Lavitzlegend

iTrader: (3)

So I recently purchased a '93 VR base with the full Turblown EFR 7670 kit, Adaptronic ECU, ID1300 and ID1700X injectors, ETS front mount, and AEM water/meth setup. Previous owner did not feel comfortable/knowledgeable tuning so I got a fairly good deal on a nice car. I mostly have the base tune working like I want with the exception of still slowly removing fuel in the overly rich boost areas of the map (currently dips into low 10s and sometimes below). However, I'm struggling to understand exactly how I'm supposed to set up the two different MAP prediction tables to properly give the correct amount of fuel on tip in. Can anyone share their throttle sensitivity and MAP prediction tables to help guide me in the right direction? Or should I just book some dyno time and try to figure it out myself? Thanks.

mrselfdestruct1994

iTrader: (1)

Been a while since I've had the car running but I think these predicted map values were half decent. I estimated the values using a tps x rpm x map histogram using MegaLogviewer.

I haven't had much luck with the async injection.

Lavitzlegend

iTrader: (3)

I will take a look at this and see how it compares to what I have. Thank you for also explaining how you set it up in MegaLogViewer. I was trying to decide how I could get a good histogram but just to clarify you are setting the X axis as RPM, Y axis as TPS (or TPS dot?) and the values in the table are the MAP values the ECU is using or what is coming from the sensor? I'm confused what the MAP value in the cell is actually telling me and how I know what to change based on that histogram. Don't most people adjust the predicted MAP table live using the plot trace?

mrselfdestruct1994

iTrader: (1)

Histogram y axis should be tps and the z axis should be your map sensor reading. You're trying to fill the table with the typical map readings for a given rpm and tps position. It's essentially switching to alpha-n for x milliseconds using a map lookup table. The tpsdot sensitivity is best tuned by feel as you mentioned. Using the histogram method my values seemed quite close but you could also fine tune on the dyno.

chuyler1

iTrader: (1)

Lol, i literally just did the same thing the other day to start fresh with my transient table. Everyone said predicted map wouldn't work with ITBs but I got it spot on now. For my Select 440d (might be different for Modular) I came up with the following for a street port 12a with a shorty intake manifold and 45DCOE throttlebody.



To arrive at these numbers, get a very long street log, in my case and hour commute to work with varying speeds between idling in traffic, cruising between 30-45 to highway cruising up to 85, keeping speeds constant or slowly accelerating decelerating but switching up gears. Add in one or two WOT pulls for good measure.

Once you have the log in MegaLogViewer, switch to the histogram and choose RPM for X-axis, TPS% for Y-axis, and MAP for the cell value. Click on each axis cell to enter a new value, and plug in the exact values from the predictive map table. 1000/2000/4000/6000/8000, 0/2/5/15/50. Make sure you filter out RPM changes and Transients. And voila, you have your predictive map values. Copy them over to the Wari table. The only one I changed was 1000RPM / 100% throttle where I didn't have a good value so I added 20% to what it likely would be to ensure I could stomp the throttle at idle.

From there, if your car runs like crap, don't mess with the map table, mess with the other settings. Transition time is how long it will use the predictive map values, if your MAP sensor is responsive, you can keep this low, for boost, you may need to set it longer to account for spooling. Gain and persistence will require going back to the log and looking at individual throttle pumps. Set gain high until you at least see it start to go rich on throttle application, then adjust persistence so it stays rich, then lower gain until you have the desired AFRs during the entire transition time.

I hope it goes without saying that you first need your RPM/MAP fuel table spot on before you start this. When it's predicting map, it's basically jumping to the cell you will end up at eventually, if that cell isn't right or requires drastically different closed loop trim levels, the math and logic isn't going to work.

Lavitzlegend

iTrader: (3)

I guess I should have started by saying I'm using Eugene. So the predicted MAP GUI and variables are a little different than what you are showing although it obviously works similarly. However, I've been fighting some strange electrical gremlins that I think are coming from an alternator on its' way out. Probably a good opportunity to have it rebuilt and cranked up to 160 amps output to cover the added electronic load (water pump, AEM water/meth pump, etc) being put on the smaller AGM battery. It's doing this strange thing randomly where the timing goes flat at 15 deg even though the part of the ignition map it's in should be around 10 and half the engine stops working. I can also see the voltage dropping quite a bit in that part of the log as well which was leading me to think alternator.

chuyler1

iTrader: (1)

Check all the other sensor inputs in the log for interference, temps, TPS, crank angle/RPM, etc. and of course check the harness for loose connectors or damaged wires. A voltage drop shouldn't cause the ECU to mess with timing.