TD_D mod journey. From stock to the 'bastardmobile'

[Td_d]

Sure - can do - I've just switched back to CAN now that the trims are dialed in. i think I have a very good idea as to why it's stumbling - the load compensation is leaning out the AFR's seriously - the last set of comps I put in were very close, but still a small stumble at around 2900-3000, so I reduced the numbers around the region - and boy did it a) lean it out badly (18 AFR at 1.1 load !! - knock, knock, who's there), and b) worsen the stumble. So I've reverted to the old comps, and massaged them in the opposite direction. So now I can assume that increasing the numbers richens the AFRs, and vice versa? I'm still a little vague on what how the % change in engine load translates to AFRs through the table.

The car, however, is driving like an absolute bastard on steroids! Weather is getting cooler here as well (don't worry, checking knock - none in OL), and the power is phenomenol... plus I suspect I can get more out, since the top end is definately rich, popping spectacularly on hard WOT gear changes.

On the down side, I friggin took out my front lip on a speed bump today (custom fabricated, fibreglass and CF) - ripped it straight in half Oh well. I had a love hate relationship with that thing, looked awesome, but too many bumps on our road, made driving a pain, constantly scraping.

 

[HolyCrapItsFast]

Looks like your getting there. This is awesome how far you have come in such a short amount of time.

Yeah so the values in the comp table represents the %error in calculated load based on manifold pressure. So when the ECU calculates the engine load, for a given manifold pressure and RPM, The load comp tables will add or remove load by that percentage and report and use that new load value instead. Hence the name load comp MAP. When a higher load value is reported, the logic will respond in kind and deliver more fuel as a result. And vice verse if the load is reported to be lower.

What this table is actually doing is compensating for flow resonances caused by after market or custom plenums. At some flow rate, the flow profile and other characteristics change and can play tricks with the maf sensor. A certain voltage may be reported to the ECU for a particular flow rate but the actual flow may be something quite different. The one thing that remains constant is pressure. No matter what the flow profile is, the pressure is the same for any given flow rate, RPM and load. So to compensate for these resonances the table references MAP and RPM and determines the amount of error compensation to apply.

This is the very reason I think Speed Density is the way to go because you take a whole slew of variables right out of the equation and the whole "Running of the engine" process becomes much easier and consistent because you are only relying on manifold pressure and the mechanical properties of the engine itself.

Sorry to hear about your lip. I did the same thing on my way to my Dad's house recently. I purchased another one and it is waiting to go in.

 

[HolyCrapItsFast]

Oh yeah your right. I should rephrase what I said...

If you hope to maintain the AVCS then you need to stay Mass Air. As AVCS changes so does VE and Mass Air will be able to compensate for those changes. Otherwise speed density would be better IMO.

 

[Td_d]

Hey George - here's a couple of logs with the requested variables, I wasn't too sure about 1 or 2 variables, so you'll see there's a plethora :lol:

I did the most stupid-assed thing today - filled up and in my rush, forgot to add the NF octane booster. Well, I can vouch for it working, pulled 11.75 timing at 5400 RPM Thankfully I instantly felt it pulling timing, and backed off. Also, thank god this is a built block, otherwise I would be seriously worried right about now. :tard: What was odd was that it went straight to FLKC (you'll see it in log 49), no FBKC. I thought that it needed to pick up a short term knock first, and then learn it?

On the stumble, further massaged the comp tables, and richened up the MAF scale at the problem area, and much better now - still a slight stumble at 2900/3000, but nowhere near as bad - check out the logs.

http://dl.dropbox.com/u/3099895/log0049.rar

 

[Td_d]

Low det tables? You definately have my attention! I do have a low boost (1.2 bar) program on the EBC that will happily handle NF free gas, but I completely forgot :tard:

 

[Td_d]

Ok, I have to say, IE 9 sucks balls, constantly crashing, dammit. Chrome FTW. Anyways...

That would most likely be the 'failsafe' OL fueling maps in O/S I think? I think it required IAM (DAM) to drop though to trigger. My failsafe maps are pig-rich, for obvious reasons.

