3rd piston shut off

[Mr Whippy]

I did see a measurable improvement on a stage 1 map, but as you say, I think it was down to better tractability at low revs and the mildly increased rail pressure below 2k rpm, which meant I was able to shift up earlier without straining the engine.

I've boosted rail pressure at specific cruise rpm, and done movements of injection advance/retard at the specific injection amounts at the specific cruise rpm too, and in all cases of steady state testing my results all seem to sit in the 0-3~4% error that comes up due to testing any way.

In many cases I think, hmmm, more advance = better torque, but higher EGT maybe, which increases temp of intake charge on a HDi90 especially = lower power again.
High rail pressure by 10%, great, 5% better fuel burn, but 5% of the 30bhp needed at a 60mph cruise is about 1.5bhp, so you can trim throttle a bit, but the rail pressure boost of 10% might cost you 1bhp? So all in the benefit might not be there?!
Really we are talking very small benefits that culminate to a reasonable amount all working in unison. But until you can really see the benefits in an honest statistical way, reproducible again and again, then you can't really make any useful changes in my view.
There is a good chance you make 2% somewhere, worth doing if it's there, but another tweak might cost you 1.5% and you'd never really know it.

That isn't to say we shouldn't look for the benefits, but it's REALLY hard when our gear is gonna be nowhere near as sensitive as OEM's with engine dynos and controlled environments


I do keep thinking of ways to improve accuracy but it's all subtle stuff. Ie, 10% better accuracy is still 2.7% noise magnitude rather than 3% noise magnitude in my results.



PCMScan and ELM are deffo good tools for this job though!

As long as you do a baseline before any mod, and then mod it and repeat the test afterwards, things should show up in the results if they do make real net benefits!

You don't even have to run both ways down a road to account for wind. As long as you always do the same thing, or as close to it each time, your results are as good as you will get them in real world testing!


Ie, these are some G-tech results from my 306 HDi FMIC tuning development. Ignore the values, they are not really relevant as this is before/after testing, the important thing is how consistent my 'before' tests were... that is easily as good as a dyno imo!

Imo if you can get a grouping that tight for mpg over 5 runs, and then do something like drop tyre pressures 10psi and run again (something you know will make a difference), and run again, and see a grouping of 5 standing distinctly apart, you know you are gonna have a chance of seeing these subtle changes you are trying to find!


I think it'd be interesting to test before/after with the spoiler on the 306 for example, or that front splitter, or the splitter/under-tray combo. I bet these things could be surprising when it comes to economy at a 60-70mph cruise!

Dave