Thank you freedom for reading what I read and not immediately going to "flame-mode"

- Steve

 

I think that the smoothness, or lack of in your case, is caused by not hitting the SMART fuel map until 41% as seen by the MAF as dictated by the SMART authority map. I had the SMART authority map set to 25% until Shiv posted not to do so, but the car seemed to drive very well as it was utilising more of the SMART fuel map and shifting over much earlier so that there was not a sudden jump or transition to the SMART fuel map. I wonder whay we cannot use the SMART fueling map at a lower MAF value? I have seen large changes in my A/F ratios when switching from the base map to the SMART fuel map when the MAF hits 41% flow. It is not as dramatic now that I have the base map set up a bit better, but you can still see it sometimes. I have considered moving the SMART authority map to 30% to see what happens.....Shiv, what is the danger in letting the SMART system have authority at lower load values?

This is my last log, which I think is pretty good....

196=2700 rpm=12.3 afr
235=3000 rpm=12.4
301=3500 rpm=11.75
306=3600 rpm=11.8
334=3900 rpm=11.47
356=4170 rpm=11.4
395=4560 rpm=11.3
443=5000 rpm=11.00
493=5500 rpm=10.78
638=6860 rpm=10.81
679=7170 rpm=10.41

I need to see how the AFR's that the wideband shows stack up to a dyno, but I think that I will start to aim for 11.1's or so from 4800 rpm on....

2nd pic shows noise floor, my engine is very noisey, 2.3, headwork, no balance shafts..... But when the rotating assembly was balanced it was done so with the front pieces i.e. dampner and pullies, along with the flywheel/pressure plate. To within 1/10th of a gram, still a lot of background noise though.

 

^^^ I'd like to know also. I just loaded a new map into my car tonight, and i lowered the fuel threshold 2 rows i think (30 or 35?) to get smart to help out a little more.

 

The LC-1's factory default is to output AFR measurements as fast as possible (e.g. instantaneous menu selection). Innovate claims that the LC-1 at its fastest rate can respond to individual exhaust events. It also can be configured to average over various periods of time (e.g. 1/12 of a second).

I remember one of Shiv's posts commenting on the wideband's update rate was adjusted to fit the SMART system. Should we be using the default single sample output or one of the averaging settings?

Also I think that you don't want SMART getting involved when the ECU is closed loop. This is why 40% was picked. Or at least that is my guess.

 

That makes sence, what parameters must be met for the car to go to open loop? I think that it is a certain MAF % or rpm, but I do not know what they are. I would think that you could have -100's where the fuel map would be in closed loop and 0's where it is going in to open (on both the x and y coordinates) so as to go immediatly to the SMART fuel once open loop is hit, but we would need a pretty accurate idea of when this is.

 

Normally load. As I recall it is somewhere in the 30's.

 

freedom,

As i discussed with you earlier, here is a datalog i had made of the MAF signal. I believe this may have something to do with some of the smoothness problems we are having.

This chart is of the difference between MAF In and MAF Out with SMART running. You can see that when SMART kicks in, there seems to be a +30% to almost -30% oscillation of the MAF signal. This seems extremely drastic to me. It appears that SMART is overcorrecting a value, then overcorrecting back, etc.

shiv

What do you make of this? has this been looked at in your testing? is it possible that the SMART correction values are too great, causing dramatic overcorrections??

 

When you guys changed the SMART threshold from 40 to 30, did you somehow modify the load points (the y axis)? Shiv has it set up so the cross over is almost instantenious. I think at 39 it reads -100 (smart off) and at 40 it's 0 (smart on). Lowering this threshold without adjusting load points will make smart engage gradually, which is bad because during that time you will have really rich readings in the smart fuel table. As a result SMART will make you run as leaner than you should. That could also cause big MAF oscillations.

 

I need to point out some results from tonights run on the road....

I started out with the template that only had smart timing / fuel and plugged in fuel, timing and boost maps that I was running up to the point I installed the SMART system. With smart fuel off this combination worked rather well.

When Shiv released the 10 map version of the template I decided to employ it with the same maps as the basis. From this starting point I have been tuning to the best of my ability. Clearly I was hitting the targets, but smoothness was always lacking.

With nowhere else to go I just tried the 10 map template with no changes, as well as with the 93 octane smart fuel table. To be honest the car runs much much better than the hybrids cited above. Notable is that I have the LC-1's integration time set to max at this juncture on the basis of smoothing out individual misfires, etc.

So assuming that with cooler weather in the morning everything is still runs well, then I would be ready to say that I'm happy with the SMART system.

 

Clearly I was seeing an underdamped situation was well. BUT my Shiv's 10 map set seems to not drive the system into this situation. (Also note that I slowed down changes from the LC-1.)

 

You can't, because the map parameters aren't editable in the currently available XMap v3.2.

 

The max corrections in the SMART Fuel table are only plus or minus a few percent. I'm not so sure why your seeing such huge MAF oscillations. There may be tiny oscillations. So small, in absolute terms, that you can't even see it on a wideband 02 trace. That should be about it though...

 

wtf

I went back and recorrected the values, it looks like the shift is more like +10 to -10, but that is still a pretty damn big percentage of the MAF signal. If you think about it, if the base fuel table takes -4%, then SMART gets overcorrected, it takes another -4%. that adds up quick.

How do you suggest i fix this?? should i turn smart off and fix the basemap first? should i lessen the correction factors of the smart table so it doesnt overcorrect? (with a good basemap, smart shouldnt have to make THAT much of a correction anyway)

Can anyone else analyze their MAF signal and see what thier logs look like?

 

That was prior to adjusting the smart threshold. Also, the load points for the SMART fuel threshold are the MAF signal. it turns on at 41. For my stock MR, the car only maxes out at a ~55 MAF, so 41 is a fair amount of load. Smart will not even turn on till ~4k RPM or so.

 

Close a few maps and hit F10 (instructions on last page in SMART Walk through). They're there and should pop up....