One change for the Cadillac Northstar in the LH2 variant was the addition of variable valve timing.  Today’s test is a slight change in the VVT timing on the intake and exhaust to see what impact on our test results.  I also restored the IAT retard at higher temps (no impact today).

This table shows the base VVT timing.  I am focused on 5200-6800 rpm, around 0.56-0.60 cylinder airmass.

These tables show the changes for Tune 6a.  The intake cam is degrees of advance, and the exhaust cam is degrees of retard.  For Tune 6a I retarded the intake cam from 5600-6800 rpm and ALSO retarded the exhaust cam in the same region in order to keep the overlap the same.

For this test we are focused on 5600-6800 RPM.   Different days, different conditions, so I see that today’s test (BLUE) is healthier at lower rpm as well — ahead by 5 whp at 5200 rpm.

For this chart I have zoomed in to just the 5300-6600 range.  Today’s tune (BLUE) appears to do much better at 6000 rpm, (9 whp at 5900-6000 rpm) although less better at 6600 rpm (5 whp)

Looking at the timing tables, I kept the retard on the exhaust at 4 degrees across even though the retard on the intake cam could not be that high (because it was already at 0).

For Tune 6b I will reduce the exhaust retard to match the intake retard in each area — that was my intent for the test.

This is the HPTuners VCM Scanner shot at peak calculated hp of 312.5 hp at 6638 RPM.

Timing is a healthy 29 degrees of advance including 2.8 degrees of IAT advance.

Less Timing?

I have been thinking about the overall test results I am seeing — and about how the XLR seems to get better results hot than cold (adjusted for temperature).  I suspect the LH2 would actually like LESS timing instead of more.

In the main spark table (high octane), the OEM calibration lowers the timing as the cylinder airmass increases.  This trend occurs up to the cylinder airmass at WOT — 0.56 to 0.64 — where it has gone flat for the rest of the table.  At high temps the engine gets retard from IAT and ETC which pulls the timing down.  At lower temps it gets advance from IAT which pushes the timing up.  I wonder if it wants less timing.

Next test:  match the exhaust cam retard to the intake cam retard exactly.

UPDATE:

I adjusted the VVT timing to a) start intake at 5200 and b) sync up exhaust with intake better.

In this chart the update run is in GREEN, the earlier today run in BLUE and the comparison run Tune 5D in RED.  What we see is close match between blue and green EXCEPT in the 4200-5800 range where starting the retard of the intake cam early picked up 5 whp in that range.

This is the same chart zoomed in.  Around 5700 the 6B change is almost 9 whp above the 6A tune. At 5200 we are still at high torque so approx 0.63 airmass.  The OEM intake advance was 12-11-8-3-2-2 from 5200-5600-6000-6400-6800-7200 at 0.64.  My test 6B is 8-7-4-0-0-0 intake advance.   It may actually want to stay at 7 degrees from 5600 to redline?

The second part of the test, trimming the exhaust to ensure overlap remained the same and try to get back to upward slope at the end of the red run into the green run.  This seems to have worked but we’ll see how it looks in future tests.

Update 2:

Started test of tune 6D which expanded the VVT mods down 400 rpm and up through 6800 RPM.

The yellow line is the 6C run.  Comparing the 6C and 6B curves, the 6C appears to start earlier, which is what we wanted.  However, it does not hold up from 6000-6800 RPM.  For the next Tune, 6D, I will return to 6B above 6000 RPM.

This run had another interesting incidence of freezing the VCM scanner, and then tilting the XLR.  I ended up in limp home mode, where I was able to use a code reader to reset the DTCs, after which the HPTuners VCM scanner was able to work as normal.  The test run for 6C may have been effected, but seems to still have been strong.  I am not sure what is causing these interrupts, but they are exciting.

I am glad to see 6C repeat the strong hump from 6B.  I plan to mix 6C below 5500 and 6B above for 6D.

The default XLR programming includes 0.5 AFR richer tuning at 6000 RPM to cool the pistons.  Ideally there would be a timer associated with it — say after 10 min at 6000 RPM it added — but there is not.  For this tune I removed that variable in order to study the impact from 6000-6400.  Since  there were few other changes in the tune we are focused on the 6000-6400 RPM range.

Specific Changes in Tune 5D:

• raise piston protection rpm to 8192
• reduce IAT retard
• point advance reductions around 5400 rpm to eliminate slight KR

The blue line is today’s test.  The red line is Tune 5B in hotter weather.  From 6000 to 6500 rpm the Blue line goes up 4 whp from 224 whp to 228 whp instead of staying flat as in prior tests.

The Virtual Dyno test results always seem muted in cooler air, which is counter intuitive.  I wonder if there are other factors such as the air pressure in the tires (reduced in cooler weather) which are impacting the test results.

The calculated hp in HP VCM Scanner showed a peak of 316 hp.

The calculated peak torque was 308 lb ft.

Overall I was glad to see a positive impact from 6000-6800 rpm and no knock retard.  I am uncertain why cooler weather tests seem to record lower results.  My intuition is there is a variable I am not controlling for.

Today’s road test is a repeat of the previous test to add to the baseline for Tune 5.  The only change from previous was to decrease the TUTD shift speed by 1 mph to 81, and to change the PE ramp from 0.1 to 1.0 (range 0.01 to 2).  This value effects how quickly the engine attempts to reach the PE fueling value.  Since the commanded air fuel ratio is now set constant from low RPM to high, we would expect this to have impact only just after wide open throttle is reached.

Today’s run is in Blue, and we are comparing to the Tune 5 run in red.  What I see is the runs are very close.  Today’s somewhat better torque may be an (additional) effect of cooler air, or it may be helped by the PE ramp up.  We’ll see how other runs look — certainly I don’t see any disadvantage from the higher PE ramp.

The calculated HP peak was at 6554 RPM.  ECT 194F, ambient 88F, Intake air 90F, MAT 115F.  Some IAT retard.  Peak advance ~26.5 degrees.

Madtuner.com offers a tune for the LH2; they did work on a 2005 STS in 2012 and showed this dyno chart for before / after:

This looks very similar to what I am seeing with the LH2 in my XLR.  They had some glitch in the base tune, but I am looking at the shape of the before & after curves — the HP peaks at 5400 or so and is flat until 6000.  They stop the scan at 6000 rpm for some reason — I suppose because the engine wasn’t making more power past that.  They don’t mention the specifics of their tune.

Cadillac XLR 320 hp picks up 12 whp with Magnaflow catback

The Magnaflow performance exhaust for the Cadillac XLR shows a similar shape in the before and after — HP peaking near 5400 RPM and dropping to past 6000 rpm.

320 hp LH2 Rear-wheel Drive Variant

GM powertrain showed the LH2 continuing to improve hp until 6400 rpm.  Redline is at 6700 rpm.  Note that the GM plot is an engine dyno, and the Madtuners and Virtual Dyno plots are hp at the rear wheels, so we expect them to be different due to drivetrain, friction,  and aero losses.  The part I am considering is the shape of the curve — the engine dyno showed that the LH2 makes a bit more power from 5500-6400 RPM.  To be fair, it is not on the same up-slope as 1000-5500 RPM.

Today’s test added to the Tune 5 baseline.  The Madtuner result is supportive that there is more to be gained here somewhere.  The Madtuner experience closely mirrors my results and varies from the engine dyno.  More testing to follow.

If you have ideas, suggestions, encouragement, or advice please hit Reply!