Al,
Thank you. I wouldn't have even tried to do this conversion project if you (x-faktory.com) hadn't put together and offered the conversion kit along with excellent documentation and email support answering my many questions. Thank you. Dan
Part 10
Now for a tuning update:
In Part 9, I described how I had found some driving conditions on a run in the hills near where I live that were causing problems with balking and hesitation. Over the last 2 days, I have spent time with my PC connected to the ECU looking at the conditions in the tune file.
The first area I needed to address was the acceleration enrichment (think accelerator pump on a carburetor). The tuner at the dyno shop had used a feature called "TPSdot Uses %WOT" which enables a curve to model effective 100% TPS for a given RPM. This allows smaller throttle movements to be more responsive at lower RPM's and ignores changes in throttle above "wide open". The problem was the curve was set so tight that at idle the TPS Accel Enrich would kick on and off and caused the idle to be erratic / not steady. I played around with the values in the %WOT (percent wide open throttle) table, but I just couldn't get it to settle down - I think the root of the problem has to do with noise on the TPS signal. The TPS gauge on TunerStudio app also has a digital readout of the current value. I watched the TPS gauge with engine idling and TPS at 0% and the digital readout jumped around from -.5% to +.9%.
Given the environment in the engine bay (heat, vibration, etc), some variation in the sensor outputs seems normal, so I turned off the "TPSdot Uses %WOT" option and started playing around with the Time-Based Accel Enrichment settings. First, I added MAP based acceleration enrichment so now the acceleration enrichment is a 50% / 50% combination (this is adjustable with a sliding bar) of changes in the intake manifold pressure and the throttle position. I made this change so the acceleration enrichment wasn't relying solely on 1 sensor's inputs. Then I set the threshold values for MAPdot (kPa/sec) and TPSdot (%/sec) and the values in the "Added / MAPdot" and "Added / TPSdot" tables so that the TPS Accel Enrich and MAP Accel Enrich functions would momentarily become active (the injector pulse width jumps from 2.5ms to 10ms then returns to 2.5ms) when I whapped the throttle, but would not be active while idling and during warm up. If the engine hesitated when I whapped the throttle, then I adjusted values until the engine revved up smoothly with each whap of the throttle (Whap = sudden opening then closing of the ITB throttle butterfly valves).
Next, I got in the car with the PC on the passenger seat still connected to the ECU with the TunerStudio app running and started the Auto Tune function and began driving. I literally headed for the hills and started up some steep grades with revs in the 1800 to 3000 rpm range, accelerating then decelerating and all the while the Auto Tune program was updating the values in the VE table (volumetric efficiency table - tells ECU how much air is going into the engine for a given rpm and intake manifold pressure) so that the actual AFR (air fuel ratio) as measured by the O2 sensor matches the target AFR in the AFR table. And joyfully there was no more balking and hesitation, the car pulled strong each time I asked it to. I drove the hill loop twice and each time the car ran well. I captured a couple of screen shots from my PC:
I drove back home, burned the changes to the VE table on the ECU, and put the car away in the garage. I went back to my PC and from the TunerStudio app I used the compare tune feature to look at the values in the VE table before and after running Auto Tune.
What I found was most of the cells with changes had lower values after running Auto Tune which means the VE table was over estimating the air going to the engine so the AFR was too rich.