Perfect Accel Enrich - datalog request

mops · Post by **mops** » Sun Aug 27, 2006 1:22 pm

Hello Guys.

I'm strugling to get my AE tuned well. I would like you guys who have perfectly tuned accel enrich post a datalog of two with some different situations for AE, like low speed, low Ae amount, off idle AE, high speed very quick gear changes, etc...

I basically want to see how it should look on datalogs, as i got my AE tuned to look, what I understand great on datalog, but it doesnt really feel good.
I can get it working in sertain places ok, but then not in others. EG. light AE is easy for me to tune, but high speed quick gear changes result in detonation (lean) and off ifle creates a little stumble (still drivable, but not really factory quality, if you know what I mean).

What I noticed is that tpsdot is fastest to respond, then it's map and x-tau appears to be mostly lagged in terms of actual PW adjustment.

I'm thinking on creating custom config with mapdot for low ae activity, with tps ae for quick gear changes and slight x-tap everywhere, mainly just to reduce/save fuel on deccel.

But to get it right I need to see some nice datalogs of perfect AE going on !

Thanks in advance

Jedrik · Post by **Jedrik** » Mon Aug 28, 2006 8:17 am

Like you probably notice, it really is a seat of the pants effect. It is hard for me to tell much from the EGO log with reguard to accell.
On occation what I thought was detonation and/or extreem leaning was infact misfire from excesive fuel... the AFR measures full lean because no oxigen is burnt during misfire. (Thats my understanding anyway.)
I do think the datalog is usefull to examine the fuel pulse width with respect to tps, map, and rpm though, to see if it is shaped in a reasonable way.

I have requested "well tuned" datalogs in the past and gotten no responce. I think for the most part, people are interested in posting problems then disapear when the problem is fixed. Understandable

Having said that, I will try to drag up a few logs for you to explore later this afternoon.

mops · Post by **mops** » Mon Aug 28, 2006 12:44 pm

Thats great.

I basically want to see the relationship between PW, gve, ego reading, mapdot, tpsdot, mapdot, how fast is map to respond with respect to tpsdot....

definitely post few logs, i'll have a look

S.Bretz · Post by **S.Bretz** » Wed Aug 30, 2006 5:44 pm

What I think would help ALOT is to see how much EXTRA fuel is added. ...SOmthing like the added acc fuel so say IDK, +.4ms that way you don;t have to guess and cross reference the load point and what that load point is at while not under acc/enr. I too have been toiling over AE's the last few days...and I think I'm on to something...but I won't post up andy solid number until I;m 100% happy. I get really good aes at low speed, but up top under hihg boost, I;m getting fluctuations. I'd know if its form the tunnel ramming going on in the intake manifold or the PWM of the boost controllor letting boost by the control valve...but I did hook up a fuel fitler this afternoon to dampen the effect, but with the Ernesto storm wetting the raods, I havn't been able to test it out yet.

I'm getting good results with a high map lag (95) and have the map bins set up something like 1000, 800, 500, 200 kpa/s, and the accel enrich kick on at faster then 130kpa/s. The x-tua is relly helpful at lite acc. I remade a new x-tua map in the last few days and set the scalar to 2...it drops down to about 12:1 under lite acc....I wish the af was around 13, but its better then the 18:1 and 25.5:1 A/F's I have been seeing.

WHile going over the DL's it seems that the ac/enr kicks in just a fraction TOO LATE in most cases.. the way its set now is that it spikes to like 14.5-15.5, then drops down to the 12 mark. I would like to get rid of that little leaness, so I figure more refinning is in order.

SO far it looks like really short acc enr times (.1-0.2 sec) is the way to go...at least for me.

LT401Vette · Post by **LT401Vette** » Wed Aug 30, 2006 6:06 pm

If we can truely peg a relationship, I'm sure we can get MegaLogViewer to work it out for us.... But there are a lot of different setups wher I think it starts to look versy different.
In reality Accell Enrich tuning has always seemed really easy to me on the MS. The default settings are pretty close, but that is probably because I think they are based on a Natuarally aspirated v8, though mine is different, it does fall into that category.

mops · Post by **mops** » Wed Aug 30, 2006 6:27 pm

thats right.
some other ecu's handle accelenrich in terms of engine cycles ~5 engine cycles. that basically means it's longer (in terms of ms) at low rpm's and really short at higher rpm's.

map lag factor 95 = nearly no lagging (100=no lagging at all, 10 = alot of lag).

You introduced a fuel filter in your map reading system, thats why now you map based AE is slow (because physically map reading is lagged). similar effect can be achieved by creating restriction in map line. that way i got my map reading VERY stable +- 0.2kpa at idle., but the map reading was late therefore mapdot was late and map based AE stuffed.

no matter what you do mapdot is always a bit late from the real thing. I heard that stock nissan ecu's run tps AE (not a confirmed source).

i think that tpsdot is the fastest to respond to changin conditions.
however in NA engine, if map reading is at max (say 100kpa) then additional movements wont change it and AE should not be applied... I think in boosted engine over 100kpa mapdot entirely should be doing all enrichments (because throttle has not much to do with boost in certain conditions).

never on MS forums I've seen explanation what happens if ae is applied over extended period of time... think of gentle full throtlle movement, where ae led is on all the fime for about 0.5 s or even more... is it additive ?

AE tuning screen in megatune is useless. there are two bars to show momentary tps and mapdot. great features they would be if they had 'peak hold last value' - they have graphical peak hold, but what number was it nobody knows....

same thing in logs. there's an AE% value... but the tuning is donw in PW extending in miliseconds... it would be usefull to see that value in the log, instead of damn guessing all the time...

