More timing than my advance table... lean spot

[DSteck]

I am trying to figure out why the VCM is commanding more timing than my main spark advance table in a particular spot.



The lean spot at 4000rpm is what I'm talking about. It happens every run, in the same spot, and I can't figure out why. My main advance table only commands ~26 degrees, yet the VCM is doing 29º. The TPS reading dips, knock retard goes into effect, and there's a lean spot in the reading. I am really struggling to figure this one out. Can anybody help? Attached is a log of my drive home. Around frame 835 you can see a 2nd gear pull starting from 3500rpm. I just don't understand why everything gets really screwy around there.

Yes, I realize my AFR needs to be cleaned up under 4500rpm. I ran out of time Monday night.

[Dozer]

ETC, IAT, AFR, Cat Light off, adder and everything else looks good you got me. have you tried a pull with m6 spark smoothing off?

[BBA]

Your PE table is F'd...looks like stock settings. The AFR at RPM is directly proportional to the inverse of the EQ ration vs RPM table. For beter mileage I'd also recommend setting the PE MAP value to something higher than your normal cruise MAP level but lower than the highest MAP level. Somewhere in the 70-80 kpa range works good.

[DSteck]

Any idea why it's adding too much timing? It's three degrees beyond what I'm commanding in main spark advance.

I know that I need to fix the ratios in the PE table, but like I said... I can only use free dyno time for so long! My friend wanted to get home to his family... can't argue with that. I will change the MAP value for PE enable. I never see the PE Enable bit set while cruising or accelerating though.

[Dozer]

At 3973rps just before the knock your getting 29* of timing and commanding 24* so i see 5* more than commanded on you timing table. But I still cant find from where?

[DSteck]

At frame 174 3973rps just before the knock your getting 29* of timing and commanding 24* so i see 5* more than commanded on you timing table. But I still cant find from where?

Me neither, and it's freaking me out. The other weird thing is you can watch the injector pulse width take a sharp stab upwards. I don't get why the ETS TPS % drops to 87%, either. The knock retard that shows up is a result of burst knock retard, and not actual knock, so something is making the vehicle think there was a spike in air flow I would assume. Could this be torque management kicking in for some reason?

[Dozer]

ETC traction control is still enabled along with Spark.

[DSteck]

This is true, but on the dyno, the car did the same thing... and we had completely turned off traction control and active handling.

[Dozer]

A friend of mine told me that some of the TBSS were having problems with commanded spark from actual due to setting multipliers to 0 instead of 1.

[DSteck]

I found it. It's a discrepancy in the way the scanner records cylinder air mass in the Main Spark Advance table for logging.

Due to my VE table being totally whacked and not dialed in whatsoever for my cam, and my dynamic airflow disable set to 4000rpm, there was a VE correction against my MAF flow until 4000rpm. This resulted in the dynamic airlow being roughly 75% lower than the MAF reported airflow. The scanner table reports cylinder air mass based on a calculation done with the MAF... however, the VCM uses a cylinder air mass calculation based on dynamic airflow (you can see dynamic air flow match up to MAF air flow after 4000rpm... which is where it switches to purely filtered MAF and no modifiers).

In the picture, where I have it highlighted, my tune would appear to command 24º advance (1.04 and 1.08 cells at 4000rpm are 24º). However, if you perform the cylinder air mass calculation at that point in the chart using dynamic air flow, you get a cylinder air mass of 0.75, which... according to my spark table... would be 29º advance. At the point in the chart, as you can see, the VCM commands 29º advance.

As a result of this, I rebuilt my scanner to calculate dynamic cylinder air mass, and I use this as the y-axis of ANY histogram referencing cylinder air mass so that it's an accurate representation of where the VCM thinks the car is operating.

The fix to the lean spot is to set my dynamic air flow disable RPM to 600 or so (something under my idle), and set the re-enable to 500. This way, the car will always operate with a filtered MAF instead of referencing a VE table that is inaccurate.

The burst knock retard was a result of the ramped up dynamic air flow at the 4000rpm switch.

[ShoddyHog]

Just curious...why don't you tune your VE table? What happens if your MAF ever fails?

Gerry

[DSteck]

a) I don't have a wideband to log along with my car (didn't buy the Pro version yet either).
b) Tuning the VE table in the E38 is ridiculous, despite Bluecat making a working assist tool.
c) If my MAF fails, then I shut the car off, have it towed home, and fix it. It's not a daily driver by any means, so it's no big deal if it's down for repair. The chances of the MAF failing are pretty low... I'd imagine a MAF on one of GM's most expensive production performance vehicles with only 18k miles is probably not going to fail any time soon if it hasn't already. I have faith in GM.

Lots of people run MAF only. Once the 2.5bar SD operating system comes out for the E38, I'll build an SD tune. For now, I'm nowhere close to maxing out my MAF (although I am approaching the 512 g/sec limit... once I port the heads, I'm pretty sure I'll have to start fudging the PE table).

[BBA]

You do realize you need the VE table for under 4000 RPM VE-MAF dependancy right?

You should also know it does not require a WIdeBand O2 sensor to set VE tables. I've done it a few times with narrow band, it's easy and always has good results. I even posted a guide on how to do it here.

[DSteck]

You do realize you need the VE table for under 4000 RPM VE-MAF dependancy right?

