VE table tuning
 

My setup is:
2001 head/block/intake
1997 fueling system
1997 sensors
1997 body
1997 alt
no A/C or P/S
949 twin disk clutch/flywheel
long tube headers
JR intake w/ MAF still installed. (Going to be replaced with a cone filter and some silicon intake pipe once I'm confident the MS3 won't strand me while tuning) A GM IAT is plugged in when the MS3 is installed.
LC-1
MS3 batch injection and wasted spark.

My issues are:
Originally when I installed the MS3 it fired right up but idled poorly and osculated from 850-990rpm, any throttle input would kill the motor. I got the idle to stabilize by turning the idle air screw until the idle RPM was right around 1k, it idled solidly there and minor throttle input caused the engine to rev. Since this isn't a street car having a highish idle isn't an issue. Not sure if this should even really be considered an issue. The original surging may have been caused by the lean condition that was present before VE analyze did it's thing.

VE analyze seems to be making drastic changes to the VE table. The cells that the engine has run in are 10-20 points higher than where they started. Most of the maps that I've seen floating around have the idle blocks in the ~40 range. Should I be adjusting req_fuel so the values for those blocks fall down into the ~40s? or is this a symptom that something else is jacked up? I set req_fuel by putting in the displacement and rated CC for the 1997 injectors and let tunerstudio calculate the value (11.6).

Any kind of significant throttle input kills the motor. I'm thinking that this is due to the drastic VE changes and the motor is trying to transition to a cell that is a much lower VE so the MS is leaning the motor which causes it to lug and will go away on it's own once the VE table is dialed in.

Theories:
I'm thinking the control for the idle air valve isn't working properly. I had an incident with the harness wiring and I may have cooked the TIP120 transistor that controls the PWM valve. Is there any real need to control the idle air valve if you don't have A/C? Idling at 1k the engine seems to tolerate me flipping the lights on straight to high beams. That would also explain the VE tables since extra air could be getting through the valve.

I'm not so much looking for a step by step guide but just some validation from those who have done this that I'm at least vaguely on the right path.

I hit 11pm last night so I shut the car down for the sake of not pissing off the neighbors. I'll be able to get it fired up and test some more things out later on today. First thing I'm going to attempt is unplugging the idle air valve to see if that causes everything to go rich.

As an interesting side note, on my 1997 C package (I think, power mirrors, leather and torsen) the AC and main fain pins are swapped.

Attachment 241580
Attachment 241581

Change Req_fuel, change the VE table, they both have the same effect. do either, but note if you move req_fuel from the 'correct' calculated value if you ever changeyour injectors you'll be unable to scale your VE table to the correct value easily.

Everything else is a symptom of a poorly tuned VE table. Instead of blindly letting TunerStudio do it's job take a log and make sure you are actually hitting the desired AFR's.

If you have AE enabled turn it off during initial VE tuning. Which TS will filter out AE events, this alos means the parts of the VE table you're in whilst this filter is on will also be fitered meaning you'll never get shrottle stabs tuned properly.

A dead IAC is a possibility, easy to check though, load a guage with Idle DC and pull the IAC plug, do the revs drop? does the ECU try to respond by increasing IAC DC? If yes, yes then software and hardware are working, no, no means neither are etc.

To disable AE I should just set the low threshold higher than the motor spins? or should I change the values in the "Accel Enrichment Wizzard" to 0?

I've been keeping an eye on AFRs, they are making it to the target AFR given enough time in that cell. The motor also runs significantly better once they get down to the target value.

Would it be worth richening up the whole table? Things do get pretty lean in the untuned cells.

Might as well use this thread for general MS3 progress on my car. If a mod wants to make the title more appropriate that works for me.

I've been tuning the VE table for a few days now. Hit some snags with an oil/fuel leak that kept me off the road for about a week. So far so good, AFRs stay somewhat close to the targets when I'm driving around. I have a few outstanding issues I'm at a bit of a loss on.

1. I haven't built the tach circuit yet I have the recommended schematic that calls for 2 1K resistors, a 2n2222 and a resistor of unknown resistance. Does anyone happen to know what resistor should be used there to make a 1997 dash cluster happy? Also where is a good place on the DIY board to pull +12v? If they are in range and no one knows a useable value can the variable resistors that are included with the DIY board handle the current of this circuit?

I typed this up and realized I don't have the PC with my MSQ available right now. I'll type up the issues that aren't wiring when I get it back.

you need two 1Ks. S19 for 12v.

Like this, correct?

Attachment 241562

yes

Should I be seeing a 10v high 7v low signal out of this circuit? The tach doesn't seem to trigger off of that but it will trigger on the 0v -> 7v transition when the MS is first powered up.

I couldn't find S19 on my board so I used S12 instead. Could that be the issue? S12 seemed to be the only post-filter spot for +12v.

http://www.megamanual.com/ms2/v3pcb_6.gif

After looking at that thread with a bunch of AFR maps in it I'm thinking that my bogging at rapid throttle input from idle might be caused by running too rich in the idle cells. Any ideas? It's pretty quick to retune those cells so I might give it a shot tonight and see what happens.

Figured out my problem. I used JS0 (IAC1A) for the output pin, looks like he outputs 0-.3V. Moved the circuit to JS11 who outputs 0-5V and everything works fine now.

Good thing there was a scope I could borrow from work or I'd be totally lost getting this together.


[edit]
Changing my idle AFR targets up to 14.0 and tuning the VE so that was the case got rid of my idle bog. Now I just have a delay when I drop the hammer. Going to mess with AE to see if that fixes it. It really messes up my heel-toe. Almost makes the car feel like it has really shitty DBW

You will need some AE to get good response.

Once you get more experience tuning, you can use EAE which will give OEM smoothness for normal driving, and you can add a little bit of TPS AE for quick blips and shifts to get quick response.

Ken

you need to jump js9 with 12v as well in order for js0-4 to work properly with an MSII or MSIII


FWIW, I never use closed loop at idle.

Since this became my generic MS3 question thread. I'm ready to convert the car from CAS to NB style sensors. Do I have to do anything other than attach the new sensors? I'm 99% sure that I don't have to modify the MS3 but I want to be sure before I make a working car non-working.

I know that I'm going to have to re-check the timing after I complete the conversion and change the wheel decoder to 99-00 but I think that's it right?

Got it running on the 99+ crank and cas. Just had to flip the trigger edge.

fans still reversed?

Yea. The only wiring I touched was the cas plug. Snipped it off and spliced in pigtails for the cam and crank sensors.

Main is 1A, A/C is 1B. if you used my diagrams, i have them labeled wrong.