Shout out to "jpnovak" and "Lapkritis" but an MS Enhanced Accel Enrichment question
 

Without "jpnovak", I would have been miles behind with getting my car to run and tuned correctly. He has endured many of my questions and at times was probably pulling his hair out on some of the more elementary questions.... :)

The icing on the cake from Mr. Mad Scientist was a plaguing sync error that caused the car to stall at lights and during cold starts right after light off. One simple turn of a resistor pot cleaned up the crank trigger signal and it now starts beautifully down to ~20 degrees! 5 minutes with an oscilloscope made all the difference in the world.

Also too, "Lapkritis" made a simple comment that totally transformed my car. Simply put, he made the comment that you have to run less injections per cycle with big injectors or the small pulsewidths will be hard to tune. Boy was he right! I spent a couple of years trying to adjust things that just didn't make sense. Now, with the correct settings, the car runs really smooth in all loads and ranges.

But, here's my predicament. Running 2 injections per cycle vs 6 I had before took away my nicely, crisp throttle response. All other areas run nicely, but it has made a lean tip-in that's been a bear to get right.........which leads me to Enhanced Acceleration Enrichment.

With "lag compensation" on with EAE, I'm getting a hard fuel cut ant mid-speed throttle transitions that makes the car buck and you see dead air in the WB02.

The lag compensation will go to 6 squirts a cycle from the 2 I have for clean throttle response, but now, it doesn't like it at all.

Any tips where to start? I had it very nice with 6 injections per cycle running without the lag compensation, but now with it, the settings are way off.

I read you only change the curves, not the min/max values. Was successful before, but not now.

I haven't a clue where to start.

BTW, my TPS is 100% (no MAP mixed in) for Acceleration Enrichment and I have it 200% rate of change.

TIA:):):)

Glad to hear you're seeing improvements. Working with the near infinite settings in MS can be daunting and difficult to learn manipulations.

EAE is a setting I would not enable. It takes a relatively simple control and complicates it. I would recommend disabling it and trying the settings that I posted in the previous thread. Throttle response is crisp, there's no bucking and the complications of EAE tuning are eliminated.

The success here will be a measure of how close your VE/fuel table is to optimum. Acceleration enrichments should be one of the final tuning settings enabled to polish up the drive.

Corey, as point of refrence from stock 3.2L tune I'll share that above 4000RPMs the stock engines have very little Accel Enrichment.
Here's why, AE is needed to replace wall fuel at the back side of the intake valve and port area. Turns out that at higher RPMs (>4000RPMs) you don't have much if any fuel build up on the walls of the intake near the valve so you have nothing to replace or compensate for when throttle angle changes.

AE was very much needed in the days of carburated manifolds that where entirely wet with fuel throughout the manifold. So many tuners simply assume that AE is essential in EFI but it's not as significant. I've seen EFI tunes that simply had way to much AE and my suggestion is to first get AE dialed in above 4000RPMs then work down in RPM.

Yes sir, it runs like a champ. My VE is pretty tight as jpnovak and I did more tuning with his method.

But, the throttle response was unbelievable before with EAE. I had it dialed in very nice where I could goose the throttle and the AFR's would barely change.

One thing I couldn't right was the decel. It'd still go a tad rich, but throttle response was superb, so I didn't mind.

Now trying EAE yields a PITA. I'm totally determined to get it though!!

Yeah Sal, MegaSquirt has a setting that you can program the slope. Basically, using default settings, from idle to 2,500 RPM, AE is the same rate. From 2,500 to 5,000 RPM, it's a linear slope decreasing until no acceleration enrichment above 5,000.

I'll go in and put in 4,000. Thanks!

And shout out to you too Sal! You've been a huge help!

Cory, One thing we have not tried in the car is to switch to a MAP based AE. I wonder if the close proximity of the injectors to the head and the large separated distance of the throttle butterfly is part of the cause. I suspect the time constants for fuel delivery might be off.

You previously described that TPS based AE should be faster than the MAP based. However, this may not be well matched to your particular build.

I have almost always used MAP based AE in the cars I have setup.

btw, we need to get that car on the road early one weekend morning and do some more work on the VE table. I still think we can solve some of those low thottle, mid rpm areas where you are experiencing the issues.

I admit the car drives really well.

EAE is a bear to get right. If your ultimate goal was the absolute best gas mileage it might be worth it but I would stick to a simpler method.

I did a little mixing without any improvement as far as I could tell, but I didn't lower the rate of change that low for MAP thinking back now.

I'll try a little this week.

I'll also try 100% MAP for data's sake.

I'm running 200% rate of change TPS and the lean tip in pretty much gone, I'll get a false trigger when normally cruising.

EAE allowed far less rate of change of both TPS and MAP when had 6 inj/cycle.

Hey Steven, thanks for your help too!

I want EAE because with 6 injections a cycle, throttle response and fairly flat AFR's were great.

Rereading EAE, I read again since I had forgotten that EAE with lag compensation will turn the 2 injections to 6 so the throttle response is better. But, I'd get a weird, nasty lean spike.

I think I had the added and sucked curves fighting each other. Put default settings back in, and it doesn't like that either.

Like I said earlier, MSExtra (I think) states to only change the curves and rate of change values, not the min/max of compensation %. That's what I did before and it was great.

Guess I'm getting OCD about tuning it perfect.

Another thing I've yet to nail down is nice AE the 1st 10 seconds the engines running.

I want it to be OEM, where it lights off, and I can give it any throttle and it doesn't have a lean tip in. I'm not talking, lean spike, I get a severe hesitation. Not good for someone expecting it to be OEM.

At 200% rate of change TPS, cold multiplier of around 250%, and cold PW adder of 6.0ms, I still get a lean tip in for the first few seconds after light off.

May try more after start enrichment and more cycles I guess........OCD, I know.

:)

I have one ms car here that lights off better cold than a couple factory ecu Bosch Motronic cars... they have noticeably soggy throttle response and cold decel is quite poor if you pay close attention. Tuning snobbery is when you start grading how well the guys at the factory did and give them poor marks! ;)

I prefer tps base ae for normally aspirated and a blend with map for forced induction. The length of line from the manifold to your map sensor can cause some lag on pure map as well as turbulence in the sign if you have individual throttle bodies or certain manifold designs. This can make tuning difficult... I match my fueling algorithms following the same concepts.

Why do you prefer blended for forced induction? Curious the logic. I'll try again but with a numerically smaller rate of change.

MS claims, I believe in the MegaManual, that hose length shouldn't matter until like 100ft; something to do with how light the air is inside the hose making it instantaneously quick even in long distances. I dunno.

Love it.....tuning snobbery.....:)

The blend is good for forced induction so you have the throttle stab addressed. One must also consider that quite often the most rapid accel / ae is at the boost threshold when tps has been constant and the turbocharger rapidly builds additional, maximum load. If you're tps ae only in those cases then you may be rough on your hardware... and find broken rings/pistons.

Ah, I see. I set my lags to 95% vs the default 50% including MAP so it should respond immediately to split-second rises in boost.