Injection - Tuning up-date, 14.03.03
This program of development embarked upon by myself and colleague Mike Barratt (MBE, Lincoln 01522 545189) was instigated by Keith Dodd, owner of Mini Spares centre. His instructions were for us to not only sort out something suitable for the already injected cars, but - if possible - apply that to non-injected cars. The idea being Mini Spares would be able to offer injection kits to boost the already injected Minis and progress right up to a system with race application potential for non-injected cars. Quiet a serious amount of work involved then considering as much of the componentry needed to be useable in various applications. So Mike and I bent our collective experience and minds to the task.
I'll deal with the kit for the already injected cars in another up-date since we have not managed to conclude on this area yet. Still having fun trying to run an engine on the dyno without using that bloody immobiliser. We know there are ways to achieve this - but it means having an immobiliser out of a car complete with correct key fob and so on. We are determined to do it without.
So, to the injection kits not using the standard OE injection intake systems. Two issues stood out here - the fact that so far nobody had managed to come up with specific programming capability for the A-series, 5-port design, and that the injection bodies used were of standard manufacture from companies making them ostensibly for modern cross-flow type engines. The electronics issues were tackled mainly by Mike and his contacts, demanding something quite different and trick were needed to get the best from an injection set-up on a 5-port A-series. True, others have made kits that run on the 5-port, but that isn't the issue. They have rarely proven much better than a well set up carbureted installation. So what was the point in spending a fairly substantial chunk more cash when no real benefits were to be made? And most of the improvements seen were more to do with tailoring ignition 'curves'. Mike also manufactured the throttle bodies manually for testing, employing various machined-in options that could be thoroughly tested on the dyno with having to have a box-full of throttle bodies to swap on and off the engine. And beautifully crafted they are too in Mike's own inimitable way. Part of the brief being that you could use the bodies as singles or twins.
I won't report the full test procedure we went through, eliminating everything we could. We tested all kinds of combinations of fuelling including those we were fairly sure wouldn't work well at all - in the name of providing as full a background picture as possible. Even the stuff that didn't work well provided useful information towards getting the best set-up. We didn't want to arrive at a lucky chance result, we wanted to understand as fully as possible all we could about the A-series 5-port and fuel injection applications. One of the best decisions made before testing began was to run multiple Lambda sensors - this provided some intensely interesting information about the various combinations we looked at. The pre-thinking that went into the throttle body design proved incredibly successful, and the trick programming firmware an essential tool. All in all we were staggered by the results gained during the last test where we tried the set-up we believed to be what the A-series needs. To display the whole test program results would take up several pages and be absolutely of no use to many, so of the accompanying graph are the results of most interest. The inset picture is obviously on the twin bodied set-up, and deservedly so!
Now - do bear in mind these are only the initial test results on an engine of unknown/unfamiliar build specification. It was essentially a 1380 mild-ish road unit that was available as a 'dyno donkey' (or 'Lester the Tester' as my mate Nick Swift would call it). As a base to work against, it was tested with the most common induction set-up used - a 1.75" HIF SU on a Mini Spares small-port ally inlet manifold and K&N cone filter. A medium bore LCB was used. No optimising of cam timing or other mechanical engine detail was undertaken. The actual peak power and torque are in the ballpark to what would be expected from an engine of this sort of build. Going to the Unibody injection didn't give a whole lot more torque - but that was expected since we ran this on the much larger ported C-AHT771 MSC ally inlet manifold - but it did give significantly more peak power and held on to the torque much longer at the top end. We were pretty impressed with that. However, when we bolted on the twin body injection, well. Flabbergasted is an understatement! We can't wait to get it on to a known, well-modified engine build! It will make a near-race spec engine feasible for road use.
One of the most impressive things is something that isn't shown on the graph - the fuelling. This is because this became such an important part of the jigsaw puzzle; we are not releasing detailed info on it. All I will say is that injector positioning proved absolute crucial. Having managed that, it became very apparent that the fuel injection system was giving far more efficient fuelling than anything else we'd tried. So that is better mpg but more BHP!
Next up-date in a week or so.