Had a really interesting Saturday on the Rolling Road, playing with the standard fuelling and ignition settings of MEMS1.9, with surprising and pleasing results!
Background
A little background: we all know the limitations of the earlier versions of modular engine management system (MEMS) fitted to our MGFs: MEMS1.9 and MEMS2J for the MPi and VVC respectively. For those who've been on the "scene" for more years than they care to remember, will remember that the only way that you could alter the maps for these cars was either to use a piggy-back chip (the SuperChips ICON-RACE was the chip of choice back then) or take it to someone who really knew what they were doing, de-solder the original chip, EEPROM flash a map on a new chip, and put this back into the case. I can't imagine why the latter never really caught on...
This has lead us all to think that these early versions of MEMS are completely inflexible and to tune your engine you need to junk it in favour of an alternative, programmable system.
What I found on Saturday doesn't entirely change this perspective, but for MEMS1.9 at least, there is another option
Within Testbook (T4; the original MG Rover diagnostic tool), there has always been the facility to alter ignition and fuelling settings. But T4 has never been a widely available tool for home use (dealer only) - and is frankly a PIA even if you do have one (three different leads, clunky interface on a long-obsolete version of Windows).
Fortunately, within recent times, there has been an emergence of aftermarket tools with some of the functionality of T4. In fact, one of the tools even offered the opportunity to alter the ignition and fuelling settings.
There is now another tool with this facility, the pscan - and as many of you know, I've been working with Philip to develop it's functionality. When I say "working with" I mainly mean badgering the poor guy so I can undertake diagnostics on my aging MG Rover (and now Land Rover) fleet of vehicles... Fortunately, Philip is very patient!
The tests
When Philip told me that his latest pscan update included the option to alter ignition and fuelling, I booked a rolling road session with Andy Bowden at Novatech in Slough (Dynodynamics test bed). Would changing the ignition and fuelling settings within MEMS1.9 actually do anything? How far can you push it? I figured that the rolling road was the best and safest place to see what was achievable.
Naturally, the test subject was Project Shed - a 1996 MGF with 1.8 MPi engine on MEMS1.9. It isn't quite standard: it has a 52mm throttle body breathing through a K&N panel filter with a cold air feed. The cams are R140 (Lotus Sport 135 cams, the same as those fitted to the TF135). An MGF Centre 4-2-1 manifold and a Trevor Taylor single exit exhaust complete the picture. As standard, the engine puts out approx. 134-135bhp - pretty much as you'd expect.
Baseline observations
Unfortunately, the graph above does not show the air:fuel mixture, but it was around 12:1 through much of the rev range. Peak power is typically made at 13:1, so what we saw is essentially in line with what everyone says: the standard K-series - and even my lightly modified, K-series with warmer cams - runs rich. If we can lean this mixture off, there will be power to be gained. We shall come to that later...
Changing the MEMS1.9 map settings
Next we connected up the pscan to the diagnostic port (easily accessed above the fuse board in a car with much of the dashboard cut away
) and set the ECU to its "default" settings. This made absolutely no difference whatsoever to the power output: pscan can't really tell you what the current settings are (a shame - not sure whether Philip can pull this data or not, but what you see on pscan is what you see on TestBook - so even the Rover tool doesn't tell you this rather important information). But in our case, it would seem that the car was in its factory default setting.Optimising the ignition advance
The next step was to retard the ignition. If the ignition setting is anywhere safe, you'd expect retarding the ignition to knock back engine power - and we did this to see where we were for the standard settings. Unsurprisingly, we found that the engine dropped power everywhere.
Within T4, you can advance or retard the ignition 5 degrees from standard. You don't want to advance the ignition too far - and generally those who remap engines tend to use as little advance as possible (and perhaps knock things back a bit once peak power is achieved) to maintain reliability.
First, we advanced the ignition 2 degrees. This netted us a +2bhp gain across the rev range. We decided to advance another 2 degrees. Perhaps 0.5bhp; we knocked back to 3 degrees - same output. On my engine, it would seem advancing the ignition 2-3 degrees gave us the best power output.
