Cadillac ATS-V Boost Solenoid B & WGDC Waste Gate Duty Cycle

Today let’s examine the turbocharger base DC (duty cycle)

The table above shows the base duty cycle used in closed loop boost control for the Cadillac ATS-V. Keep in mind that notionally for a waste gate, 0% is fully open, and 100% is fully closed. Left axis is desire pressure ratio, which is total manifold pressure over barometric pressure. So if the baro is 14.5 inHg for example, a desired pressure ratio of 1.2 would equate to  [(14.5 * 1.2)-14.5] = 2.9 psi, desired pressure ratio of 2 would be 14.5 psi, and desired ratio of 3 would be 29 psi.

The table above then can be read this way in PSI:

This will vary with Barometer reading; I used 14.5 here as an example to help me relate what the system is doing.

We are interested at WOT, when the system is building boost from 3600 RPM to 6250 RPM.

Here is a chart of what the Boost Solenoid B is reporting in that period on one run of my ATS-V:

The gray line at the top is the boost solenoid info. It is reading on the right axis, as %. So it appears to show 50% at 4800 RPM, rising to 60% around 5800 RPM, and then falling to 50% around 6200 RPM. The rapid fall after 6100 RPM may be part of the gear shift logic near that point, I am unsure.

The boost pressure goes from 15.7 psi to 18.4 psi across this run range & gear. So if the boost is a result of the desired boost, then the desired pressure range is operating in the 2 to 2.2 rows. The base DC from the table in that range is this part of the table:

so at 4797 RPM and 16.4 psi it would be commanding around (look down the 4750 PRM column to between 14.5 and 17.4, so around 2/3 of the way between 14.5 and 17.4 so 2/3 of 4% plus 42.84) for a ~45.5% duty cycle; the observed “Boost solenoid B” value is 45.9%.

At 6000 RPM and 18.4 psi boost it would be commanding around a 60% duty cycle. The observed is 60%.

There is a nice article on this topic — waste gate DC versus boost and system design here.

My conclusion is that because we are operating at less than full waste gate DC we are ripe for adding more boost to the system by commanding higher waste gate DC.

I am a bit unsettled about the overall strategy the OEM tune follows. Let’s look at 1.8 pressure ratio. The WGDC goes from 100% at low RPM (spool up fast!) then falls to 34% at 3600 RPM, and then rises to 42% at 6250 RPM. I am unsure why there is a dip in the middle, possibly to bring boost on more smoothly once the turbo is up to speed? Similar curve at 2 pressure ratio, but there is not the same dip in the curve at 2.2 or 3, where the commanded WGDC steadily rises from ~40 to ~60.

So what is the ideal turbocharger base DC? If we raise this focus range from pressure ratio of 1.8 to 3 by 5% we would go from 40-60 up to 42-63%; 10% would be 44-66%. A different approach would be to set the table to a flat 60% across the range; remember in the larger context of the whole row we are basically curving from 100% at low RPM down to 42-60% at high RPM depending on the row. Another thought is to stay close to the OEM but to remove the divot in the 1.8 row, smoothing this row, then raising each row from 1.8 to 3.

Here is a table where the high RPM WGDC is set to 50-60% and then is smoothed from 1900 RPM to 6200 RPM.

This would impact when boost is above 8.7 PSI or so up to max boost, should spool faster, and ensure that the desired boost level is hit. The caution would be that this may produce too much boost too soon, exceeding the desired boost and causing the throttle to close. It needs some testing to see what works well within the torque management scheme.

Where in a current tune actual boost is below desire boost it would make sense to increase the WGDC base DC in that area.

Cadillac ATS-V HPTuners Stock Scan & Throttle closing

I am starting to run driving scans on my 2016 ATS-V to get a feel for how the engine operates. The LF4 twin turbo 3.6L engines use a torque management based strategy. The control sorts out what torque is desired, then manages the powertrain to make that torque, independent of throttle input. The throttle input (gas pedal) is treated as a request — so you might have the pedal on the floor, but if the engine is making over 100% of the torque planned at that moment, the actual throttle may close to maintain the torque to the planned levels.

In the image above, the Cadillac is accelerating through 61 mph, at 5313 RPM. The pedal is at 100%. The actual throttle however, is at 67.8%. The ATS-V was making “too much” power at that moment. If you examine the 2nd row in chart vs time, the green line is throttle %, so where we see dips in the green vs yellow line is where the ATS-V is reducing engine power to desired or planned output.

In this case, at this point, the total boost pressure of 31.5 psi (without adjusting for baro) was higher than the Desired Boost pressure of 30.7 psi. So the engine controller is pulling back the throttle because too much power is being made.

I also note that the stock file is running a few points of KR, or knock retard. There is a ramp up of the planned advance in that range. The engine controller goes from a spark advance of 8 degrees up to 10.5 then down to 10 degrees. That appears to be too much since the KR (knock retard) is going to 2-3 degrees there. Although it is possible that a richer air fuel ratio (AFR) would be fine.

The ‘stock’ commanded AFR is 12.8, lamda 0.909 equivalence ratio commanded. This is within the range of what most tuners are doing with the LF4 direct injected V6. This ATS-V may prefer to be slightly richer, but needs more research.

Correction Factor: Ambient air: 72F, Intake air 75F. Baro 29.5223 inHg, 83.13% humidity so SAE J1349 correction of 1.01555. 453.6 hp corrects to 460.7 hp.

Cadillac XLR LH2 V8 Tuning – More Tune 9a

Tonight I had cooler weather so I re-tested Tune 9a for baseline comparison.

XLR Tune 9A2 Comparison Calc HP

The 9A2 runs showed some KR spikes at 4150 rpm and 5000 RPM. This appears to relate to a high IAT advance multiplier at those RPMs, which I will zero out for Tune 9B.

XLR Tune 9a2 to 9a comparison

The Virtual Dyno for 9a2 was equivalent to 9a1, which is good news (same tune, different day).

Tune 9B will also reapply the MAF 2% touch up around 8,000 hz since there are some signs of running lean there.

XLR Tune 9A2-0139 HPTuners

The issue of KR due to multiplied IAT advance at lower temps could be why the XLR has consistently run better at high temps (which is counter-intuitive). In the next tune I will change the multipliers below 6000 RPM/below 104F for IAT advance to 0. The alternative would be to zero out IAT advance adders at cooler temps, but there may be some value in keeping them at higher RPMs.

Because the KR doesn’t go up at higher RPM it may be false knock?