any tips to tune out an off idle stumble?

[teg]

My reqfuel is 9, w/ a PW of about 1.8ms at idle. I'm tuning closer to 12:1 as the 5 port engine seems to idle best there so far, and I use that for full throttle too. At cruise I'm running it at 14.7:1 and it does seem to run well there.

I have a Opel Kadett with 5 port and I'm running idle as 15:1 or maybe even leaner (16:1). At full throttle I run with AFR 13-12:1 and cruise 14-15:1.

I'm using normal AE (no decay) with mapdot accel. MAP bins of 1.0/50, 2.5/100, 5.5/200, 8.0/300 and a mapdot threshold of 30. Accel time of 0.2s.

I have tried everything. Best so far I have found is rpm ae. I'm using it with threshold of 80kPa/s and 6/1000, 4/2000, 2/3000, 1/4000. I'm using 29h (some version) code.

But still I have problem when I slightly open throttle, my AFR jumps over 20 and will come slowly to correct AFR (14-15).

[whittlebeast]

Guys

Try the gauge called accelEnrichGauge "AE Driven Fuel Pct" On the opening page of MT that showes the 8 guaged, just right mouse click on any gauge that is not important to you and sellect swap to and then this gauge. If right mouse click does not do anything the gaueg has to be turned on in the ini or just update to athe newest MT. Once turned on it is easy to watch the gauge and feeling how motor reacts makes if fairly easy to figure out what the motor wants. This gauge knows about opening time and the actual percent increase in fuel above the amount caused by the VE table. On AE at the lower RPM ranges a begining percent in the 150% range (assuming you are running decay) appears to be about correct. 5% or 250% will never work. On the high RPM ranges the correct answer apperes to be in the 1 - 5% range. I am seeing mapdot thresholds in the 50 -70 kpa / sec range appear to work. If running TPSDot the thresholds tend to be in the 1 - 2 v/sec range. Start with 1.5v/sec

In MS2 the trick in X-tau will be watching essentualy the same gauge called "X-Tau Correction"

[jayroc]

I had some trouble with this too, but I managed to tune it out pretty well eventually.
What are your map bin settings for mapdot? I had to go up to 400kpa/s with a 12ms pulse at highest and maybe an accel time of .5 or .6 IIRC.. That took care of a big part of my WOT hestitation issues. Also, as Mike already mentioned, enrichening your high kpa low RPM VE values helps a lot too. Doing those two things resolved about 95% of the problem for me.

EDIT, oops, just saw your post with the MSQ file. You might try increasing your upper value to 400kpa/s and 10ms or so.. If it seems like it's falling on its face after WOT under load, I would increase accel time too 'cause .2 is pretty quick.

[mikesb]

Andy sent me an MSQ for msnse 26h4. Burning a new version is a little more than I have time to get to right now. But I did have a look at what Andy had done, and tried to emulate it in my 24s13 version that's currently installed.

First big change was in the accel wizard - I had my MAP values entered upside down - low to high instead of high to low. D'oh! Good catch Andy, thanks!

Andy also suggested a longer accel time, and to use the decay.

A bit of fiddling and I had it sorted for the drive home. My peak Kpa/sec is around 250, so I think I'm ok using 200 as my top table value for now. I changed from 0.2s to 0.5s accel time and added a decay to 0. This let's me use fairly high map based AE values which really helps get past the lean pop. The decay helps reduce the stumble following the pop, great!

I'm still getting a lean stumble just as the throttle opens, but it seems that it's much less now. I'm going to fiddle for a couple of days to see what I can do there.

I'm looking forward to trying rpm based accel, fingers crossed for good weather this weekend.

[whittlebeast]

At some point you mentioned the it was too cold topop the case apart to burn a new code. Do you realize that once you have extra loaded on a squirt the days of junming the short are gone.

As a side note on my test MS2 box I have a switch drilled into the case for flashing new MS2 code.

AW

[mikesb]

No, I didn't know that! I've been using easytherm since I've got a non-standard coolant sensor, I never even looked the download-firmware.bat.

Thanks again Andy!

If I had a spare laptop or snse supported wireless I'd never even have to leave the office .

[mikesb]

Did some playing yesterday till my laptop battery died, it went pretty good. Installed 26h4 (it's so easy now!). Got my best tune using MAP based acceleration enrichment. I played with RPM based for quite some time and got it close too, but not quite as consistent as MAP based.

With RPM based I needed to set the low range richer than with MAP based to avoid a pop with quick throttle changes. But then I'd get a seriously rich stumbles with a slower throttle. I still get a bit of a stumble with MAP based at higher ranges that I tuned out with RPM based, but it's very slight. I'm pretty sure what I want is a 4x4 table, or maybe a %change for RPM based AE off of the MAP AE table. Where do I add my feature requests for MSNSE ?

There's still a very slight lean hesitation when the throttle is first cracked, but I can drive right through it and it doesn't interfere with heel and toe either.

Now that I've got it running so well (thanks again everyone and especially Andy!) it's time to try to get the "'n spark" part going.

[mikesb]

I promised to post back here if I got the off idle stumble sorted, and I did, so I am. This is for the 26h4, 6 or 8 versions of MS1 on a 2.2 board.

Remember, this is for a 5 port MGB 1800cc engine with a very big single throat throttle body from a bike. FWIW I've attached my msq.

I had to use MAP based AE, TPS changed too slowly to keep up with the engine needs and I'd lean pop like crazy.

I couldn't use RPM based - I need a very rich AE for low RPMs, but if the engine stumbled at all (and for some reason it did) I'd get a bunch of very rich pulses a low engine speeds which would flood and stall it.

The AE algorithm always starts with the first bin for AE, so I needed to have that low but rich enough so that the start of the pulse was good enough for all conditions. I used 2.0ms at a rate of 20kpa. Incidently, that's my threshold too. Any less than that and it'd trigger at idle.

The critical thing was to find the breakpoint and common kpa/s points while running - I found that was between 40 and 70 kpa/s, so I fiddled to find some good values there. Off idle I went very rich, using 14ms @140kpa. Typical pulse off idle was 12ms or so, probably 100-120kpa/s. But I chopped this down using decay to 6ms to avoid the engine stumbling when blipped from idle in neutral - this works well when accelerating from idle while avoiding chopping off the AE when I accelerate from a lazy cruise.

I used engine cycles - 15 (1.5 in the megatune menu) rather than time to end the AE. It's sort of an engine speed based way of reducing AE at the higher ranges where it's not needed, as is a low AE PW for slower MAP changes.

My mid table bins, those just above & left of my cruise bins I left a couple % rich - a slow throttle that fails to trigger often drifts too lean there otherwise. A quick ego adjustment seems to help at higher speeds.

I needed to write some log analysis code to figure this all out and it was a big help. I hope it's something Phil is able to add to the megalogviewer.

HTH!

[woh]

Mike

I've been following this thread with interest. I'm using TPS acceleration enrichment but I want to try the MAP based enrichment. I'm running MS1 v3 with 29j and I don't want to load the .msq.

Could you give me your AE bin settings, your requ fuel, and the injector open time?

I could not figure out where to set the number of cycles. I set timing to cycles but did not see where to put in number of cycles. I wonder if the screen is different for the MS1 v3

[whittlebeast]

get 026h8 from http://www.ExtraEfi.co.uk under downloads

AW