Never heard of the fuel filter trick. Interesting idea though.

I assume you can't feel the MAP bouncing around like that and it's just the sensor? Have you tried removing the tee to see if that's causing issues?

Also, Fire, thanks for reminding me I need to spend money on a CAN Lambda gauge

 

What rate are you sampling at in your datalogs? I believe the Maxxecu Race can sample at up to 1000hz. If you could do a pull logging at 250hz or so I suspect you could see individual intake pulses. Might help you determine the cause.

 

Oh dang, that's actually a really nifty trick with the filter! Ok, if nothing else, I'm glad to hear that this isn't a one-off case, as someone's obviously thought of this before. I'm gonna see what happens if I add some mild smoothing above 100kpa and will give the physical filter a go depending on results.

As far as the integrated wideband controller.... yeah I'm still running an analog signal from my AEM for now. I do have it set up to where the gauge and ECU read within a tenth of each other, though. I didn't think I'd want/need to set up the controller so I clipped the wiring at the ECU end and now I'm gonna have to run another harness through the firewall into the engine bay to get that up and running lol. That's on the to-do list, but the list is currently looooong.

Sim, no, the bounce isn't noticeable via the butt dyno. I can try capping off the tee to the boost gauge but I'm still under the assumption that there just needs to be some level of smoothing via the ECU or in the MAP line as Nate suggested. I'll cap the line off if filtering fails.

Watterson, damn, good catch. I just checked and I'm only logging at 25hz. Never would've thought to check logging frequency. Still, since the car showed this behavior with the previous ECU, I assume it's just a result of the hardware? Lots of valve overlap while in boost causing noise in the MAP line?

I set up a table to start smoothing MAP once the engine's in boost. No idea if a 4-event sample size is high or low. The preset filtering options showed 4 events as the smallest option and 32 as the highest. We'll see what she does on the drive home.

 

Well that worked an absolute treat.



Much smoother now. Previously, there was so much noise that EGO wouldn't even activate. More tweaking tomorrow.

 

Not sure if you mentioned this, but where are you getting your map signal? I know using a port too close to the throttle body or the intake port can be a problem

Looks way better already!

 

Yeah, pretty happy about that! I dropped the smoothing factor down to 75kpa today as the signal was still noisy and affecting fueling from there to 100kpa as well. Seemed to help. I'll post up results once I drive it more.

MAP signal is coming from one of these front ports on the underside of the IM plenum. Should be a non-offensive location from my understanding?

 

Ah yes! Yeah that spot is fine, though it's proximity to the runner entrance might be causing some of the noise. Not like you really have an option with that manifold, or that it's a big deal. Software smoothing probably fixes it, but the fuel filter trick was night and day on my car and I've always run one since I don't see any downsides.

 

You better not be pulling my leg. These little guys cost me all of $6.45.




I'll pop one in next week. The filtering settings I'm using on the MAXX seem to be working well but you saying "night and day difference" has me wanting to see what effect the literal filter has versus the software filtering lol.

Tune's getting pretty dialed in. This was me all weekend.


The part of my fuel table where I spend 90% of my time is dialed in now. I still need to massage some of the fringe cells (I.E. 50kpa @ 6,500rpm), but all of the cruise cells are within 2% max EGO correction and I've got the 100-160kpa rows hitting their targets within 4-5% in most cases. Still needs a tiny bit of fine tuning but it's good enough to rip on for now. Gonna up the boost and get to dialing in the 180kpa row here next. The VE table has a couple interesting "humps" that developed, but the engine wants what the engine wants, right? I'm not too worried about the peak in the 3k rpm cells or the shelf that seems to have developed around 5k rpm.



I copied most of my timing table out of my ECUtek tune, but pulled a couple degrees out of the low load cells, I was running as much as 40-41* at 30-40kpa in some areas, but am going to hold off on that until I get knock set up. Knock settings with a 2.5-swapped NC are kinda weird. If you're using a 2.0L knock table, you'll sometimes get false knock readings especially around 3k rpm before the engine's fully warm. I experienced this more than a few times, pulled timing, and proceeded to get the same knock retard values, albeit intermittently. Again, I still gotta set up knock on this ECU, but I'm going a little easier on the timing in that area until I do.

