Traction off, Torque Management Continues

Scan D5B

Today’s scan was with traction control disabled, to determine if the Torque Management Advance (retard) (TMAR) was originating from TCS. I observe that we still have TMAR with us, so I conclude it is not, and further study will be required.

Comparing today’s D5SB with prior D5SA, today’s peak was 468.9 whp vs prior peak 476.8 whp, or 8 less. Prior was a warmer day with somewhat higher humidity and earned a larger dyno correction of 2.4% vs 1% today. I am looking for repeatability here, and under 2% may be par for this type of testing.

I have added relative humidity to my scan table, so that the resulting scans can be self-correcting for conditions via dyno correction factor. The car tracks barometric pressure, relative humidity, and ambient temperature. With those figures an SAE 1349 dyno correction can be applied, and is figured in prior to graphing. I am also using a 400/464 ratio here for crank to wheels; I find this matches precisely with stock results.

Stock vs GM Dyno

This graph takes the GM LF4 engine dyno, converts it to wheel torque/hp, and compares it (dotted lines) to a scan from a stock ATS-V also converted to wheel hp (solid lines). They appear to overlay nicely.

HPTuners to Hp/Torque Analysis — defactored, with dyno correction, tranny losses

This is an analysis of a on-road hptuners scan. It is easy enough to export from hptuners to a csv file. This lends itself to graphing in excel or similar tools. However, the key info we want to graph, HP and Torque, are sometimes factored upward in tuning. By analysis, if we can discern what the factor is relative to a stock tune, we can de-factor these values prior to graphing.

This chart shows such an approach. Export from hptuners. Add a column next to Torque, and then modify that column to de-factor the values, and also to correct for SAE conditions.
For defactoring, consider the stock values and the tune values for airmass coefficient(s), and resolve a factor by RPM. IN this case there was a fairly consistent factor across the range, so I choose a value slightly higher than the max factor in the range to cover the entire range. I am considering modifying the tune to make the factor consistent, but it is precisely correct as is.

For the dyno conditions correction I used an online calculator, with baro and temp from the tune; I need to sort out the formulas so I can simply do the math in the excel.
Also I added a correction for wheel versus crank; the ATS-V stock averages 425 whp for a 464 crank hp rating, which is 91% or 9% transmission losses. This could be used, although I instead chose to use 400 whp / 464 hp or .86 or 14% tranny losses which is more conservative (produces lower numbers) because we are going in this case from crank to WHP.
Add another new column for WHP, calculating it from the torque and rpm. Then chart as an XY/Scatter diagram, for defactored and corrected wheel torque, WHP vs rpm.

The blue line is WHP, the read line is WTorque. I added the boost pressure in gray, choosing boost/vacuum in this case.

The large event 5000-5300 rpm does not repeat in every WOT run; I wonder if it is a physical event on the course, such as a large bump or dip. This needs more study. Update: This appears to be a traction management advance event pulling timing. More study needed to lock it out.

There are three gold lines. The solid line shows the VVT intake cam angle. The dashed gold shows the intake cam angle commanded. The dotted gold line shows what I suspect is an optimal tune. I arrived at these values based on the stock tune’s estimates of hp generated at various airmass, RPM, and intake cam angles. This needs more study.

ATS-V LF4 Boost and Baro

I have been thinking about boost and barometric pressure, as what I see in the scans confuses me. Complicating the issue is source of the signal, meaning there are multiple sources of related information, and the poll rate or how often that source updates.

Basic Issue snapshot:

At a specific scan snapshot I see this:

Boost Math 17.7 — this is a parameter in hptuners math that takes MAP minus Baro
Boost / Vacuum 18.7 psi — this is a sensor that compares total boost to vacuum (baro)
Absolute boost (not referenced to baro) 32.7 psi
MAP 31.6 psi
Baro (SAE) 13.49 psi
Baro (mfr) 14.43 psi
Baro (SAE) chart update 13.69 psi

First, let me say I love the detail that we can see in the scanner.

I do get how each of these numbers should relate — for example, absolute boost minus barometer should be what Boost/Vacuum is; that’s the definition. But the math does not quite work. There are 3 baro readings, and they differ slightly, so there’s that. But none of them fit perfectly. There are 2 total boost readings, MAP and Boost Total, so there’s that, but neither of them fit perfectly with any of the 3 baro readings to result in the parameters shown.

Example: Abs Boost – Baro (sae) = 32.7 – 13.49 = 19.21 psi

2348 Boost Math says it uses 50032, the Generic Sensor for Baro, and 50030 the MAP.
Boost Math scans at 17.84 psi. 50032 Baro reads 14.06 psi in that instant. The MAP reads 31.90 in that instant. 31.90 – 14.06 = 17.84, so that works but depends on recognizing the EXACT baro and MAP used in the instant. So Boost Math is Manifold Absolute pressure minus the current (instant) baro reading.

The “Baro SAE” on the hptuners chart at that instant reads 13.49. But when I ‘chart’ baro sae it shows the 14.06 psi at that instant. I suspect the difference is the refresh rate for each.

Boost/vacuum in the chart at that instant reads 18.75 psi, which is almost 1 psi higher. In the channel list it shows at 18.34 psi, which again I suspect is refresh rate. Notionally it is reading a similar thing — total MAP boost minus baro. MAP at that instant (chart) is 31.9 psi versus the channel shows MAP 31.6 psi. We know the source is not the same as boost math, since they differ. So what are boost vacuum’s source values? I observe that Boost Pressure sensor at that instant is 32.84 psi; if we subtract the chart instant baro of 14.06 we get 18.78, which gets us close to 18.75 psi, although not exact. So Boost/Vacuum appears to be the Absolute Boost sensor minus the current (instant) baro reading.

SO which is more accurate, Boost PSI (boost math) or Boost/Vaccum? The difference appears to be the difference between the 32.84 psi Boost Sensor and the 31.9 psi Manifold Absolute pressure. Why do these two sensors for notionally the same thing vary, and which is ‘more accurate’? My intuition is since there IS a Boost sensor, to use it for boost. That makes the Boost/Vacuum the “correct” boost versus the Boost Math.