I'll check in with the welder when I speak to him and ask what he did for prep, etc. I remember him giving me the rundown on how he prepped/executed the turbine housing weld, but it kinda went over my head and I wasn't really listening because I was in a rush to prep the car for that weekend's race.

If your friend wants some side work, I can send him my next hotside failure lol.

 

 

Welding to cast is always fun. Pre heat and controlled cooldown are exactly what I did when welding uprights on my Jeep. I'm sure a pro welder is all over it though.

All this work to get the stock ABS working? C'mon, get an MK60E5 in there. It'll take you like 4.5 hours, tops.

 

Yeah I'm hoping my guy can handle it. That lead I had on a manifold was a bust, I asked Michael at Kraken what the lead time and cost would be on a spare manifold and I imagine he's busy as hell and it'll be a hot minute before I get that answer.

Don't you put that evil on me, Ricky Bobby. I'm already 12-15 hours into this project with god knows how many to go until I'm banging shifts with the 8 speed lol. I can't stomach the thought of additional work.

On the coattails of the ABS discussion, I got the ECU wired up into the car 99% of the way and flipped the key on tonight. The MAXX isn't wired into the CAN yet, but I still have the secondary factory ECU connector plugged in which controls the ECU's connection to the CAN and other power/ground/relay/body functions. No ABS warning on the dash with the key on and engine off. I'll check for ABS codes via the OBD port tomorrow.

Just about everything's wired into the MAXX now aside from the CAN and alternator wiring, and the CLT sensor ground which I forgot to splice in. Got the ECU up and running tonight and confirmed some of the outputs work as designed. I wired up two outputs to control the fan relays. Output 1 triggers fan relay one for low speed, output 2 triggers relays two and three for high speed. I haven't looked up ideal temp settings for fan control yet. Set low speed to come on at 196*F and high speed to come on at 205*. I also got electronic throttle control working!


I initially turned on the ECU and proceeded to get throttle error codes and limp mode enabled as soon as the throttle pedal was depressed. Double checked accelerator pedal voltages and TP sensor voltages and confirmed all four sensors were reading as they should. Putting the throttle in test mode showed the TP sensors were moving as the throttle plate actuated. After ten minutes of head scratching, I swapped my throttle motor + and - inputs in MTUne and the throttle started working immediately. Not sure if I wired or configured it backwards, but it's good now. I think it'll need some massaging of the PID settings once the car's actually running. Didn't check take any logs yet but looking at the live TP sensor voltages, quick throttle demand changes seem to produce a minor overshoot. I'm just running the preset Bosch 68mm DBW settings in my tune as I know that TB is a drop in replacement for the stock NC TB. All I did so far was set closed and open TP sensor values.



Waverly was in charge tonight, I was just assistant. Don't mind the janky power switch to the ECU. That's just a placeholder until I grab a relay and wire it into an actual switched 12v source (I'm probably going to put that off for way too long).

 

Just get on board the MK60e5 bandwagon.

 

I'm gonna go against the flow here and say don't do the mk60



FOR NOW.

but with your wiring plan on it.
Individual wheel speeds is really cool and you can do really tricky things with them, and an IMU, and DBW

 

You guys are all horrible influences. I hate and love you all for it. Unfortunately, MK60 is going to be on the backburner for a while. I got the car running and leaving the OEM ECU CAN wiring seems to have the factory ABS working.

Wait, holy sh*t the car's actually running, like decent!?


Got the manifold back from the welder after work yesterday and slapped it back on the car last night, along with the turbo, etc. I'm fairly confident it'll hold. If it doesn't I might not even know for a bit, seeing as I was driving around with it cracked for over a week without noticing.



With the hotside bolted back up, I went back to work this morning finishing the MAXX install. The last bit I needed to do was simply wiring up the alternator and confirming base timing. I rigged up a spare spark plug wire to use with the timing light and... forgot my paint pens at work. So I set the crank to TDC on cylinder 1 with my timing tools and marked it with a bit of gold spray paint applied via a pocket screwdriver. Perfect and not sketchy at all.





I went to check for rpm signal in MTune, cranked the engine over and... no rpm. The built in oscilloscope showed action happening on the CKP but nothing from the CMP. I checked my wiring and found that I'd forgotten to splice in power to the CMP. Got power to the sensor and boom, rpm and sync present. Cranked 'er over and holy hell she fired right up and idled smooth!

