I'm hitting boost targets perfectly with the following....which I believe you helped me with

Attachment 44398

@rfinkle2 is your load dynamics still zeroed out and you are using the load error comp for hybrid purposes or did you switch to pure load tuning?

While not nearly as versed with dialing timing as you are, I have noticed the same thing. I recently spent quite a bit of time trying to smooth out the OTS timing tables while adding a little something in the higher load cells thanks to my 2.5/10 E85 mix :headbang:
(side question, why are those things so damn jagged in areas, especially the lower load? Seems like there will just be random low or high values thrown in that cause big dips in the graph or don't smoothly follow load/rpm.)

Anyway, I aslo matched the OL/CL tables and wanted to compare these to the Max A/B. Both the commanded tables call for up to high 50s in some of the lower load/rpm spots and I routinely observe values in the 50s while cruisin on hwy. Looking at the Max tables though they seem to top out in the 40s.

You guys are still being pussies with Boost and Load Error Comp....Real mean zero both out and run em to the max.

I was trying to pure load tune, but was told by Lex and David @ Cobb, that it is unlikely I will be able to do so, given the nature of the underlying ecu logic.

So for a boost tune or a hybrid tune the "whole" numbers are fine correct? I am currently using the cobb OTS values divided by 4 I believe due to my GS...but they are not decimal numbers like my BEC.

Just curious if I need to change them for some reason.

If you are happy with your tune, I wouldn't fix what isn't broken.

Gen2 does not have an absolute load targets table, and fwiw, even the guys "load tuning", are using the boost comp tables as well.

It seems that the Mazda factory tune is a "hybrid".

I can give that a try...

I had been using the following:

However, that doesn't "feel" like the right answer since I should be stupid overdampened if the loop was off by that much...

TBD...

Motivated Car Events
 

Nothing feels better than building your car with your own two hands

I am using a 4.5 gallon E to a full tank mix. This is my proposed fueling to increase gas mileage/part throttle response. Fueling is set to transition at 1.5 load on the spreadsheet. Please provide feedback on what you all think.

On my latest log initial WOT showed 1.76 load, climbing to 2.7 load and tapering to 1.95 load by the end of 4th gear log. This was NOT using the proposed fueling I am attaching.

@phate @Enki

I totally agree with you. (not feeling right), although in terms of kpa vs. load, it seems closer in terms of units of measurement as far as what correction should be made.

I am still in the conversion process (or hope to be to full load tuning).



It is either the grimmspeed's operation that is causing a difficulty, or it could be a combination of my tables and their interaction.

Fwiw, the grimmspeed is silly good for boost tuning, although I know that you are not fond of it.

Alright, here is the new Fuel Table with the New CL Max Load Table. All other tables are exactly the same (with the exception of the knocking fuel table) Feedback, thoughts, and opinions before I start running this map...

To answer your question, the Max A/B tables don't control the maximum timing allowed, it controls the maximum timing allowed by the previous table lookups in that "Ignition Tables" folder. When you car moves into an over-run (engine brake) state, it'll usually shoot up timing close to 50-60 to compensate for the assisted energy from braking. That value is at 60 degrees, and is the true Maximum Allowed Timing.

We haven't had many requests to expose this value, as most care about minimum timing allowed (for things like shifting). I'll see about squeezing it in to our next release if I can.

Thanks,

-David@COBB

Thanks, David. What I am most concerned about is that even when comparing each of my timing tables at a given load and rpm, we cannot limit or even direct timing in that range unless it is at full throttle in open loop (at least I think so - I haven't tested this extensively enough).

As an example:

I was trying to command ~25° @ 3k rpm @ .8 load - just to test the condition because this is typically where I cruise around 65-75mph (due to MAF scaling with straight E85).

All of my commanded tables show the same value at each rpm/load range, so every one of them was commanding this 25°. My max A/B tables are set to 1° above those commanded targets. My car was actually running anywhere from 40-55° of timing in that rpm and load range, which is well beyond my commanded or max allowed :/ Edit: that is all at steady state cruising on the interstate, so no engine braking and with very slowly rolling hills.

I'm not worried about the absolute maximum, I'm concerned that we don't really have control of timing in that range.

Please let me know if I can send you anything - I already have some data logs and I can send you my tune or tables or whatever you need.

Any chance the next release could have some updates to resolve the load targeting issues detailed in the recent posts in this thread?

Phate...I too find that the commanded timing is different from logged values while steady cruising...confusing to say the least.

Here is where I am at now without meth and 30% e85. I have a small range to adjust for my MAF cal as I targeting 12.3 AFR. As you can see it jumps lean for a while at the same time on both runs.
Nice and cool 37* amb for this.
VirtualDyno CF used.
http://img705.imageshack.us/img705/4557/dyno111011.png

The next release will have a massive backend overhaul of ATR, which will allow for rapid deployment of changes and beta tables via updates.

I can't promise it will be in the upcoming release, but it should be easier and quicker to get those changes to you after this new ATR backend is released.

Thanks,

-David@COBB

@cld12pk2go...

Putting some thought into the inability of the ecu to bring the the wgdc up to meet your load.

Maybe using the rpm comp A and B tables is the solution.

i.e., It is my understanding that the logic (@ least with the boost tuning toggle switched on) switches ramp limiters when certain conditions are met.

Maybe, increasing the ramp limit @ the high end of the rpm range will keep your low limits correct, but allow the ecu / wg more authority to increase load up top.

