rich misfire? lean afr reading

[kaeman]

I believe I am getting a rich misfire when I stab the throttle during a launch. I am noticing that the fuel pulse width is large and the accel enrichment is about 40% at the point where the afr is reading 22.3 and then later it rebounds down to 11 and then back to 13's all the time while the pw is getting smaller. do I need to lower the top end of my accel enrichment curve? I know I need to recalibrate my tps because it thinks wide open throttle is 135%.....
see the attached datalog and msq.

2014-04-16 code 1.3.0 first tune.msq

2014-04-20_12_modified.msl

[billr]

I'm not of much help here, I have no experience tuning the AE. I did notice, however, that there is a drastic rpm spike right as you pop the throttle open, it spikes up towards 12K rpm. I think that is meaningful, and probably related to the crank trigger. Maybe logging sync-loss and such would help. The VE table also looks strange to me, but if you are running ITBs with specialized intake runners I am way out of my element there, too.

[kaeman]

I'm not of much help here, I have no experience tuning the AE. I did notice, however, that there is a drastic rpm spike right as you pop the throttle open, it spikes up towards 12K rpm. I think that is meaningful, and probably related to the crank trigger. Maybe logging sync-loss and such would help. The VE table also looks strange to me, but if you are running ITBs with specialized intake runners I am way out of my element there, too.

yes its itb's, which is why the odd ve table, the best I can figure is that the rpm spike is the flash stall of the converter when I hit the throttle hard and the megasquirt is reading the rapid rpm climb and miscalculating the rpm spike, the engine and tach never see that much of an rpm jump.... I am not running crank trigger or anything yet, just a fuel only for the moment.

[Matt Cramer]

You will need to get that RPM spike issue fixed; that's definitely going to mess up the fueling. Some more aggressive noise filtering may help; where exactly are you triggering off of? A tooth log of that spot will definitely help (I don't believe fuel only ever declares a sync loss, so the sync error logger will not help much here).

You also have the battery voltage starting to fall off beginning at record 309.

[kaeman]

yes I am running fuel only, triggered off the distributor lead for the neg side of the coil. I noticed the battery voltage, the alternator failed, at the end of that launch, I am replacing the alternator tomorrow. what kind of filter setting should I be using for distributor/coil input. will negative coil filtering get rid of the spike. I think that is the spike created by the flash of the convertor stall, but the engine never revs up in the 12,000 rpm. the tack in the vehicle only bounces up to about 4500 when that spike occurs. I am just figuring it(my ms3 ecu) is looking at the rate of change and miscalculating the peak rpm...... the rev limiter is set to 6700 rpm.

[Matt Cramer]

I'd like to see a tooth log - this can tell us a bit about what sort of filtering is needed.

[kaeman]

can you do a tooth log on a neg side of coil input? I thought that you could only tooth log when you are using trigger wheels... I will see about the tooth logging today....

[Matt Cramer]

Yes - or fuel only may use a trigger log like EDIS does, but they're functionally the same thing in this mode.

[kaeman]

I will try to get a tooth log this week, I had to change out axle studs to pass tech for the drag strip. so the car has been up on jack stands... will try to get the log and post this week.

[kaeman]

here are a couple of trigger logs. I checked and have no filtering on the ignition input currently, after looking at the logs, I was thinking that I should set the filter to 2ms, which would allow for 7500 rpm before the signal was filtered out. my 6700 rev limit is about 2.24ms trigger timing. the 23.27 file is idling, the 24.23 file is about 2500 rpm and the 26.04 file is about 4500 rpm.

2014-05-28_20.23.27.csv

2014-05-28_20.24.23.csv

2014-05-28_20.26.04.csv

[Matt Cramer]

The way those logs look, I'd turn off noise filtering completely.

[kaeman]

Well I have done some more testing, I am getting wild rpm spikes(the engine is not revving up that high) and my fuel pulse width is dropping to 0. Here is a new datalog and msq.

2014-06-03_21_modified1.msl

2014-06-03 tune.msq

[Matt Cramer]

Any way you could use something other than the coil negative for RPM input? What sort of ignition are you using?

[kaeman]

I am using a Mallory magnetic pickup distributor and a Mallory coil. I am wanting to change to a 36-1 tooth wheel and vr sensor. and make a cam sync out of an old distributor. and run distributor less.
Or possibly just go with the 36-1 and use the current distributor to share the spark and run the timing with the ms3x ecu. My final goal is to run cop distributorless.

[Matt Cramer]

In the meantime, try splitting the VR sensor signal with the MS and the ignition box.

[kaeman]

I am currently not using an ignition box, just driving the coil with the distributor output, and ecu is getting signal at distributor also. I was looking at the datalog the other night (the last one posted in this conversation) and noticed in the log that there was a timing error log. I opened up the timing error channel and everywhere the engine sees the funny rpm spike the timing error is crazy. the big spikes have timing errors in the 127% range... and the PW goes to zero right after on of those events.. normal timing error seems to be less than 3%. I am going to turn on noise filtering for the tack period rejection and use the 1.9ms time and 50% for a test. not sure what the right direction(starting value) for the % should be. I don't know the correct operation of the % portion of that filtering. any explanation would be helpful.

[Matt Cramer]

The timing error is a measure of the accuracy of the RPM input. It is, unfortunately, not sampled fast enough to be a very useful tuning tool. But it does point back to the fact that you're having RPM input spikes.

[kaeman]

Well I am back at it. I changed out the POS B&M 3600 stall convertor and replaced it with the 2400 stall convertor that was in it before. No more wild rpm spikes, the logs look clean. the convertor isn't slipping and allowing the engine to free wind. Now what I am seeing is what I think is a rich misfire when stabbing the throttle hard for a launch. I have a large input of fuel followed by a real lean afr reading and the engine stumbles for just a moment, then takes off like crazy pulling hard to the shift point. If I roll into the throttle the engine just runs great and pulls hard during accel. I would like to have some people post examples of their itb v-8 engine ve tables. Just trying to get a little info to compare.
I will be posting my new msq and a datalog with the misfire/lean spot tomorrow. the laptop is at the shop.

[kaeman]

Here are the new datalogs and the new msq. the first couple datalogs are hard throttle stab and the latter ones are slower throttle stabs.
I think I still need to take fuel out of the ve bins in the lower rpm / high kpa ranges.

2014-07-29_18.03.44.msq

2014-07-26_12_modified1.msl

2014-07-26_12_modified2.msl

[kaeman]

last logfile rolling into throttle

2014-07-26_12_modified3.msl

Any ideas, I am going to see about removing more fuel from the high kpa idle bins to see if I can get rid of the misfire. I think its a misfire...