ben73

Guys,
I have attached a WB log below showing what my AFR does on a gearshift. (700r4)
The log shows rpm in blue, thru 1st gear up to 5400rpm, shift to 2nd and then a run thru 2nd from 3450-5800rpm.
I have tweaked the VE table to give a pretty flat AFR at WOT, see the circled section, but its got me stumped why it always has a big lean spike on the shift that takes quite a bit of time to richen up.
The TPS stays at WOT, the MAP doesn't change, so I can only imagine that it is pulling fuel based on the rapid change (fall) in rpm???
Does anyone know how to fix this?
The pink is TPS, green is AFR.
Ben.

RBob

iTrader: (1)

AFAIK there is nothing in the '747 ECM that pulls fuel during a rapid RPM drop. IIRC Jon also posted something along the same lines with the '8746 ECM.

As for the cause I can only offer a theory. At high RPM the air flow is high enough to strip the intake manifold walls of fuel (and head runners). When the shift occurs the velocity of the airflow drops as the RPM does.

This causes some fuel to be re-deposited onto the intake manifold walls. In turn less fuel reaches the chambers until the fuel film is built back up.

There may also be a small decrease in MAP at high RPM. With the shift the MAP will increase a little with the lower RPM. This too will affect the amount of fuel on the intake & runner walls.

To correct for this add fuel (increase the VE) at the lower RPM area that is entered after the shift. May also add a little fuel to the top end RPM for a before the shift enrichment.

Does the 2-3 shift show the same magnitude of a lean spike?

RBob.

P.S. May also need to extend the AE period. There is a lean spike at the same RPM in both 1st & 2nd. . .

ben73

Thanks for your comments RBob. I'll try richening the VE's at the rpm points just before and after the shift.
I'm not sure what happens on the 2-3 shift with this tune as it's moving at well above the posted speed limit by that time Thats enough for instant licence suspension here....
But last time i was at the track with a slightly different tune it did the same thing..
I'll try it out.

ben73

Just wanted to refresh this topic now that many more guys have wide band setups.

Mine is now a LOT better than in the plot above, and I no longer use a 7747, but i was wondering if other guys were having problems with a lean spike on the shift points, and what you have done to correct it??

The really odd part is that my Injector PW goes UP during the shift, at the same time the WB reports a leaner mixture!! I'm starting to wonder if the lean condition is 'real'.. Altho, I am a bit confused, as i swear i can feel it..

Also interested to see if this is a TBI characteristic, or if PFI setups have the same issue...

Comments??

Ben.

BMmonteSS

I don't have the lean spike on the shift, but I'm having a heck of a time getting the pump shot right. When I blip the throttle from idle I get a huge lean spike like your seeing, but then it goes strait to 11 or 10 to 1 for a good half to full second. It seems like Rbob is right, it has something to do with fuel being stripped from the manifold walls. Once I get the air velocity up it strips the extra fuel from the walls and causes my rich condition while at the same time it takes too long to get my pump shot to the cumbustion chamber so I also have the lean spike.

Is there any way to control how long the pump shot last? I'm only seing delta tps and delta map with a quanity and not a length.

ben73

Yeah, I think AE tuning is a bit of a black art, especially when you are using one of the older ecms that only have a 25% TPS AE table.

HaulnA$$

I have spent countless hours messing with AE and PE on my 16197427 $0D TBI PCM and witnessed the exact same thing you show in your WB log. Here's what I found: The flatter that I goy my AFR curve by adjusting VE and WOT AFR vs RPM to match my WB, obvoiusly the spike on gear shift got a lot smaller (from a 3 to 4 AFR spike down to less than 1), but I could never totally get rid of it and believe me I tried everything. The other thing I noticed was that I had to drastically increase MAP AE by almost 300%. I started out with BLM's of 128 everywhere but noticed a lean bog on WOT exactly like yours. I initially got rid of the lean bog by fattening up the WOT AFR vs RPM table down low and all seemed good with a flat AFR curve but the numbers in the table were way off from my WB so I started adjusting AE. I quickly figured out that TPS AE only lasts about .1 seconds (but my table goes to 50%) and that MAP AE lasts about 2.5 to 3 seconds and/or is done by 3000 RPM. After getting AE dialed in, I fine tweaked it by adjusting the DIFF MAP AE CORRECTION vs RPM table. Now it doesn't bog and the AFR's match the WB. HTH

Grumpy

Naaaaa
Just takes a couple thousand pages of notes, to get it perfect.

I think alot of guys, just have the PE enables set too high, and wind up with needing alot of AE, to cover for that. That's also why I mentioned using a MAP and TPS PE enable when you can, but that takes being able to do a patch.

ben73

Ahhh, we need a lightbulb emoticon ;-)
My PCM does have seperate MAP and TPS AE enables.. I'll have to have a play with that!

ben73

Thanks for that, i'm inspired to keep working on it ;-)

BMmonteSS

Yeah your explanation confirms some of the suspicions that I had. I'm going to play with my car some tomarrow, I'll try to post some of my findings.

I'm not sure if the 747 has the DIFF MAP AE CORRECTION vs RPM table. I'll have to go dig through the hack and see, I know my ecu doesn't have it. That would be very usefull to have, seems like I need more down low and less up top.

ben73

Can anyone explain why my BPW goes up during the shift, and the AFR has a lean spike at the same time??

ben73

HaulnA$$, would you mind sharing your .bin with me for a peek?? I know is a different PCM etc, but the principles are the same...

HaulnA$$

BTSOM but mine does the same thing. Hopefully when I get my PCM bench finished something may turn up. So far I am inclined to go with RBob's take on it.

RednGold86Z

iTrader: (1)

The BPW goes up with VE and MAP, so, your shift is probably going to a higher VE point or MAP in the map. The lean is almost definitely caused by the manifold wetness requiring more wetness at the "shift to" point than the "shift from" point. I don't see any other way around it (looking at it blindly). It's probably not noticed when using stock 11.5:1 PE A/F's. The only other thing that can cause a lean reading is misfire, which may be occuring because of leanness (chicken, egg), but doubtfully from an ignition problem. The wetness is also a little A/F dependant, where if you can bear (and probably benefit from) a richer A/F in PE at the "shift to" point, you may alleviate the problem.