I finally got an Innovate LC-1 and AFR gauge, and have been doing some no-load tuning with it. I think I'm making some progress, but I have a question about this:
Notice how my O2 (red on middle graph) dips in the middle of this throttle stab? I would guess that this relates to under-correction by EAE? If so, what should I change to help correct this? Would I just increase the stick-to-walls constant at the top end of the % of MAP range?
After further review... I think my setup is incomplete.
I am using the Analog output 1 for the MicroSquirt and Analog output 2 for the AFR gauge. It looks, though, like MicroSquirt expects to use some other calibration curve... the default for AO1 being very similar to a standard narrow band. I'll have to reprogram the LC-1 and recapture the data. We'll see what that looks like then.
Often with the quick stab, EAE (due to being MAP and VE table lookup based) cannot respond quickly enough. If you keep trying to tune the initial lean spike out, you will just get frustrated because it won't go away, and the behavior will match what you are seeing.
This is why I added the ability to use normal AE at the same time. You can use a bit of that to get EAE started enriching sooner, getting rid of the lean spike and stumble.
This behavior is also why I recommend tuning it using slow throttle movements in high gear without RPM changing much. This will get you the correct curves, then a little normal AE can be used to get rid of the quick stab lean spike.
Thanks for the advice. I do have the lag control in place.
I am fighting with a couple of things here. First of all, I've configured the LC-1 for the same output on 1 as on 2: 0 - 5V, which shows 7.4 at the bottom end and 21.7 at the top end on the Innovate AFR gauge. However, I can't seem to get MegaLogViewer to see this. The best I've gotten there is to set up MLV with a custom AFR field of 0 - 0.5V (seems to capture WB02 at V/10) and 7.4 to 21.7, which gives me a graph a bit leaner than I really am (compared to the gauge on output 2).
This seems close enough, though, to use MLV to see what's really going on. The part I'm more concerned about is behavior like what's in the attached log. Basically, if I stab the throttle once, hard, it will probably be OK, and maybe even decently responsive. If I stab it two or three times in a row, the engine just dies. After looking at the grapha again, I now see PW going to 0 after the stab, a second stab leaves me to too much time with PW=0 (too much "sucked from walls"?) and the engine fuel starves before EAE can switch compensation from enleanment to enrichment.
Thanks for the advice. I do have the lag control in place.
I am fighting with a couple of things here. First of all, I've configured the LC-1 for the same output on 1 as on 2: 0 - 5V, which shows 7.4 at the bottom end and 21.7 at the top end on the Innovate AFR gauge. However, I can't seem to get MegaLogViewer to see this. The best I've gotten there is to set up MLV with a custom AFR field of 0 - 0.5V (seems to capture WB02 at V/10) and 7.4 to 21.7, which gives me a graph a bit leaner than I really am (compared to the gauge on output 2).
This seems close enough, though, to use MLV to see what's really going on. The part I'm more concerned about is behavior like what's in the attached log. Basically, if I stab the throttle once, hard, it will probably be OK, and maybe even decently responsive. If I stab it two or three times in a row, the engine just dies. After looking at the grapha again, I now see PW going to 0 after the stab, a second stab leaves me to too much time with PW=0 (too much "sucked from walls"?) and the engine fuel starves before EAE can switch compensation from enleanment to enrichment.
Any suggestions here?
Thanks,
Paul
Did you calibrate AFR Tables (under tools)?
I don't see AFR data-logged.
Peter Florance
PF Tuning
81 BMW Euro 528i ESP Car
60-2 Wheel LS2 Coils, Low Z Inj
Co-Driver 1999 BMW E46 DSP car.
The behavior that's in the log on quick stabs is likely because you were trying to compensate for EAE lagging a bit on a quick stab.
If it pulls too much fuel when you lift after a stab, you probably need to decrease the amount it thinks is getting pulled from the walls.
A better idea would be to get it acting really well at all loads and RPMs on slow throttle movements. EAE cannot be properly tuned using the quick stab method.
After you get it working well for slower throttle movements, you can add a little standard AE to get it the bit of extra fuel it needs (not getting there due to MAP responding too late to get the fuel in).
Trying to tune the quick stab using only EAE will cause you to overcompensate and will likely cause issues like the stall one that you're seeing.
If AFR isn't being logged, you probably didn't set up for wideband when you were creating your project. (TS project settings or MT configurator).
muythaibxr wrote:
If AFR isn't being logged, you probably didn't set up for wideband when you were creating your project. (TS project settings or MT configurator).
Ken
I've seen this before but couldn't remember what caused it.
Peter Florance
PF Tuning
81 BMW Euro 528i ESP Car
60-2 Wheel LS2 Coils, Low Z Inj
Co-Driver 1999 BMW E46 DSP car.
You're right, this project was set up for narrow band. I just got the wide band last week. I had forgotten that that was one of the configurator settings. I will clone the project over so I have a wide band version I can work with during tuning.
Regarding the earlier response, how do I reduce the amount of fuel it thinks it's getting from the walls? Is that just reducing the "sucked from walls" constant at the low % kPa end of the range?
I'll take a look at the not-too-quick thread as well.
Yeah, if it's on lift and it's pulling too much fuel on lift, it thinks more fuel is coming from the walls than there is, so I'd reduce the sucked from walls coefficient or the RPM based part of that (if it's not doing it at all RPMs).
Hi. I'm back to this point in my project and have made some new progress. I am now on 3.1.0, but was still having the same bog or stall with a quick throttle stab.
I've got it basically solved now. As far as I can tell, there were two aspects to my problem:
1. Way too much fuel during accel enrichment. I found that by switching to pure TPSdot AE and playing with the curve there. I ended up needing between 1/4 and 1/3 of the stock pulse adders.
2. I was confused about how EAE works. I had originally tried the stock MS2 code on my bike, and had tried X-Tau. My thought was that X-Tau was from Ford and EAE from Toyota, but that they were both basically the same thing - yes and no. They both use the idea of compensation for puddling / wall-sticking to try to make actual fueling under dynamic conditions match the map, but they don't seem to do it the same way. When you enable X-Tau, normal MAPdot and TPSdot AE is disabled, and X-Tau does all your AE. I thought that that's what EAE was doing, which is why I wasn't very successful in getting it to stop flooding my engine: I didn't realize until switching EAE off, that EAE is (apparently) using the regular TPS/MAP AE curves to drive its AE as well.
So, the upshot is that, once I got the quick stab stuff tuned in with pure TPS, EAE works a lot better.
Hi.
Once you got "normal" AE working you should increase the TPS treshold so high that it is only triggered at heavy Throttle stomps. "Normal" transient fuel should be managed by EAE, then.
If the default EAEs won´t work probably have a look at another thread where I had posted my EAE settings.
I thing it would make sense to collect some EAE setups mo make it easier for user to find a good initial setting.
It´s maybe a bit hard to get EAE running but it´s worth it. Definitely.
Thomas
Ah. Great. Hidden settings. (The thresholds are outside of the normal window boundaries, at least on my machine, when opening the Acceleration Wizard. I was only seeing the top three graphs.)
I may play with those a bit, if only for the "cold" options. I'm not too inclined to change much else, though, as things are working pretty we as-is. It would be nice if (if we're expecting to blend EAE and regular AE) that the thresholds could be specified when enabling EAE, so that you don't have to make two or three changes to compare EAE to non-EAE behavior.