You should also know it does not require a WIdeBand O2 sensor to set VE tables. I've done it a few times with narrow band, it's easy and always has good results. I even posted a guide on how to do it here.

Read my post right before ShoddyHog. I switched to pure MAF.

For as far out as my VE table is, I'd be better off with a wideband. My VE air flow registers about 150% higher than my MAF airflow until going WOT.

[BBA]

I hate to be the one to tell you, going pure MAF loses performance. Why do you think people go Speed Density so often? Once you mod to a certain point, MAF becomes more of a problem than a cure. Especially if you exceed the built in MAF limits or go past the PCM's configured tolerances.

You need to do a VE tune, even if you want to go pure MAF, which I don't think is even completely possible.

[DSteck]

Engine -> Airflow -> Dynamic Airflow

High RPM Disable set to 600rpm
Low RPM Re-Enable set to 500rpm

It's not pure MAF, it's just filtered MAF like what is used over 4000rpm anyway. If you log MAF air flow against dynamic air flow, the two are now VERY close to each other 100% of the time... the only difference being that dynamic air flow doesn't show the oscillations of the MAF air flow as much under steady state or loaded conditions.

The car drives a lot better, and there's no cam surge (granted, I played with my timing map to accomplish this by pulling a lot of timing in the low rpm and low cylinder air mass area). At WOT, the lean spot is gone, and the car runs much more predictable. Also, in switching my disable rpm to 600, my throttle has instantaneous reponse now, whereas before it had a hard delay when revving from idle.

My car isn't maxing out the MAF... It's within about 10% of the 512 g/sec peak, which occurs at 9800 Hz at just over 7000rpm.

When the 2.5bar operating system is out, I'm going to switch to that.

Now... I had a problem with the scanner reporting more timing than my advance table had, and had a lean spot along with a weird injector duty cycle trend around that lean spot. No offense, but you didn't help with either, yet I fixed both problems and the car operates very smooth now. The problem wasn't in the PE table, and changing the MAP enable for PE didn't change anything, as my car still doesn't enter PE mode on the highway, and made no change to the calculated fuel economy onboard (even though the PE enable bit was never set while cruising or normal driving under either setting). With that said... what's wrong with the way I did it? There are a ton of people who don't touch the VE table in the Z06, and still manage to run 9s and 10s consistently while maintaining a drivable car.

http://www.hptuners.com/forum/showth...mic#post121675

I attached the latest tune for anybody curious to see. I filled the tank, drove the car like this for a while on a 65mph highway cruise, and filled the tank again. The car got an actual fuel economy of 26.5mpg on the highway. The best this thing ever did on the same stretch when stock was just under 30, so it's plenty good on fuel economy for me.

[BBA]

Well, if it works for you then I am glad you figured out how to get around your problems.

Sounds like your running under the TPS enable setting when cruising, which makes sense that your not running enriched when crusing. You are running TPS controlled PE enable instead of engine load controlled, I just don't think throttle controlled PE is better than vacuum controlled PE but I guess I'm old fashioned that way (head stuck in carbs).

[WarWagon]

a)
c) If my MAF fails, then I shut the car off, have it towed home, and fix it.

You won't have to turn the car off... it will likely stall with a MAF failure.

You would think a 1985 Olds 98, one of GM's top of the line cars, would have a good MAF too. Yet mine failed and flooded the engine to a stall while taking a right turn fast. The 98 is a FWD car for 1985. The stalling engine drag wanted to kick back the steering wheel and no power steering put me into the door with all my strength on the steering wheel to finish the turn and not plow through a pickup and trailer instead. I clearly recall screaming while trying to rip the steering wheel around because it was that close to a bad wreck!

Turns out just tapping the bad MAF would stall the engine with black smoke out the exhaust. The 1985 primitive computer could not figure out the MAF failure. Also it was from a known bad reliability MAF sensor production run. (It failed at 80K, 10 years after production.)

I disagree with your trivial assessment of a MAF failure and you can see why. I also really hate FWD for some reason...

[DSteck]

You're also speaking to an engineer that came from the Chrysler product development world. Parts 20+ years later are a little different. Production techniques have made leaps and bounds in that time, namely in the electronics world, bringing increased reliability and durability. In the event of a failure, it's probably not going to just flash open the injectors. Also, the VE table isn't totally disabled... it's just that corrections against my MAF while the MAF is operating are disabled. On top of that, I have good insurance. I really don't care.

[WarWagon]

Actually you are 100% right in the fact that your changes will not affect the outcome of a MAF failure. Kudos for the tow it home safe attitude rather than try and limp it home. In this MAF failure case I am talking about the sort of working but readings are out there. Like the failure in the 85 olds: cracked internal connection that would read high airflow when there wasn’t. (The MAF is not stone cold dead.)

You can prove that the 'modern' engine will stall with a bad MAF just by repeating a mistake my tuner made one day. In the MAF Airflow vs. Output Frequency take the lowest MAF reading number, copy and paste the same number in the entire table. (Instead of multiply by % like he intended to do.) This will show the effects of a MAF stuck at a low reading: shifts hard, stalls the engine, and throw a pending MAF out of range code. The engine stalls before the code is set. This was done on a 2006 TBSS.

I agree reliability has come a long way. However the modern computer, like it’s primitive 20+ year old pioneer, is still too dependant on the MAF working within spec or not at all. If you go to the GM parts counter I bet they have a MAF in stock...Perhaps from all the K&N oil?