Miata Turbo Forum - Boost cars, acquire cats.

Miata Turbo Forum - Boost cars, acquire cats. (https://www.miataturbo.net/)
-   Adaptronic (https://www.miataturbo.net/adaptronic-63/)
-   -   Tuning/Logging help (https://www.miataturbo.net/adaptronic-63/tuning-logging-help-45041/)

dgmorr 03-16-2010 03:24 PM

Tuning/Logging help
 
Hey guys,

I want to dial my fuel map in more. Right now my AFRs are pretty wild between 10's to 12's. How do you guys get a good log? Just do a 4th gear pull from low speed to the rev limit? I haven't been over 5k rpm yet, but my timing map should be ok, if I set my fuel cells to 90+ VE above 5k will I be relatively safe to do a WOT pull to redline? I still haven't done WOT yet either. I'm using the stock injectors now and running base boost of about 7psi.

gospeed81 03-16-2010 03:59 PM

Ease into it.

Drive around like you normally would...you're trying to hit as many cells as possible, so you need to get high vacuum, on throttle, and boost for every rpm you can.

Work your way slowly into boost and up the rpm range. If you hit say the 5500rpm 135kpa cell and MLV tunes it for you, the area between it and the neighboring cells is now extrapolated to that accurate VE number, and you can push a little further on the next log.

You should see the "max cell change" going down on each analysis report. Usually starting around 18-20, and slowly getting down to 10 or below. Once down to 5-7 max cell change you're pretty well done. It will continue to make small adjustments, but they won't affect driveability much, and can be affected by other factors (road grade, changing ambient temps, air density, lunar cycle).

Once the map feels good and drives well I start exporting the table under a separate file name so that if the next wash and rinse goofs something up I can go back to the other one and try again.

You're usually in pretty good shape after 10 10-15min datalogs.

dgmorr 03-16-2010 04:25 PM

The car already feels very drivable, but I know my fuel map isn't properly tuned because one cell could be 20 and a neighboring cell may be 40 or more. There are a lot of spikes and dips in the fuel map.

I'm going to find a couple long hills tonight and try to hit more cells.

TurboRoach 03-16-2010 06:05 PM


Originally Posted by gospeed81 (Post 539114)
Ease into it.

Drive around like you normally would...you're trying to hit as many cells as possible, so you need to get high vacuum, on throttle, and boost for every rpm you can.

Work your way slowly into boost and up the rpm range. If you hit say the 5500rpm 135kpa cell and MLV tunes it for you, the area between it and the neighboring cells is now extrapolated to that accurate VE number, and you can push a little further on the next log.

You should see the "max cell change" going down on each analysis report. Usually starting around 18-20, and slowly getting down to 10 or below. Once down to 5-7 max cell change you're pretty well done. It will continue to make small adjustments, but they won't affect driveability much, and can be affected by other factors (road grade, changing ambient temps, air density, lunar cycle).

Once the map feels good and drives well I start exporting the table under a separate file name so that if the next wash and rinse goofs something up I can go back to the other one and try again.

You're usually in pretty good shape after 10 10-15min datalogs.

MLV isn't used to tune the adaptronic so none of that info is available to us. We only use MLV to display the log files.

dgmorr 03-18-2010 09:39 AM

Alright, I was able to get some good data through a few load ranges.

I am seeing a maximum knock count of 52 at 5498rpm and 151kpa. My background knock level setting is 75 from 3500rpm to 5500rpm. Is 52 an acceptable level, or do I want to have 0? If the engine sees over 75 counts, does the ecu start to pull timing, or is there a setting that needs to be on for that? Is this only if I have Ignition Corrections set to adaptive? Yes, I should read this in the manual, it's been a while since I've studied it, so if someone has quick knowledge - very appreciated.

Or is the 52 value really a function of actual count subtracted from the background level of 75...being 127?

Thanks.

gospeed81 03-18-2010 09:44 AM


Originally Posted by TurboRoach (Post 539194)
MLV isn't used to tune the adaptronic so none of that info is available to us. We only use MLV to display the log files.

Damnit, I hate it when I do that...thought I had adaptronic subforum updates turned off.

There is only one ECU to me...sorry I couldn't help.

dgmorr 03-18-2010 09:48 AM


Originally Posted by gospeed81 (Post 540305)
Damnit, I hate it when I do that...thought I had adaptronic subforum updates turned off.

There is only one ECU to me...sorry I couldn't help.

lol, I figured this is what happened. No judgements :laugh:

TurboRoach 03-18-2010 07:52 PM


Originally Posted by dgmorr (Post 540302)
Alright, I was able to get some good data through a few load ranges.

I am seeing a maximum knock count of 52 at 5498rpm and 151kpa. My background knock level setting is 75 from 3500rpm to 5500rpm. Is 52 an acceptable level, or do I want to have 0? If the engine sees over 75 counts, does the ecu start to pull timing, or is there a setting that needs to be on for that? Is this only if I have Ignition Corrections set to adaptive? Yes, I should read this in the manual, it's been a while since I've studied it, so if someone has quick knowledge - very appreciated.

Or is the 52 value really a function of actual count subtracted from the background level of 75...being 127?

Thanks.