I'm thinking that perfect at would be compose of 0.1s of aggresive tpsdot and short period of xtau (primarly for it's fuel saving features).

Can somebody trully say that they got their AE tuned perfectly ? both off idle and high speed, quick gear changes, and everywhere else ?

Post by **Bruce Bowling** » Thu Aug 31, 2006 6:19 am

mops wrote: never on MS forums I've seen explanation what happens if ae is applied over extended period of time... think of gentle full throtlle movement, where ae led is on all the fime for about 0.5 s or even more... is it additive ?

Good question on this. What happens is that once the AE trigger (either MAP or TPS) exceeds the threshold, the enrichment is triggered. Once it is finished it can be immediately re-triggered. So a slow throttle movement, one that exceeds the threshold setting, will immediately retrigger the AE over and over.

Also, there is debate over if a short enrichment time vs. a long enrichment time is better. From what I find it depends on the application, and also how well the VE is dialed in.

I also know that many use the VE table itself to introduce the enrichment and only add very small amounts of AE. The VE table areas at low RPM and high MAP can be jacked up a bit to do this.

mops wrote: I'm thinking that perfect at would be compose of 0.1s of aggresive tpsdot and short period of xtau (primarly for it's fuel saving features).

Can somebody trully say that they got their AE tuned perfectly ? both off idle and high speed, quick gear changes, and everywhere else ?

In one install, a SBC with TBI, I got extremely close with a combination of X-Tau and a very small TPS-based AE. This is on a good VE table, which made all the difference. X-Tau took away all hints of lean tip-in and the AE dialed in the WOT tune.

For those really interested in AE, I really recommend spending the $12.00 (each) and going online to http://www.sae.org and purchasing the Aquino paper (number 810494) and the Hires/Overington paper (810495). The wall-wetting papers by Cowart at MIT (also at SAE) are also excellent!

- Bruce

hobieboy · Post by **hobieboy** » Thu Aug 31, 2006 7:17 am

I'm not sure there is a perfect AE log because it would depend on the config (e.g. normally aspirated vs boost). And to know its perfect, I think you'll need to be able to compare it either on a dyno (most dyno runs are for steady state hp/torque rather than accel I think) or a GPS based system that records accel performance based on exact same criteria (which is impossible).

I do agree that even the log is pretty useless. In fact, I have asked the question before: what does the TPSacc column means without any response. I see a big negative # when I accelerate (engine code becomes 17).

My thinking is that: if I use speed density (MAP based) VE table, then AE should be based on TPS to give quicker response and rely on dialing in the VE table to supply correct steady state fuel?

mops · Post by **mops** » Thu Aug 31, 2006 12:20 pm

Thanks Al, good info there ;)

hobieboy:
correct steady state fuel is not a really well defined value. if you tone the car to the very edge of misfires, you are then very dependednt on good AE, as you will get misfires if it's not fast/good engoug. BUT if you dont tune your car to the very edge of misfires then AE is much more forgiving.

I do realize that AE will look differently on different engine, but still I really want to see a datalog of good AE, nevermind what type of engine. As i said, i'm not after hard values, but the relationship/behaviour of mapdot, tpsdot, map, rpm, ego feedback, pw, spark, etc...

on my engine it seems like xtau can handle low AE just fine, but WOT gear changes require still alot of tpsdot AE...

moreover, i'm noticing following things... with just standard (mapdot, tpsdot) AE, when I releasse the throttle then ego immidiately reads rich for up to even a second (10.something afr's)

and the more concerning matter... rpm based AE ?
when i set my wot vemap bins... say i dot a wot run on 2nd gear. i adjust the bins aiming at 12.5:1 (na engine). i redo the 2nd gear run and fine tune to 12.5:1.. that way i can get the whole wot pull to within 12.3-12.7 afr's from 2k rpm to 6.5k rpm. i can repeat it on the same day and it's ok. BUT if I then make a wot run on 3rd gear I immidiately see slightly richer afr's.. even even riched at 4th and 5th gear...
Basically it looks like whle rom's are changing slow then engine is going rich on WOT, while when rpm's are growing fast then engine is going lean on wot... feels like for low gear acceleration a 'rpm/sec' AE is required....

did anybody noticed this ?

S.Bretz · Post by **S.Bretz** » Thu Aug 31, 2006 12:24 pm

I got alot closer today. I followed the suggestions in the Mega manual for x-tau...yeah, what a big help

. It said if you get a lean stop, try turning that wholeare around there up....well. I kept uping an upin, and it got teh more woresreres and worsrererrr.

Turns out that it was getting TOO much fuel in most cases. I was lucky enough to catch an a/f in the 9's afte one of the caughs and figured out what was going on. I turned the x-table back to "my" base map nad turn the puddle factor down to 25%. It HELPED A BUNCH. Under fast tps, it was still leaning a little. I moved the sliderto 75% tps, and turned the AE numbers way down .1 @10%/s and .4 @ 80%/s. It go a little worse. I turned them up to .3 @10% and .5 @25% and it elimintaed the lean fast tps acc.

Like Mops said, the tps is faster...the air pressure has to make from the t-body to the map senor for it to read...by that time, the engine has already hit a lean spot....with the large intake and t-body I;m using on a 4 cylinder, it was just too laggy to have more map based...lean the enrichment toward the tps seems the way to go. Currnetly set to 75% TPS in my application.

Good luck mops. Its taken me a whiel to figure out what was happening. If you can borrow a faster computer, it helps alot. My new computer logs 100X better thne my old one. WIth the old comp my DL's were riddled with the "lost comm" lines. Thisnew one isn't the fastest in the world, but far better. ....plus this laptop doesn't turn off if I go over a bump wrong...that helps a bunch too