//steve\\

^I wish you were a geek in my area.

scottinal

Is that on 93 octane?

Gernby

I do use 93 octane.

imstimpy

I'd love to see a plot of torque versus throttle pedal position. You might need a load-based dyno to obtain this, unfortunately... I'm curious if you are reducing the transient conditions below WOT or if you are increasing the resolution in an already small area (make sense?). I'll be honest, I haven't looked closely at the the pedal and throttle plate relationship.

I've seen this throttle plate strategy on boosted cars. BMW and Subaru use variations but they have boost acting as a load modifier in a requested torque strategy. On the S2000, we only have RPM and throttle plate. If RPM remains constant, and the throttle plate is forced open aggressively, does that mean you are setting the torque curve through your timing curve?

Whenever somebody claims "better throttle response" the red flags are raised in my motorsports brain. A more aggressive throttle to pedal relationship is not the same as torque response to a throttle plate delta. Linear torque production versus pedal position is preferred for motorsports whereas the Sport button on an E46 M3 is an example of "better throttle response" obtained by changing the ramp-rate of the throttle plate to the pedal position. It feels faster, but at a loss of pedal resolution.

Gernby

It's super easy to control load with the DBW throttle map. If you want to see how much power you make at 50% throtlte, you just set the max throttle to 50%, and do some pulls. That's how I do my part throttle tuning.

I really didn't understand what you were asking in the other questions... :|

On a side note, my fuel economy has gone up so much more than I expected.

imstimpy

Let me try to explain my thoughts in a different way.

I'm curious what the torque output looks like a a constant RPM across all throttle plate positions.

In a fictitious world, 0% throttle plate position would produce 0 torque, 50% plate position is half possible torque, and 100% throttle position is all possible torque. This would be described as a linear throttle.

In the real world, the movement of the throttle plate is not linear and varied effect on torque output at the various throttle positions.

Quoted above is the part I don't understand fully understand. If I'm reading this correctly you are stating that the remaining 50% of the pedal position is now unused? ie: 50-100% pedal position produces the same load because it fully opens the throttle plate? That correlates to the E46 example in my post where "better throttle response" is achieved by reducing pedal position resolution.

Just to clarify, you are trying to solve the issue of two different throttle plate positions generating similar load but requiring different tunes? Your solution is when two different throttle plate positions generate similar load, you force the throttle plate positions to be the same?

Rounding back to my original question, how does the torque output of the motor correlate to your remapped pedal position?

Can this problem be addressed by increasing the resolution near 1 bar, as you've done, without altering the DBW table?

MegaTRon

Good stuff, gernby.

Gernby

Increasing the resolution at full load defijnitely helps on its own, but it doesn't totally fix it. Since barometric pressure varies from day to day, there's no way to guarantee that a particular load level will always be WOT vs. PT. That's why the remapped throttle plate will help reduce PT at full load even further.

My goal with the throttle mapping is actually to make the torque / throttle relationship even more linear than stock. In stock form, the ratio is logarithmic. When companies impliment "sport mode" pedal maps, they are actually making the throttle even more logarithmic.Basically, the regular map causes the "usefull" part of the throttle very small. For example, at low RPMs, the only part of the throttle pedal that does anything is the bottom 10%. Everything in the 10%-100% range does virtually nothing until RPMs start to rise.

I created a thread here about remapping the throttle plate so that it is more emjoyable to drive. This made the pedal less "sporty" at low RPMs, but much easier to drive smoothly at slow speeds (parking lots, etc.). I'm now taking it a bit further to make the full load / PT portion of the map very agressive so that the throttle plate will be fully open more of the time while at full load. I think it gives the throttle a much, much more linear feel.

imstimpy

After speaking to a number of people, and consulting a few books and SAE papers, I think I have a better understanding of what is going on.

You are addressing two issues here:
1) Multiple throttle plate angles will generate similar load on the F20/F22 under certain conditions.
2) The F20/F22 suffers from excessive intake resonance and requires additional load resolution.

You have addressed the first issue by recalibrating your DBW table.


If max load is achieved at two different throttle plate angles, the throttle plate is no longer a restriction to airflow beginning at the lesser throttle angle. Under this max load condition, the throttle plate affects airflow without necessarily restricting it- different angles will generate different resonance, pressure waves, turbulence, whatever you want to call it. The lower the RPM, the lower the throttle angle necessary to eliminate the airflow restriction.

Based upon that, I'd think the DBW table probably ought to be recalibrated by logging all loads against all pedal positions (throttle angles) against all RPM. I was thinking the end result might look more like this:


In the picture below, is that the maximum throttle angle necessary to achieve full load for a given RPM?


Certain throttle angles might be more efficient than others. It is a very real possibility that the ideal throttle angle isn't always 100%. In your DBW table, you have decreased the throttle angle at lower RPM, but is that correct for maximum efficiency? Did you remap your DBW table to generate a consistent load for a constant pedal position? Once again, the best way to figure out the ideal throttle angle would be to see torque output for all throttle angles at a given RPM.

You have addressed the second issue by expanding the resolution at critical areas near max load.


This allows for more precise control of resonances, or slight fluctuations in VE, and variations in barometric pressure. If you simply interpolated the additional columns you have only replicated what the ECU already does. However, if you logged specific loads and altered the tune you have now taken advantage of the additional resolution.

Gernby

The only things I would add / change about what you said are:

1) I wouldn't call it "suffers from excessive resonance", since that resonance causes significant performance increases. We like resonance!
2) When the throttle plate is open just enough to achieve full load, it isn't the slight restriction or turbulance in the intake air that is preventing resonance. It is the reflection of the pressure waves (negative pressure waves from the intake valves that bounce off the back side of the throtlte plate, and positive pressure waves, or rarefraction waves, from the air filter inlet bouncing off the front side of the throttle plate). When the throttle plate is fully open, those pressure waves travel freely up and down the intake plenum, which gives a "super charge" of the intake air in certain RPM ranges.
3) The reason why my throttle map doesn't look like the one you suggested is because I don't like the 1:1.5 pedal to plate ratio of the stock map. I wanted to maintain true part throttle / reduced load for the bottom ~60% of the throttle pedal across all RPMs.