I might add a bit of timing to the 70 and 80* rows as well, the dropoff there is kinda aggressive.



Aggravatingly, my VVT settings are still unfinished due to ECUtek hardware woes. I was going to copy my tables from that tune, but my ECUtek USB key malfunctioned and can't be recognized by the laptop anymore. This is only five months since I paid $150 to replace the last one that did the same thing. This time it should be warrantable, but it's still annoying as hell. Glad to be done with that stuff.

I did my best to remember what I was running with my previous tune. The Fab9 intake manifold likes a lot of advance up top. Going from 5 to 15* while NA resulted in notable gains, but we messed with all sorts of angles on the dyno after I slapped the turbo on and the engine didn't seem to care what amount of advance it was given while in boost.



A couple other little bits. If you're gonna run an NC on this ECU, the preset Bosch IAT sensor settings are the same as the NC's IAT integrated into the MAF sensor. This CLT table seems to be working spot on as well. I should probably enable failsafes on both though.



I also got auto-blip to work, although it's a bit more annoying to use on the street than ECUtek's option. I didn't realize it, but ECUtek would only blip the throttle if the brake was pressed BEFORE the clutch was. The way I'm running it on MTune, the ECU will blip whenever both inputs are pressed. So if you have the clutch pressed halfway in and then you hit the brake, it'll blip and buck you a bit. Can't find a way to require the brake to be pressed first then the clutch switch. Aside from that, it works well and should be fine on track for the one race I'm gonna do before slamming the 8HP in.





More fun to come. Next week, though. Headed out to Thunderhill Thursday morning to race Lemons with @rb92673!

 

timers:
Set Internal Ouptut using brake to start the "brake timer" IF brake is pressed but clutch is not pressed. (A!B)
Set Internal Output using clutch to start the "clutch timer"
Set a user table of brake timer vs clutch timer.
Set Internal Ouptut using user table to trigger blip

This way you can set at what times the clutch and brake have to be depressed for the kick to happen as a table.
like if clutch is depressed for 1s then you brake, you probably don't want the blip.
But if you clutch for 0.1s then brake... might want the blip.
Or if brake is depressed for 15s then you clutch... probably don't want it.
But brake for 2s then clutch, that's likely valid.

Also, you could 4d table that user table with vehicle speeds. so you don't blip if you're under 10mph.
Or you could change that 4d table source instead of speed to be in what gear you are and set times for brake/clutch for specific gears (though this might be a problem if it calculates the gear position wrong due to RPM change with clutch press)

 

for info share.

Re: o2 feedback - this is actually a feature OEMs use. On the turbo 4 engine I used to do calibration testing on, we could get some individual cylinder fuel control feedback from the wideband o2 after it went through the turbine. Of course it had much worse signal:noise ratio than N/A engines after effectively going through a blender, but the software and dyno wizards did make it work in a small window of the speed/load range.

 

Thank you for interpreting my post as a cry for help. That's what it was

Amazing, I'll give that a go later. Knew there were probably one or twelve ways to get the intended result there, just didn't know where to start. Once I'm in there, I'll see if I can figure out a way to work RPM into the fold as well. Another issue I'm having with it is that it'll blip if RPM is raised above the threshold as well. I.E. if you're slowing down below 3k rpm, heel-toe to rev-match, and rpm exceeds 3k, you get a blip. The guidance you listed above is great, I'll try to figure out something for that while I'm in MTune messing around later.

That's the basis for your P219A-P219F trouble codes, yeah? I was really surprised the first time I saw a car throw one of those codes when I was in school. Actually couldn't fathom the O2 was able to identify individual cylinder feedback like that.

 

Man, do you still have all your 5-speed stuff? We talked on IG a while back about it. My NC1 six speed just completely locked up on me and stranded me today. I just got the car and registered it earlier this week and it is already broken…

Anyways, please let me know if you still have all your 5 speed stuff, as I really want to swap to it!

 

Yo dude! Yeah I remember talking to ya a while back about it. Sorry, but I ended up selling the transmission. I still have the clutch and driveshaft laying around collecting dust if you want em for dirt cheap, though! PM me.

