 20% variance between meters.

 

Well Mitsubishi's logic includes the Karman MAS sensor. Load is calculated from measurements from that sensor based on the calibration of the sensor and the ideal gas law.

As many have said already, anything that alters the Hz reading from the MAS sensor's original calibration can and will affect load numbers (not to mention manually chaning the calibration with the maf scaling tables). A simple air filter or intake pipe can have a pretty big difference. A 20% error up or down at a 400 load level is +/- 80. It doesn't mean the mass airflow is different by +/- 20%, but the calculated mass airflow (load) is.

Different baro and intake temps also can change the load numbers dramatically as well, but that should (as long as intercooling is similar in the two cars) correlate with torque.