Yep the knock level is actually 127. That is about what I was seeing when I was N/A but at a different RPM.

The ecu will reduce timing when set to closed loop timing". When set to adaptive it will reduce/advance trying to find max torque.

I would do another run with the master ignition trim set to -3 or -4 degrees and see if the knock level changes or use detcans if you can. If the knock level doesn't really change then you should be ok.

One thing to watch out for is that in closed loop timing some of the default values are really aggressive when pulling timing. Make sure that they aren't going to pull all of the timing out.

dgmorr 03-18-2010 10:48 PM

Thanks, I was actually reducing some timing cells in the area manually. I'll give the master ignition trim option a test tomorrow morning.

18psi 03-18-2010 11:11 PM

yeah man, take your time, richen it out so you're safe, then let autotune do its thing. When all or most are marked tuned, mark them "untuned", change the precision of the autotune (by raising time it takes to tune a cell and acceptable margin of error), or changing it to "slow converge" (essentially the same thing) and do it all over again. It will keep smoothing it all over for you as you're doing it over and over til you get a fairly decent map. Then if you want perfection you'll have to fine tune any remaining cells by logging and by hand

dgmorr 03-18-2010 11:37 PM

Thanks for all the tips!

18psi, did you ever dyno tune yours? Or did you do it all by autotune?

18psi 03-18-2010 11:41 PM

No dyno tune yet, but I will have one in the near future: a friend of mine is getting a mustang dyno in his shop in a month or so.


I used autotune for a semi decent overall fuel curve, then started fine tuning it by hand.

For the timing map I still haven't touched it from the base map I made right after I rebuilt the motor. I took a couple VERY knowledgeable guys timing maps, and using them ALL as reference put together my own SUPER conservative yet very good/smooth map. So far the car's been running great power wise (still have that stupid lean issue when the fans kick in at hot idle or hot restart) and it HAULS ASS.

I just picked up a set of electronic det cans from the group buy and will start fiddling with the ignition map. The car is a beast already but I'm never satisfied:)
Just recently raised it up to 12-14psi

dgmorr 03-19-2010 08:35 AM


Originally Posted by 18psi (Post 540782)
Just recently raised it up to 12-14psi

Need vids :bigtu:

I was a bit skeptical about autotune when I first started out, but I've been recalibrating my LC-1 every so often and it seems to make the fuel map a bit more logical.

dgmorr 03-20-2010 04:22 PM

Alright, I was able to do a few more runs this morning. I set the master ignition trim to -5*. I am seeing 83 to 89 knock count at 6600 rpm and 158 kpa. I was previously seeing 52 at 5500 rpm. How do I know if this is not the resonance of the parts vs actual danger? Considering that I reduced the ignition trim by 5* today, is it enough to rule out actual knock if it is still occurring? My ignition at 6600 rpm and 158 kpa is 19.8*.

I do not have det cans handy at this time.

I also had massive flames shoot out when I hit the limiter in 1st and/or 2nd, and the AFRs shot up to 22.3. I'm assuming the high AFR is expected since I have cut fuel and ignition for the rev limiter?

AIT was 8* C
Water was 93* C

Thanks!

18psi 03-20-2010 05:13 PM

how the hell were you able to see 22.3afr? what wbo2 do you have? Mine only goes to 17 iirc


And I havent heard of anyone having false knock way the heck up there which leads me to believe you actually have knock.I have less timing up there. More like 14

dgmorr 03-20-2010 07:06 PM

I'm using the Innovate LC-1. The range in the Innovate software shows 7 to 23 AFR. I guess Wari shows whatever the controller sees.

Do you use Closed Loop ignition or Open Loop? I'm going to play with the Close Loop parameters tomorrow as well as back off my timing map a bit more. This is already using 94 octane too :vash:

Anyone have any idea what an acceptable level of knock is? I'm guessing 0, since I can eventually create hot spots if I sustain these levels over a period of time?

Vlad, did you capture any during your catastrophe?

I am using NGK BK6RE plugs, should I be using the next step colder?

Thanks for all the pointers, everything is starting to make a lot more sense now.

TurboRoach 03-20-2010 08:48 PM

Closed loop is what you use if you want the knock sensor to have an effect on timing.

Open loop just applies temp corrections to the map.

dgmorr 03-20-2010 08:50 PM


Originally Posted by TurboRoach (Post 541645)
Closed loop is what you use if you want the knock sensor to have an effect on timing.

Open loop just applies temp corrections to the map.


Yeah, I'm gonna see how running in closed loop works out. If I can get it to work without pulling all the timing, it might be good. Any idea what good parameters would be? Default looks terrible with 1* retard for every knock count.

TurboRoach 03-20-2010 09:12 PM

I think I went with 1* for ever 5 or 10. Its going to depend on the shape of your background knock and how much total advance you are running. My timing isn't very aggressive so I don't need to pull much /knock count.

dgmorr 03-24-2010 09:38 AM

I'm at 5* with master ignition trim and reduced the areas I've seen generating knock at least another 4* or so and still seeing knock, and it's actually feeling a bit slower :facepalm:

Time to take away more :crx:


All times are GMT -4. The time now is 02:07 PM.


© 2024 MH Sub I, LLC dba Internet Brands