For about 45 seconds, then the battery went into failsafe mode and shut itself off. I checked voltage and it was drained down to 9v. Gave it a charge, fired the car up again and yep, the alternator ain't alternating. I checked my wiring and outputs in MTune and found everything as it should be, then I dove deeper into the forums. Apparently this is a common problem I'd never heard of, and the NC alternator is a PITA to control via a standalone for some reason. Most standalone guys swap an internal regulator into the factory alternator and call it a day.

F*ck it, I'm driving this car today. Since I was already piggybacking off the ECU for some other things, I wired the alternator back into it as well. Wiring up the alternator, as well as sending CKP and CMP signals to the factory ECU got it charging again, albeit at only 13-13.1v. Some further digging showed that the ECU needs to see IAT to fully enable alternator control. I ordered a pack of 2k ohm resistors to simulate that and will wire one in next week. For now, though, the car was driveable!


Fueling right off the bat was surprisingly close to where it needed to be. The MAXX makes fuel trim adjustments at all times, and you can apply them to your VE table at any time. I took a quick 20-minute drive around where I live and let it do its thing while monitoring for anything that seemed off. Didn't really need to do a ton off the bat except let the fuel adaptations do their thing. I'd set my minimum throttle angle at idle to 2*, which was a bit much and resulted in a 1,200-1300rpm idle, but aside from that the car just... behaved while cruising around. I'm definitely gonna have to tweak the throttle tables, as they're super "soft" at low pedal demand. Also still haven't set up VVT or knock sensor settings, or taken the car into more than one pound of boost, so that'll be on the list this coming week as well. And then there's the nest of wiring that needs to be shortened and cleaned up. Plenty to do, but I'm very smitten for the time being.

 

What do you mean you don't want to swap the car to a standalone ECU, while putting a transmission from a Jaguar into a Mazda, AND yanking the ABS out of a BMW all at the same time? Lol

Nice to see how good the car seems to run almost right off the bat on the MAXX.

 

Might as well just yank everything out of the car and put a V8 in there while I'm at it if I'm going that deep lol. And yeah also no joke with the MAXX doing its thing. I think I just got lucky with some of the things working right out the gate. Can't take credit, I took lots of advice and borrowed bits and pieces of multiple peoples' tunes (albeit only one tune was from an actual running car lol).



Took the car to our Mother's Day celebration today and continued tinkering with the tune. Couple crazy things right off the bat. Idle control is somehow better than the factory ECU, maybe because the factory strategy was never meant to deal with big fatty cams? Either way, it hits the target and idles smoother than the OG ECU did, and I really don't miss the ECU needing to relearn idle for 15 minutes and having it jump around and nearly stall after every reflash. After tweaking with the throttle tables a bit, throttle tip in is smoother than I was ever able to get with the factory ECU as well. There was always a very small dead spot between 0 and maybe 2% throttle, almost unnoticeable, but kinda annoying. That's gone now. Also gone is the annoying pop from the exhaust that happened every time (every single f*cking time) the car exited fuel cut overrun. No tweaks for that in ECUtek so I was stuck driving the car like such ever since I installed the turbo on it.

I used the autotune feature and also manually massaged the fuel table enough to get some pulls in today up to 7psi. The fuel map in boost still needs more tweaking but is getting close. Currently, I have no VVT action when commanded, and am assuming that's because I forgot to wire power into it as well. The lead on my test light at home is too big to check the connector pins, so I'll confirm that later this week.

I played around with accel enrichment a bit as well, just because I figured I should. Can anyone confirm I'm going in the right direction here? I noticed I'd get a super quick lean spike on the wideband gauge (didn't register on the butt dyno, though) going from part throttle to WOT quickly. Checked a couple logs and it looks like the spike occurs due to the enrichment not lasting long enough. I was just running the standard accel enrichment settings in MTune, so it made sense that there would probably be some tweaking necessary. Base settings have the enrichment duration set at 2 injection events with a 50% decay rate. I incrementally adjusted these to 4 events with a 35% decay rate and it seemed to take care of half of the lean spikes, but not all of them. More testing/tweaking needed but more duration needs to happen based on these logs, right?


Example before increasing AE duration.


Example #1 after increasing duration. Successful.


