Farting Around again
After many pleasurable miles driving my seven, my ear became tuned to a distinctive and annoying sound - pinking. This was occurring between 2000 and 3000RPM just as the engine was coming on cam. In order to allow the engine to idle and give a decent progression I have used a lot of advance at idle (14-16 degrees) with a tight advance curve that does nothing until about 2000RPM. The problem with an engine using 'big' cams is that the burn characteristics are very slow at idle and when off the cam, changing rapidly to very fast burn as soon as the engine harmonics improve (coming on cam). This means that if you get the idle advance right and progression right, the engine pinks like a bastard the moment it starts to come alive. Backing off the timing killed the idle and progression, causing the engine to nosedive when pulling away, not too good when first gear is as high as mine.
What was needed was an ignition advance system that was not simply determined by engine speed, the standard method uses weights, springs and centrifugal force to provide the advance with a vacuum operated advance for smooth idling. This is fairly crude and will only allow a simple linear advance curve. Most people disconnect the vacuum advance mechanism when tuning an engine because it is crude in operation, and when at idle (where it is most useful) it is not easy to get good vacuum characteristics when using sidedraught carbs which have one barrel per port. What you get is a fluctuating vacuum and therefore fluctuating timing, accompanied by a fluttering noise from the diaphragm in the distributor vacuum mechanism.
Pinking in an engine with 11.5 to 1 compression is at best irksome, at worst bloody frightening, having already experienced a 'happening' with the original Pinto lump, I was not keen to have the Cossie melt a piston or worse. Shit or bust time I thought, the way to go is mapped ignition, simply dial in the timing where and when you want it, no springs, weights etc to worry about. I set about evaluating the various offerings, MBE, WEBER-ALPHA, EMERALD etc. I even considered the prospect of building my own (fortunately I have the right skills available within my company). After looking at the costs of building and the inevitable bugs in hardware and software it didn't seem worth re-inventing the wheel so I contacted EMERALD about their M3D system which on paper looked quite good and competitively priced. I duly purchased an EMERALD M3D unit and set about reading all the bumph.
The M3D ignition system controls the advance from a datum point 66 degrees before TDC and then determines for a given engine speed and load when to fire the ignition. It uses a throttle potentiometer to determine throttle position that allows the advance to be tailored by engine speed and throttle load. The unit incorporates a shift light, hard and soft cut rev limiter and Tacho drive and telltale. If you add up the cost of a normal electronic ignition such as Lumenition, a shift light/revlimiter and a competition distributor, the M3D unit makes a lot of sense. It can be triggered from a Hall-effect distributor (ESC), Lumenition eye, or a std Sierra Magnetic reluctance distributor.
As the advance was going to be controlled electronically, the existing distributor advance mechanism was not only redundant but also a bloody nuisance. It is absolutely imperative that the distributor does not supply any advance centrifugally, and indeed the standard way to cope with this is to 'lock' the distributor advance mech. with a self-tapper or similar. I managed to track down a distributor from a Sierra which used the old Ford ESC system which also had mapped ignition. This distributor has no advance mechanism, a Hall effect trigger and was 35mm shorter than my old distributor, allowing a better fit under the inlet manifold!
The M3D unit was installed and a new wiring harness placed but not connected, ready for setting up on the rolling road. EMERALD is run by Dave Walker who is the Technical Editor of Cars and Car Conversions magazine, he recommended a set-up at Apollos rolling road at Erith near the Dartford crossing, with him doing the mapping. Fortunately when the day arrived, it was a fine one and I despatched myself down to Apollo, having fun goading a few hot-hatches (Golf VR6 etc) on the way.
The car was plonked on the rollers, and the new distributor fitted together with the new wiring harness. The distributor was aligned as per instructions, but the engine wasn't happy, it kept dying when the throttle was opened, finally dying completely. After much headscratching I spotted that the fuel pump had been turned off as a precaution whilst we were connecting the new harness, and it hadn't been turned back on (my fault)! The fuel pump was resurrected and all was well. The engine was then mapped at various speed sites and loads to give maximum torque with minimum advance.
Dave Walker is a very knowledgeable chap and we had a chat about Cossies vs. Zetecs etc. The best aspirated Cosworths he had seen before were HT Racing prepared and only made about 150BHP at the wheels causing him to be disparaging about their potential. After a short debate about expected power output, we both agreed that substance (E.G. spread of torque) was more important than a big number and that my engine was delivering shedloads of torque from as low as 2500RPM. A maximum power run was done to see what the engine was giving (at Daves insistence) and it recorded 190BHP at the wheels at 7000RPM that Dave reckoned would be around 225-230 at the flywheel. This was clearly a lot more than he expected and he was gobsmacked. Dave mentioned that the rollers at Apollos were the most conservative ones he used and were if anything inclined to under-reading. By this time I was a little nervous after hearing the engine at maximum load at 7500, isn't it strange how we are all happy to gun the engine when driving, but scared shitless when standing next to it being gunned on the rollers?- a humbling experience.
The advance is dialled in on the fly using a laptop PC connected to the M3D using an RS232/DIN cable (supplied with the M3D). A quick look at the map revealed what had been done, at idle the map is giving 35 degrees of advance, which gives a good idle, backing down to 12 degrees under full load. On partial loads the advance is around 20 degrees until the 2000-3000 range when it backs of to about 12-14 degrees, at 3000 when the engine starts to come on cam the advance ramps up to 24 degrees or so on full load, and about 30 degrees on light load, on full load above 4000RPM, advance is around 30-32 degrees increasing to 34-38 on partial throttle. This gives tremendous throttle response and no more pinking
During the drive home I could hear a metallic clinking (pinking - OH NO!) which foxed me somewhat, after stopping to investigate I discovered that it was the buckle on the rear of the passenger harness clinking on the roll-bar, what a relief. It's true what they say about gauge/instrument paranoia, the more instruments/gizmos you have, the more you worry, scanning them constantly for anomalies, one of my neighbours summed it up, ignorance is bliss all you need is a speedo.
The engine is now considerably more tractable, and will pull strongly from 1500RPM right up to 8200RPM in any gear, it is very quick on the throttle and requires much less initial throttle to maintain idle. Starting is easier and the mapping around the idle position keeps it steady at any temperature. I have had a play here and there with the mapping using a laptop and this gives tremendous satisfaction as things gradually improve. Obviously it is not possible to map every load site on the rollers, so those which were approximated are now being optimised.
I would recommend this conversion to anyone using a cam with more than 290 degrees duration, as it is engines of this sort of spec which would benefit most. But it is relatively expensive and not as visually exciting for your money as a pair of Webers, four into one manifold etc., but if you run an expensive engine and want the best from it, you could do a lot worse.