Camshaft Data

Digital Corpus said:

The correction factor doesn't account for the difference in cooling, or lack thereof, of the intercooler.

Click to expand...

Yes, the intercooler performance would change on a warmer day. BUT by the second or third dyno run the engine and the intercooler are heat soaked anyway. We see the second and third runs putting out a bit less power on most cars with no changes except the heat soaked engine and intercooler. But that near 20 hp drop in horsepower from the 195 you put down when the engine was first run a couple of years ago to the 175/180 on June 20 was not all due to the higher temperature. My car picked up about 5/6 hp and 20 lb/ft with no change but the 11mm pump. My tune did not change. Just more fuel and much more smoke from the 11mm pump.

So I am thinking the summer temperature is not really anything close to the 20 hp change we see on your car. Most of the delta has to be due to changes in your tune?

Correction factor is linear... it works well for N/A engines but with compressors, you have a non linear response to reduction in density, like it does with ambient pressure.
The compressor flow maps have 2 axis, one for pressure ratio and another for air mass. Changing the air mass changes the efficiency of the turbo, which can make it use more power to maintain the same pressure!

BTW, I can't access the vcds logs, says cannot connect to website?

Try the logs again please. I'm not seeing any issues on this end.

Found the cause of the lower dyno numbers. This happened within about 3 weeks of the dyno. Caliper slide pins gunked and seized.

I got the logs now, thanks!

Some weird stuff is seen. I am looking at where the maf shows below 1275.

Looking at timestamp around 24 seconds:
3213 rpm 1195 maf 3093 mbar.-> 3.1 bar -> 1195/(3.1)= 386 (mg/s maf per barA)
3549 rpm 1080 3.1 bar -> 349
3864 995 3.1 -> 322
4158 925 3.1 -> 298
And so on it drops.
There are similar readings on the second run in the log.

But it seems like it is happening too soon (see my log below)

With this data, I can't see where the ecu can tell if there was more airflow or not from the cam.

But, I can say that when you max out the maf, it shouldn't drop that soon in the rpm range. I can't find the post but someone calculated where the stock maf gets maxed out.
For example from a log of mine from this spring after the maf drops below 1275:
4305 rpm 1215 maf 2.6 bar - > 470 (mg/s maf per barA)
4452 1175 2.54 463
4620 1145 2.5 458
4767 1105 2.46 449
4872 1085 2.42 448
4977 1065 2.39 445
5124 1010 2.37 426
5271 990 2.34 423

See how the number drops, less efficiency, less VE. But still even if your maf was saturated, you should be getting same or higher values than I get with less boost. A saturated maf would show as a sharp drop in the maf reading- because the maf is reading the same mass flow (kg/h for example), but the rpm is climbing- so mass/rpm factor would keep dropping fast.

So, I don't know if you tried another maf, but yours might have a problem.

As has been already mentioned, it's not useful to look at the absolute peak MAP/MAF value (i.e. at the torque peak) when comparing the OE vs Colt cams and the respective timing variations due to sensor saturation. More useful would be to superimpose all the curves on common axes and look for differences in MAF mg/stroke away from this saturation area, at the same given RPM to give insight of the volumetric efficiency. By eyeballing across several graphs, they look very similar, but putting them all on a single one will show more clearly what if any differences the cams and timings have.

Another graph can be generated for the actual volumetric efficiency - since there are also MAP and IAT logs, we can find the density.

The volumetric efficiency is then mass flow [kg/stroke] / (density[kg/m³] * 0.25 * displacement[m³]). The factor 0.25 is because only one cylinder is having an intake event per stroke.

Picking out a convenient data point for eyeballing, I chose the Colt standard at 4250 RPM:

MAF: 900 mg/stroke = 0.0009 kg/stroke
MAP_actual: 3050 mbar = 305000 Pa
IAT (R1, just guessing at 9 sec corresponding to 4250 RPM in the other graphs): 45°C = 318 K

Density = P/RT = 3.34 kg/m³
Displacement = 1896 cm³ = 0.001896 m³

Volumetric efficiency = 0.0009 / (3.34 * 0.25 * 0.001896) = 0.568 = 56.8%

This all assumes of course perfect sensors and measurements, which they are not; there is the possibility for considerable error, so a value of 56.8% should not be judged in isolation but rather compared across other data sets, in statistically significant sampling and against what can be found in the literature.

Edit: I must note that the MAF readings seem to be saturated in the higher RPMs too, because of the hyperbolic shape of the curve - just like a torque curve with constant HP @ increasing RPM will also have a hyperbolic shape. Physically, this makes sense - the MAF measures a time averaged-heat loss from the hot-film and then the ECU takes this analog data, then transforms and discretizes it to RPM to arrive at mg/stroke.

Yeah, the readings aren't a good gauge but if they weren't way low (compared to my numbers) you could use it as a comparison in flow between the changes.

The key issue is that his maf reads very low compared to mine. I don't think it should drop that low. With the same maf - one that is saturated shouldn't give lower readings than an unsaturated maf.

TDIMeister said:

Edit: I must note that the MAF readings seem to be saturated in the higher RPMs too, because of the hyperbolic shape of the curve - just like a torque curve with constant HP @ increasing RPM will also have a hyperbolic shape. Physically, this makes sense - the MAF measures a time averaged-heat loss from the hot-film and then the ECU takes this analog data, then transforms and discretizes it to RPM to arrive at mg/stroke.

Click to expand...

The maf readings hit two different limits, 1250 mg/stroke (IIRC 1000mg/s in the stock file) which is a software limit, and another figure in kg/hour higher in the revs that is the actual hardware limit of the MAF itself.

I'm running a 3" maf with the linearization scaled so that the computer "sees" 50% of what the MAF actually reads, so the software limit is taken care of from the scaling, and the hardware limitation is taken care of by the larger MAF body.

You just double the smoke maps to suit, so you put in the same amount of fuel for the halved MAF data.

Why not run MAF delete?

Because MAF Delete tunes (I've had many from different tuners) are generally not as good as ones that include MAF.

MAF based tunes generally are more responsive, less tip in smoke, less temperature sensitivity to smoke, etc.

For maximum performance it may not make much difference, but for every day driveability and consistent performance over ambient temperatures it does

oldpoopie said:

Why not run MAF delete?

Click to expand...

Because I'm a weenie.

It just seemed obvious to me, if I ran into anything that was a dead end I'd probably have given up pretty quick, it all went pretty easily (although getting the linearization proper for the larger MAF took more excel skills than I was able to muster)
It has a sweet-ace device to measure the air that it's breathing, makes tuning for smoke free operation real easy.

I was seriously working towards going mechanical on the pump while trying to figure out how to reflash the ECU. Just to give an idea on the kind of mindset I was coming from going into all this a year ago.