3.0 Litre Update

Yeah - I'm happy with what happens on the GTA - fuel squirted on all cylinders every two revs, spark every third of a rev and the dizzy sorts out where it goes to - just couldnt understand the statement about identifying individual cylinders from crankshaft rotation thats all. I guess it would work if you had 3 coil-packs each feeding both of the cylinders that could be firing at each third of a rev - one spark does the job, the other is "wasted".

*( for those who might want to learn here, having one missing tooth per engine rev gives the ECU enough information to know which cylinder is which - vital if you want to run coil packs/wasted spark. It does not need to know which cylinder is which if you run a distributor and batch fired injection - as we are here, so having a missing tooth for each cylinder pair is fine)

I "want to learn" but I cant see how you can positively identify a cylinder from the Crankshaft rotation - surely you can only narrow it down to one of two, given that each cyliner only fires once every two revolutions and you dont know which of those revolutions the engine is currently on. Dont you need a camshaft sensor in order to positively identify which particular cylinder is firing at any given time? I know its irrelevant to the GTA - just wanted to get all this stuff straight in my head.

Yes Mich, you are absolutely correct. You need a cam angle sensor to do fully sequential injection, or ignition - the best you can do is wasted spark. Without a cam sensor, it can never know exactly what cylinder it is working on..

Cheers for that DG.

On a completely different subject - is your engine management analyser available for hire at the moment? I'll be down in Chelmsford next Wed/Thu so could perhaps arrange to pick it up? That will be a larf when checking in for my flight home - "Has anyone given you anything to take onboard sir? - Yes, this box of wires and electronic gadgetry - What does it do? - Feck knows... I'll get the train shall I?"

Cheers for that DG.

On a completely different subject - is your engine management analyser available for hire at the moment? I'll be down in Chelmsford next Wed/Thu so could perhaps arrange to pick it up? That will be a larf when checking in for my flight home - "Has anyone given you anything to take onboard sir? - Yes, this box of wires and electronic gadgetry - What does it do? - Feck knows... I'll get the train shall I?"

Yep, no problem.

No, you don't fill in the boxes in the setup window for the teeth that are missing, so you are not directly telling it how many teeth there are (well you are - only the 'real' ones)

Yes you do. Fill in 60 teeth and three missing, it'll drop out that each division is 6 degrees. Tony's original setup was 36 teeth, 1 missing. The angles drop out at 10 degrees. Trust me on this..... (66 teeth is a bad example because the figures that drop out are rounded to the nearest degree on the display)

(for anyone not familiar with the Adaptronic software, you run a little wizard then it draws a table of all the teeth positions)

Why? If I tell the ecu its 10 degrees before TDC or 50 degrees before TDC, you can just adjust the map to suit. Doesn't mean its not going to run - the evidence of that is it does run!

I'm not arguing that it runs (or more accurately IDLE's) correctly. But I want to know HOW.

What you've said here is correct - I could set up the TT DeLorean telling the ECU is triggering at 30 degrees as long as you add 30 degrees to every figure in the ignition map.

But you don't REALLY want to do that, do you? You want to the figures to actually mean something rather than being an arbitary offset.

The fact is that at IDLE, Peter's figures are correct. I put his ignition map on the TT DeLorean which is set to idle at 13 degrees ( because there's a notch on the DeLorean timing cover at 13 ) and it's timed perfectly.

So why on an unmodifed flywheel does it require this magic figure of 108 to work?

NO, its 108 degrees of teeth! Because, like Ive said numerous times, just because a flywheel can have 66 teeth, doesnt mean they are going to be at regular angles

If they're not ar regular angles, it's because the ECU is using the irregularities to establish #1 in order to run coil packs. This engine does not. The SDS EFI Ecu uses four magnets on the flywheel and one of those is placed slightly offset from the one that's referenced to #1. This is not an argument that applies to this engine.

[quote]

The ecu is told there is 20 teeth per period, over a 108/9 degree section, = 5.45 degrees, which is correct. I can't explain it any simpler...

[quote]

Except that's not how the software is configured. You don't need to tell it the sector covered by each period, which is what you're suggesting, because it can work that out for itself based on number of cylinders. It needs to know the pattern and the position of the reset in relation to TDC ALL Ecu's do this!

What you're saying would probably be correct IF TDC was referenced directly by the reset. In other words if you want the ECU to run 10 degrees of advance, then fire just before it "sees" the second tooth after reset (5.45 x 2)

Back to the magic 108 figure, by your explanation, it's wrong and should be 109 because 20 x 5.45 = 109. A discrepancy of 1 degree would be visible at idle with a timing light, and it's bang on the money.

I "want to learn" but I cant see how you can positively identify a cylinder from the Crankshaft rotation - surely you can only narrow it down to one of two, given that each cyliner only fires once every two revolutions and you dont know which of those revolutions the engine is currently on. Dont you need a camshaft sensor in order to positively identify which particular cylinder is firing at any given time? I know its irrelevant to the GTA - just wanted to get all this stuff straight in my head.

Yes, you need a phase sensor (camshaft) to be able to run sequential injection or individual coil packs

The basic EFI setup is the GTA's. Run batch fired injection - which means all the injectors are fired simultaneously once very one or two revolutions (depending on the ECU, the Adaptronic fires every two), and run a distributor. The ECU only needs to know each TDC and there are three per revolution on a 6 cylinder engine. Which spark plug fires is governed by the rotor arm pointing at the right HT lead, so the ECU doesn't need to know engine position, only where TDC occurrs

Wasted spark is the next order of complexity where The ECU just needs to know where #1 is (which is paired with #5 on the even fire PRVs). It uses three coil packs and fires both paired cylinders at the same time. This has some benefits and some drawbacks. It can potentially light up a mis-fired cylinder on the exhaust stroke causing a back-fire, and it'll cause any digital tach timing light (like mine) to read double the actual revs. The Laguna PRV uses this and has a 60/1 trigger. Shame they were only ever autos....

The next order of complexity requires a phase sensor on the camshaft so the ECU knows where #1 is and when #1 is at firing stroke. You can then run sequential injection which primarily helps with fuel economy and stops the really loud ticking from the injectors in favour of 6 individual little ticks per rev.

Much of this stuff I learnt from reading the Electromitive Tec3 ECU manual. The best one I've ever read, shame the ECU is so bloody expensive!

I'm not arguing that it runs (or more accurately IDLE's) correctly

It works all through the rev range right up to where my limiter was set at 6300 and was smoother and more progressive than the standard ECU....it also showed 209@wheels on 1 bar of boost and 300lbs of torque.It does work!!!!!! Next time I speak to Jonathon I will get him to email you, he never batted an eyelid when I described the Renault teeth pattern and told him it couldnt be done, he told me it was no more difficult than many others he worked on.....lo and behold he had it running in less than a day.