Screenshot of proposed idea:

I just spent an hour searching the forums for exactly what those first 2 tables do without luck. Got a link?

@ your service:

hyperlinked to explanation from David @ Cobb:

http://www.mazdaspeedforums.org/foru...ml#post1009814

That seems to be defined within the context of using boost targeting logic.

Any details on the functionality of those tables when using either stock logic or "Abs Load Table" targeting?

I have asked a few other guys that use atr frequently, and it seems that the factory set up works in much the same way that "boost" tuning does.

I.e. The weighting of the load comp tables is so great (as mentioned earlier), that zeroing in on torque targets is the ecu's primary job in factory form.

Once within a specific threshold, if the boost dynamics is checked, it seems the ecu concentrates primarily on boost, and ignores load @ that point.

Steve from Versatune alluded in a thread on his forum that the mazda factory tune's boost targeting logic is effectively turned off from Mazda.

I honestly believe that the Grimmspeed is what causes the confusion, given how sensitive it is to the dynamics tables.

Absolute load targets tables do not exist in gen2 atr, and the only load tuning experimentation I've done was to use the trl - x gear tables.

I was under the impression that most atr users were taking baro comps and iat comps out of the equation by setting those 1 line tables to a multiplier of one, but I've seen Dustin recently post that calculated load is dependant on those tables, which leads me to believe he is using them in factory form.

All the rambling above leads me only to lend one piece of info that seemed to work well on my gen 2 (when I was attempting to STRICTLY load tune it), i.e. removing all ability of the factory ecu to change wgdc via the wgdc boost error comp table.

Setting the trl -x gear tables .05 above what you want to see logged as a calculated load value was the only time I could get within load targets.

Gas mileage update:

1 tank over the weekend headed to Tennessee: 24mpg

1 tank headed back from Tennessee: 26.5mpg with an awesome tail wind

1 tank to Indy and back within 5 hour window today: 24.5mpg

All were done at 70mph with 1 or 2 short stops each. Indy trip had 3 stops.

Trip to Tennessee dropped fuel pressure at mile 50, lol. Pulled into abandoned gas station and cleaned up the spill valve. It's been ~1,000 miles since then and all is still fine.

Phate, is there a writeup anywhere on the spill valve cleaning? I don't even know where it is located or what is involved in taking it off, but i'd like to know for precautionary purposes.

Yep, silvapain just posted one:

http://www.mazdaspeedforums.org/foru...l-valve-96970/

So how lean are you running at part throttle....

Just curious as to where you are compared to what I tried.

Baro Comp and IAT Comp do also effect the WGDC regardless of what Error you use.

Your right about that.

And on a load tune I don't see any reason not to remove the boost error comp table from the equation as your are relying on load to adjust. If your trying to verify that it is indeed effecting your load tunes.

1.00L - 14.68:1 targeted in ATR.

I've been playing with some other things in ATR trying to pick up gas mileage, and it seems to be working so far. I'll make a thread with my findings, soon. I have the same changes implemented in a few other tunes I'm doing - and it seems the results are positive :)

Thank you.

I have not worried about those two comp tables all that much (baro and iat), given I have set them to 1 so they have no affect. I was under the impression most atr users were doing the same.

The more I weight my personal tune toward load, however, the more I realize how functional those tables truly are.

As for the boost error comp tables, there was a point I was trying to tune my car without a particular "psi / kpa" target adjustment, and allowing the wg to adjust based on a load request only.

I gave that up and populated both the boost and load error comp tables, and quit battling Mazda's inherent ecu logic.

PSHHHT.....


16 or gtfo....JKJK

I tried making everything under 1.5 load target 15.5 and then 16......Seen like 2 or 3 mpg increase but hated cruising around at part throttle that lean.

hehe, eventually. I want to get my EGT gauge back in before I lean out like crazy. Most cars can handle part throttle leannes without issue. We only target stoich for emissions ;)

Right right....do it up son

I got the opportunity to tune @mazdafreak's car on the dyno, today - straight E85!!!! Same dyno I tuned my car on and the results were almost identical. We ramped in boost a little later, since the low end power is so horribly unusable, but from ~4k+, the power curves lay on top of each other. Very impressive, if I do say so. Great car with more serious power out of a K04. GT28/GT30 gas cars beware, lol - there is now video evidence!!

Tune went great!!! I'm more than happy with the outcome!! If anyone has seen my recent post on my thread, you'll see the results, if not, ill post it here if @phate doesn't mind lol.

http://i1205.photobucket.com/albums/...A0NTguanBn.jpg

I should mention, though, that mazdafreak's car was getting some spark blowout on the top end. Had he not had that, he probably would have carried more just a touch more power through the very upper end of the power curve than I did ;)

Finally stages of tuning and it started doing it :/....Ohh well...u got lucky! Hehe jk jk.....on a side note, I did order the plugs tho lol

Just filled up my tank 13.473 gallons; 7.120 E85 / 6.353 93. Anyways had put on 263.3 miles of mostly highway cruising with SOME boost and only managed 19.54 mpg. Still rather have the smile on my face though. Clint I'll be on the rollers sooner than later!

I am currently using 6 gallons of E to 10 gallons of 93. This is stepping up from 4.5 gallons E to full tank. I will Dyno next month.

I plan on hitting the dyno in Feb once I am back and revamp my tune. Phate, I'm gunning for you ;)