Lemons weekend up in NorCal with @rb92673 was absolutely stellar. The trip had about a million firsts for me. First time racing any sort of wheel to wheel, first time driving on track at night, first time driving hour-plus stints, first time driving at Thunderhill, first time doing a 10pm transmission swap in the paddock, and so on.



We had planned to run the second half of practice on Friday afternoon from 5-9:30PM so that everyone could get acquainted with the car and maybe get some practice in the dark. That went sideways pretty quick as it was discovered the 6 speed we swapped in a couple weekends ago didn't want to shift at speed. Almost simultaneously, the engine also developed a misfire, which we narrowed down to a failed #3 fuel injector. After we swapped the injector, I went out on track for 4-5 laps and confirmed that the trans only shifted marginally better after getting some heat in it. That being said, I was pleasantly surprised by the Ecotec swap, which felt very similar to my N/A 2.5, just with the powerband lower in the rev range (I think the shift light was set to 6.3k).


Anyways, it was decided that we were better off swapping in the old 5 speed for the race. We went through light tech at 8PM or so, then got to pulling the transmission around 9. It was much less of a bummer than I expected it to be. The garages were bustling with activity Friday night, so we had lots of people to hang out and chat with while doing the job. Having four sets of hands made everything move quicker too. We ended up doing the job in about two hours and eight beers with plenty of messing around. Ended up pulling the clutch off for inspection and replacing the pilot bearing too, which fell out when the clutch was removed.



With the five speed in place, the car was ready to rock for Saturday. I was equal parts nervous and excited knowing there were going to be 110 cars running simultaneously on the 3-mile course. RB had originally told me he'd probably put me in for the third or fourth stint, as the first and last stints were usually filled with the most mayhem. Saturday morning, though, I was told I'd be running the first leg of the race, with a short stint from 11AM-noon. Sh*t, ok time to spread those wings, baby I gridded early, being about the tenth car out of the pits. Got a handful of parade laps in before hearing RB call out "green flag" over the radio. I had only the slightest clue of what I was doing, and didn't want to do anything too dangerous with 25 hours and one minute left on the clock, but I hit that gas pedal and started trying to go fast.

Within ten or fifteen minutes, the nerves faded a bit. Almost all of the other drivers on track were aware and predictable, and setting up passes was easier than in TT events. I was still driving conservatively, but wasn't absolutely terrified as I had expected to be. The fact that the Nyan Cat ecotec miata was one of the fastest cars on track helped too. This made for an interesting/fun dynamic which largely exposed my TT/HPDE background. Battling a similarly-classed car, I found I could put a gap between myself and the other car(s) most of the time if there were just two or three of us. However, when we came up on packs of slower cars, those same drivers would more often than not close the gap and get around all of us while I was still trying to weasel my way through. It was both interesting and a ton of fun. I got a lot better by the end of the first hour but still ensured not to take any big risks despite all the fun that was being had. Pulled into the pits at the end of the first stint with a big sh*t eating grin.


The first few hours of the race went well until we had a rear brake caliper blow a seal at the parking brake arm. We cut the arm off with a cutoff wheel and tried having the area welded over twice but weren't able to get a good seal on it. After a couple hours of struggling with it, the decision was made to pull the caliper entirely and cap the line. The MK60 ABS was going to have to work overtime as we'd be running the next seventeen hours or so on three brakes lol. The lack of one rear caliper ended up slowing us down a bit but wasn't as catastrophic as expected. Like RB said in his post, we still had probably 80% of full braking capacity.



My next two stints would be from 10PM to midnight and 3AM to 5AM. I was equally excited and nervous to race at night for the next sessions as I was for the first session.

At 10PM, I pulled out of the pits, tried to rapidly acquaint myself with driving on track at night, had a few fun laps running around in traffic, then promptly pit-maneuvered another car in a long sweeper and sent us both into the dirt. Utter embarrassment and dread. Guy left the door open for me to pass (or at least I thought) on the inside, mid-corner, and I misjudged everything from my rate of rotation to trajectory to position on the track and glided outside right into him. Luckily the other car wasn't damaged (well it was already a little damaged, it was a Lemons car) and I was able to get the miata back into the pits where RB and one of our teammates banged a dent out of the fender and bent our light bar back into place. I went back out from 11-ish to midnight with the fear of God instilled in me and came back into the pits still feeling a little shaken. Then it was time for a quick nap as the new plan was for me to drive a 4AM to 5:30 stint.

