Julian Edgar, 50, has been writing about car modification and automotive technology for nearly 25 years. He has owned cars with two, three, four, five, six and eight cylinders; single turbo, twin turbo, supercharged, diesel and hybrid electric drivelines. He lists his transport interests as turbocharging, aerodynamics, suspension design and human-powered vehicles.
The benefits of road testing modifications
Yesterday I did some serious testing. Well, ‘serious’ as in I set out to specifically test and fine tune the air/fuel ratios in my project car.
Despite frequently writing that you should always have an assistant in the car to read gauges and hold hand controllers and make tuning decisions, I must admit that circumstances often force me to do it all myself, as I am driving. (I am cognizant of the dangers involved in doing this; once many years ago while on-road tuning, I ran straight into the back of a car that had unexpectedly stopped.) So yesterday, for the fine tuning, I did in fact sit in the passenger seat. My lady drove the car and our little boy was in the back.
Despite needing to use full throttle as part of the tuning process, the performance of the car (a turbo hybrid Prius) is so slow that it can all be done on open public roads without breaking the law or posing danger to anyone.
While setting-up this car I have at times run numerous in-car instruments, but on this occasion they were limited to a MoTeC air/fuel ratio meter running from a tail-pipe mounted probe, an LCD intake air temperature display, a LED indication of factory oxy sensor output, and an LCD hand controller connected to a Digital Fuel Adjuster kit. I also brought along a screwdriver and spanner to allow me to adjust the high and low fuel pressure regulators (this car uses a system that switches out the closed-loop oxy sensors and simultaneously switches in higher fuel pressure). A smaller screwdriver was also carried that could be used to adjust the switch-over point between fuel pressures, a change which is triggered by a Simple Voltage Switch kit monitoring airflow as measured by the airflow meter.
The first step was to check intercooler efficiency. While I had previously measured intake air temps, I’d only done so with the front bumper and number plate off the car. I’d physically felt the temperature of the plumbing after driving the car hard, but hadn’t quantified the numbers. And I must admit, watching the LCD temperature display, I was appalled. Initially, on the long trip down from the hills on which we live, the temp had stayed low – about 10 degrees C above the 20 degree ambient. But then, whenever boost was called upon, the temperature rocketed. Like, it would rise to 20 degrees C above ambient after just a few seconds of boost, and 40 degrees C above ambient after perhaps 10 seconds of boost!
I looked glumly at the display, wondering whether the presence of the bumper and number plate in front of the ‘cooler could really make that much difference. Or was it the fact that the undertray was now back on the car – I’d have thought that would’ve improved intercooler performance, but perhaps not…
Once we reached our empty country road test location, I got Georgina to drive at full throttle from 40 km/h to 100 km/h. The intake temp rose to 67 degrees C. Ouch. I then took off the front number plate and we re-did the test. Same temp….
But something was wrong – there was no way that my previous reach-down-and-feel-the-pipes hand test had shown intake temps anywhere near this high. I popped the bonnet and saw the problem. Duh! I had fitted the temp probe to the plumbing before the intercooler, not after! Gee, that explained a lot. In fact, it made for a really good picture regarding turbo efficiency in this application – if the max temp coming out of the turbo was only 67 degrees C at 7 psi boost, the turbo was working well within its areas of peak efficiency. Out on the road I couldn’t easily swap the probe position over, but again the ‘hand’ test showed intake air temps were happily low.
Hmm, OK, so that part of the morning’s testing hadn’t actually been testing what I thought it had been!
Now for some air/fuel ratio tuning. For the previous 3000 kilometres, since the turbo install had been finished, I’d been running a full load air/fuel ratio of about 11:1. That’s pretty damn rich but it’s also a suitable mixture for shaking down any problems. For example, if the car had a propensity to detonate, or if the fuel system somehow gave problems, better for safety to be running richer than leaner. But there had been no problems, so now it was time to lean out the full-load mixtures a bit. I adjusted the high load regulator and fluked it – in one go I got the full-load mixtures back to 12.5:1… the ratio I’d been aiming for.
However, the switch-over point to high fuel pressure occurs before full load, and in between the switching point and max power, the mixtures were a bit richer than needed. I then used the Digital Fuel Adjuster (working on the airflow meter output signal) to pull these transitional mixtures back a bit. And after only about 30 minutes of on-road tuning, the mixtures were as desired – stoichiometric in closed loop, 13.5:1 immediately after the fuel pressure switchover was made, grading to 12.5:1 at full load.
We drove around for another four hours, visiting some shops and doing some sightseeing – I wanted to check that there weren’t any urban driving situations that would catch the system out. All looked fine and so we headed home – up the big hill we’d previously descended.
It’s a very big hill, and with the air-con on and 2.5 people in the car, it’s a hill that in this car needs full throttle. And here’s where I found a problem. If the Prius is baulked by a slower car, its maximum speed up the hill is less than usual. In fact, the speed can be dragged back so far that the engine airflow signal drops below the fuel pressure switch-over point, resulting in a return to stoichiometric mixtures… even at full throttle. (There’s no way of changing gears in the Prius.)
The way to overcome the problem is to set the hysteresis of the Voltage Switch to a higher value, so that once tripped, the load signal has to drop by a greater amount before the switch again turns off. Or at least I think that’s how to fix the problem – I didn’t come up with the possible solution until later and so I haven’t had a chance to try it out yet.
But I think the significant point of the day’s testing was that on a dyno, or with road testing that was quicker and less broad, I’d never have found the problem. After all, in over 300 kilometres of test driving on secondary roads, freeways, suburban man roads and back streets, the problem occurred only once, on one piece of road that provided unique demands.
It’s always good to do lots of road testing…