Optimising the fuelling
Now for fuelling.
Unfortunately we don't really know quite how these ignition or fuelling settings effect the standard map*: whether adding ignition or taking fuel does this over the whole map, over selected parts, or variably across the map we could not really ascertain. It "feels" like it is working uniformly across the whole map. The fuelling can be altered by 120 units either way - adding or removing. What those units are, we haven't worked out.
[* when we're talking about the map here, we are only talking about the "open loop" map - how the engine is feed spark and fuel when the accelerator pedal is depressed; the "closed loop" map, which adjusts spark and fuel in accordance to the Lambda reading when the engine is in steady state is, we believe, completely unaltered]
We started gingerly - taking out 10 units. This added a little power, took a little out of the air:fuel mixture. We decided to man-up and took 30units out the fuelling. This moved more decisively. 60 units out, still heading in the right direction. Then 70 units: too much! The air:fuel mixure at low rpm was far too lean, and the run was aborted at 3000rpm. Then 65 units out - the Goldilocks region it would seem: we still didn't reach the 13:1 sweet spot, and the standard map at peak power was still running a little too rich - but with 3 degrees advance and 65 units less fuel we saw a round 140bhp on the dyno. That's an extra 3 bhp by taking out the amount of fuel added! More fuel doesn't mean more power; more doesn't always equate to more
We went back to the ignition settings - and elected to take 0.5 degrees out of the advance we put in to give us this:
The proof-of-the-pudding is the driving - and driving back into London on a Saturday lunch time gives a mixture of motorways, main roads and urban roads with a mix of free-flowing traffic and congested stop-start frustration.
Driving impressions
We've added five horses, but have added a kangaroo.
I'll explain. The acceleration felt more instantaneous and smoother. I thought this was placebo effect, but if you scrutinize the torque curve, the fuel and ignition changes have flattened out the curve substantially - and it transitions from acceleration to steady state to deceleration very smoothly. The extra power is only just noticeable, but that is to be expected: the rule of thumb is that only gains of around 10bhp+ are generally really noticeable. The alteration in the low-mid-range torque is the most palpable: where typically there is a bit of a dip there is now none.
And the kangaroo? Only noticeable if, like me, you trundle along in very slow moving traffic in second gear and modulate engine speed with the accelerator with the engine turning at around 1500 rpm. The "re-map" only effects open throttle states, not when the engine is in "closed-loop" - and this remains the factory, full-fat-fuel default map - and consequently the transition from open- to closed-loop and back again isn't as smooth as it ought to be. Project Shed is about 150kg lighter than standard, amplifying this effect, but for the gains, I can live with the slight loss of this low-speed driving refinement. This is my race car remember.
Conclusions
The 5 bhp gain and the general improvement in the torque curve was very pleasing. It just cost me some rolling road time. And now I know the safe limits for this particular engine, I can tweak the settings at my leisure using a pscan. The improvement in power and general acceleration is pretty much exactly what you get from Mark Stacey's MEMS3 reprogramming - but without the option of putting in exhaust pops on overrun, and doesn't have that last word of very low-speed transition between maps that occurs as a consequence of tweaking just the "open-loop" table. But if you have the appropriate tool, you can adjust this out with time to give the best possible result for your taste.
What it isn't is a true re-map. Without doubt, with a fully re-programmable ECU you could probably massage a bit more out of the engine and maintain driving refinement. But that'll cost you a bunch more cash. In my case, it cost me a fun hour of rolling road time and the tool I already had (or in this case, actually Philip's as he came along to see whether the software and hardware performed as it should; it did). I think I'll take my other MEMS1.9 car for a bit of rolling road fettling too - and that would be my recommendation: play with the fuel and ignition settings as much as you like, but do establish the "safe limits" for your car's engine on a rolling road first. What worked on Project Shed may not be quite the same for yours.
