But once I pulled the Hyundai out of the garage so I could work on the next series of videos of a different project, the Hyundai's troubleshooting continued. Outside. In 98°+ weather. It started with, "why are my fuel trims still all discom****ulated?"... then I found the spade connectors I used on the coolant temperature sensors were all loose and intermittently open-shorting... which is another reason why the fuel trims are right one minute and wrong the next. Coolant temp is part of the fuel trim calculations. After crushing those so that they grab the electrical connections tightly, the tune stabilized slightly. The log image above is what my idle was initially doing after that fix.
There's lots of stuff wrong with that idle. For the tuners in the crowd, you can look at my combined fuel trims and see most of what you need to know about how far off my tune was. 23.6% is terrible. That value should be closer to zero. Looking further, AFRatioEst vs. my LC1WB (Wideband o2 sensor) values are WAY off. AFRatioEst is a calculation, so the only way it should be that far out is if my airflow and/or fuel settings are wrong. If we want to exhume this casket and see the bones, note the airflow-per-rev value is also pretty far off. it's at .23 g/rev instead of .25... and the STFT, LTFT LO, MID and HI values are non-zero numbers. Ignore the ISCPosition value for now, that one is fixed, and there was just a bunch of electrical crap running. I've done a video on that one before.
So what do I make of this? My injectors were lacking deadtime. I adjusted that from 360µsec to 430µsec to fix AirFlowPerRev and bring it to .25. I added about 10% fuel to the idle cells of the airflow table. By adding airflow, naturally the ECU sees more air and squirts more fuel. The math starts to work and the STFT values zero'd out, or I should say that they fluctuate from ± 2.5% around zero. I'm displaying averages from a large range of time when I was idling here. I broadened my idle settings a bit in the airflow table and brought the 50hz MAF range (LTFT LO) up a few ticks in the airflow table and quickly watched the LTFT LO zero itself out after a minute of sitting still at light throttle. With the airflow and dead time values closer to correct, the AFRatioEST value now closely resembles what the Wideband Oxygen Sensor is reading. That means the ECU's math is starting to work out. Idle was rock solid and stable. Below is an example of what I ended up with after about 2 hours of tweaking settings and then sitting and watching them take effect.
I can't help but point out how much better this is for someone who considers himself to be a bit of a novice with tuning. Note how close LC1WB (metered a/f ratio) vs. AFRatioEst (calculated a/f ratio) are... Note how close STFT, LTFT LO/MID/HI are to zero. Look at how close to zero the CombinedFT value is, it's barely making any corrections at all. AirflowPerRev is .25 where it should be. This is the way you want your idle to look. But beyond the "look", when you get these values to mimic these, there will be a HUGE difference in the way your car FEELs. What happens when you coast off-throttle... what happens after you blip the throttle in neutral, what happens after you turn the key and start the car. All of these corrections will make your car run more stable and predictably out on the street. They'll prevent you from having to blip the gas to keep the engine running or babysit the gas pedal on a cold start. These don't sound like big issues to deal with, but the less of it that you have to do makes the car FEEL much better because it takes less of your attention to operate it properly. That should be your idle tuning goal. A manual-transmission car will spend more time off-throttle on city streets as it will in on-throttle conditions. So if you can get your idle tune nailed down tight, you've completed more than half of your tuning. Look at it that way.
I understand tuning from a basics perspective, but certainly not from an engineering or professional auto technician's perspective. I know what I want to accomplish, it just takes me a little more time to get there. This is the reason why I haven't made videos about tuning, but as time progresses this might change? I would prefer that people get their answers from the professional auto techs and engineers, but I recognize there's a gap between the terminology they use and the language the typical automotive enthusiast speaks. Perhaps this is the kind of information I can arrange in a coherent way to help bridge that divide? The minute someone goes over your level of understanding, your eyes tend glaze over, the yawns begin, and you start thinking about, "did I feed the cat?", "Did I turn the oven off?", etc... You may even truly want to know what that technical person is talking about, but focus becomes a struggle when you don't know the words, the whats, and the whys. I sorta feel like as I gain a better understanding of tuning, that I'll be better equipped to speak on that matter. I probably lost half my readers by this point anyway.
There is progress on the GSX block and videos. Less than I would want there to be right now thanks to some random person that left a germ trail and gave me a cold that I've been carrying around for the last two weeks. I did manage [with enough cold medicine] to do a collaborative video with Joe Bucci (go check out the tech video at Joe Wrenches), but my voice is rekt right now, preventing me from finishing the voiceover on the video I have. I hope you all get a chance to drop in on Joe's channel and share the love. In case you didn't know, back in the day... Joe ran a 9 second pass in a 1g with a 14b turbo. He's got skills. Check out his current racer project!
If you haven't seen Joe's tech phone call video, it goes like this... Tom trolled Joe in a video, gave out his phone number, and I helped Joe troll Tom back. Tom didn't know we were doing this. Joe and Tom called back all the people who left messages, though. That phone call between Joe and I was a real phone call. We couldn't resist the opportunity and it was all in good fun. I confess that I embellished it a bit with some video trickery so that you guys would know it was me. I respect all other DSM channels because they're helping to keep our spirit alive. We're all different in how we produce and deliver our content. The more people share about their own approaches to their DSMs, the broader the palette is for us to enjoy and learn from. It's for this reason that I will support all things DSM, and that includes the channels run by other great DSMers. This isn't a competition after all. DSMs are always community-built cars even if only one guy turns the screws. No man is an island. That man learned which screw to turn because someone else turned it and posted about it somewhere.
So because today's forecast is rain, rain and more rain, so the track won't be open in the rain for me to make a pass with the Hyundai... I'm going to take a stab at completing other video projects this weekend, and for those of you who did see Joe's video collaboration where I said I'd see him at the shootout this year... yes I am really trying to go to the shootout this year. That is also one of my goals. You heard it on Joe Wrenches first unless you're just now reading it here for the first time. Going to the shootout. Yep.
I hope you enjoyed the little tuning discussion. I drove the Hyundai to work today and it was a grin-inducing kick in the pants to drive. It's running better than it ever has right now. It's ready-to-race, so I'll bring you some drag videos the very next time the track is open (weather permitting). Or you can come up to the shootout and see it yourself. ;)