Example 2 after increasing duration. Better but still getting a lean spike.

 

For reference I hold 10 events with a 50% decay. Is that wrong or right? Idk, I just threw some number at that untill it didn't stall or buck during throttle changes.
There is an entire table below that for % fuel added based on RPM and TPS rate, so you can also up those numbers to your liking.
Personally I don't sweat transitory lean blips like that, it's not long enough or loaded enough to hurt anything, as long as the car isn't hesitating I rock it. While I've kicked many rods though blocks, I've yet to melt a piston with these half assed methods.

 

AE is just as much art as it is science. You will always have some amount of lean or rich spike during a quick throttle event, so AE is best tuned by feel. Maybe @redursidae has some input based off those logs tho!

 

Awesome stuff dude! Very impressive and glad to hear it’s going fairly well. I’m reading every post and learning as you go

AE is always a do it by feel area. You’re on the right track with a longer decay for your engine, but don’t expect to get rid of the lean speak while achieving the best feeling result. On single throttle engines with injectors close to the valves you tend to have relatively lower percentage of enrichment with a medium tail of decay (so increase the decay events). If it feels good and the lean spike only lasts a few engine cycles then it’s ok. Since it sounds like it’s only happening on bigger stabs you can also try increasing the higher rate areas of the table some.

If you get rid of the lean spike with normal AE it will likely run rich and feel like it’s chugging, at least in my experience. You would have to use a fuel puddle modeling approach and MAP estimation to have perfect AFRs and good feel. Even then it’s unlikely it would be 100% perfect in the data. One thing to remember, as Curly pointed out in Nate’s RX7 thread, is that any time you have a large jump in pulsewidth you’re asking for a lean spike. AE is basically doing that.

 

I'd just echo what everyone else said. IMO if you can't feel the AE then it's doing it's job. Personally I usually add way too much fuel, pull some out until I can feel the car stutter when AE is active and then add some fuel back in. Like Wingman said, I'm not sure that's the right way to do it, but my pistons are still in one piece, and with how stabby I am on the throttle I think they would've melted by now if it was going to cause an issue.

Congrats on the first drive. Always interested to see people's experiences with other tuning platforms. Glad to see it's already improving the experience. Honestly I wonder how much you'll notice going from Megasquirt to Maxx. I was blown away with the G4x when I turned on any of the closed loop corrections and they worked perfectly out of the box. Coming from Speeduino it was a walk in the park to get the car running properly. I'd wouldn't be surprised if going MS to Maxx is similar

 

Wow, three four perfect and differentiated responses. Thanks guys!

Aight, I'm at ease regarding the lean spike then. Honestly, I couldn't even feel it before I increased the AE duration, and was more just thinking of things to attend to while messing with the tune today. My previous setup with the stock ECU and MAF-based fueling did just about the same thing, but ECUtek doesn't have any AE options, so I just rolled with it and didn't worry. I'll probably increase my AE events from 4 to 6 just to see what it does, then move onto the next thing.

Actually, I don't know how many other things there are tuning-wise. Gotta set up VVT and knock for sure, but otherwise it just needs some more VE table-massaging and I gotta tone down ASE a bit. Engine wants to run at like 10.5-11:1 for longer than I'd like it to when I cold start it. The stock ECU would target 14.7 basically from the get-go and I never had any issues with it.

I thought I was gonna need more DBW tuning after watching my throttle angle overshoot on quick transitions while messing with it last week, but that actually hasn't been an issue. Before starting the car, I changed the D value in the PID settings from 100 to 150 and that worked well enough for me to forget about it. Throttle feels responsive but doesn't overshoot. It might be a little more damped than the stock ECU setup. I'll probably have to revisit this after driving the car in anger, actually.

Sim, night and day difference for sure, although it's been so long since I had my MS that my memory may be a little fuzzy. I remember having to mess with PID settings on my MS for most closed loop functions like you said. So far this time around, the only PID setting I've tweaked is the DBW stuff like stated above. EGO correction/fuel trim adjustments seemed to nail their targets right out of the box, same with idle (I haven't even looked to see if there's PID settings for this because so far it's just worked fine). Pretty cool. After this, I'll have to look into some of the more advanced stuff as I'm barely scratching the surface of what the hardware can do. Similar to your journey with the Link haha.