4AM came around and I got back out on track still a little shaken, but feeling about as good as possible for running on almost zero sleep. I was tossed out into a group of cars that was about my speed, and we proceeded to run several laps together, all on pace with each other. Such a sick experience rolling around a track in pitch black with a bunch of cars all wrapped in silly colored lights. Then at 4:45 or so, the sun started coming over the horizon. That was icing on the cake. Much appreciation to RB for tossing me into that time slot. By 5:10 it was light out and I got to run my last seven or eight laps being able to see the track again. I definitely improved at night driving but was still a hell of a lot better at driving in the daytime haha.

Handed the car off to our next driver at 5:30 and rolled over to the cafe area to watch for a bit. The lighting outside was amazing, or maybe I was just sleep-deprived and dehydrated.


We'd end up having one more mechanical, the fuel rail to AN adapter clip phased out of existence which caused the car to stall on track around 7 or 8am. Luckily we had a spare on hand and took care of it quickly. The car made it across the line, still running, at 12:02PM on Sunday. I don't think any of us knew where we placed in our class, just crossing the finish line felt like a win lol.


Dang, that was a way longer recap than I meant to type out. Shoutout to anyone who read this far. Back to NC stuff this week. Got a couple things to take care of before the next race at Buttonwillow in a couple weeks, then it's 8HP crunch time starting the weekend after.

 

Oh hell yeah. Enduros are always a blast. I've never been able to do one at night but being able to run wheel to wheel (ish) with a track packed full of cars is always a blast. Seems you got the full Lemons experience with on track incidents, major repairs the day before, and a few "minor" issues while running... It does seem to be mandatory for Lemons 😆

 

Zak failed to mention that he had the fastest lap time on the team by 3 seconds. Sorry Zak, I changed my mind, you are no longer welcome on the team because you beat me so badly.

 

If you have a turbo you need MAPdot enrichment as well. Both TPSdot and MAPdot should have de-enrichment also to avoid rich spikes.

A fuel flow rate derivative (perhaps normalized) might be an effective way to combine TPSdot and MAPdot, but haven't tried it.

-Chris

 

Lemons sounds like a blast man, glad you got to go out and do that!

Also Watson, interesting point on MAPDot, I haven't though about using that and TPSDot, but sometimes driving on bumpy roads my MAP swings around like crazy and causes lambda to fluctuate. Not sure if this is what you're suggesting, but a tiny bit of MAPDot correction might help smooth that out

 

Turbo spool at WOT is a simple case to consider. Fuel demand goes from atm to full boost with TPSdot=0. TPSdot is still useful as MAPdot lags it slightly. IIRC I run about 50/50 from each source. Another common case is sudden load change at idle where a closed loop idle uses timing before throttle to make a change (also a reminder to use TPS or TPS target (if DBW) rate and not pedal position, which doesn't change at idle).

That said, MAP changing due to bumpy roads sounds like a mechanical problem?

 

Might be a mechanical problem with my foot. Interesting concepts though, hadn't considered any of that but it makes sense.

 

Yo, thanks all re: Lemons! It was a party for sure. I feel like the experience would've been lessened without the trans swap, mechanicals, and other drama lol.

Regarding the talk above, I'm gonna go back into my tune later and see what I've got for TPSdot in the AE settings. IIRC, I may be only running MAPdot. I've got an issue on cold starts where the engine will go lean and lag for a second or so upon throttle tip in from a dead stop. It starts to go away once CLT hits 120-130*F. My hunch is that it needs some sort of extra enrichment while cold. MAXX doesn't seem to have many settings for E-throttle-based idle. I just set min and max throttle trim along with my target rpm and it hits/holds it well. I haven't really tried to figure out the cold tip in issue because it's a minor annoyance that goes away quickly and I'm still focusing on dialing in my fuel map at higher boost levels. I'll take some logs and post settings later.