 

[HolyCrapItsFast]

Yes but as suggested that table does not activate until DAM as fallen to a specified value. Too late IMO. Albeit the value can be increased so you can have that table activate sooner. There are also timing tables which do the same thing. If I remember TD_D's DAM threshold is set much higher so it should become active much sooner than stock. This was a strategy that his original tuner used and I agreed with it.

TD_D's logs shows that DAM/IAM never goes below 1 so the table is never invoked.

Correct me if I'm wrong George, but shouldn't he be able to set the low det tables to handle this in case he forgets the additive again? I don't completely understand all of the tables in ATR yet, but this seems to be the purpose of this table. Open source should have a comparable table, right?

Ok, I have to say, IE 9 sucks balls, constantly crashing, dammit. Chrome FTW. Anyways...

That would most likely be the 'failsafe' OL fueling maps in O/S I think? I think it required IAM (DAM) to drop though to trigger. My failsafe maps are pig-rich, for obvious reasons.

 

[Td_d]

Yup, I figure if you land up pulling 15 - 20 degrees at 6500 RPM plus before the failsafe kicks in, the engine is probably toast. Well fueled toast, but toast none the less :lol:

 

[Td_d]

And yes - my failsafe maps kick in at IAM below 0.975, basically instantly, as opposed to 0.35 for stock (which in my case would be way too late).

 

[HolyCrapItsFast]

Now you know why I pretty much know nothing else about the car. I spend all my time with the engine and tune... Mostly the tune.

 

[Td_d]

^ that's how I feel about some of your engineering threads Fuji! Doesn't stop me from peeking in and getting confused :lol:

 

[HolyCrapItsFast]

Ask away on any questions. I'm trying to succinctly explain some rather difficult concepts, and questions will allow me to refine my explanations. The thing about tuning is that the tables and their associated definitions does not allow me to fully understand the link to the internal behavior of the engine in most cases. It takes me longer to understand the tuning process because I have to visualize the mechanical effects resulting from a changed value.

I have this same issue at times. Especially with the AVCS.

Hey I do have some time coming. I can take it any time I want really. You tell me when you have time and I'm there!!!

Sent from my Desire HD using Tapatalk

 

[Td_d]

Happy birthday to me, damn, one year closer to 40... Maybe I should get my ass to bed

 

[HolyCrapItsFast]

Welcome to middle age... almost.

You will know when you get there when your knees start forecasting the weather :rofl:

Happy Birthday MAN!

 

[IGOTASTi]

Happy B-Day.

 

[Td_d]

Thanks guys - I was wondering what that thing with the knees was I was under the car 2 days back removing the remnants of the front lip, cursing at the limited working space - man did I wake up sore the next morning :lol:

 

[Td_d]

I meant to ask you Holy, see anything enlightening in those logs?

 

[HolyCrapItsFast]

Yeah I do... Still not done looking but take a look at the intake AVCS. They seem to be inconsistent from one bank to the other. They should really only ever vary 1-2* between each other Sometime the right side is higher than the left and others it is lower. There are some spots where the right side has no advance at all and the left side has full advance.

This could indicate a faulty solenoid or possibly a clogged filter screen. The exhaust AVCS looks fine to me.

I would also do a TGV delete if you have not already done so. That should help out greatly with closed to open loop transitions

 

[Td_d]

Hmm... I see what you mean. It seems to be much worse in C/L - which would explain why a zeroed out AVCS table in those areas would result in a much smoother ride.

I'd like to test completely zeroing those areas (like the base tune - well zeroing and then tapering) - before I create havoc - timing obviously has to change - down I would assume?

EDIT had a look at the two maps, the timing is less aggressive with the lower AVCS overlap, so to be safe (and since I know that map does not knock in those areas) I have transplanted the timing across, and smoothed it lower on some of the 'edges'.

 

[HolyCrapItsFast]

Yeah I'm pretty positive that there is an issue with your intake AVCS but I would want someone else to chime in and confirm this.