Possible Fiat L Jetronic FI upgrade?

[Marc]

Hi Guy,

Thought you would like this one...an airflow meter from a Datsun 280ZX fits perfect on a Fiat 2 liter Spider and works with 40/80 cams without tricking the ecu!

I just finished a conversion on a car, tried some other airflow meters and this one works well, its a Nippondenso model, looks just like the stock unit, plugs right in.
To make it work you must use 40/80 cams or more to increase airflow from the head to put it in proportion to the characterisics of the larger airflow meter. Since a 40/80 cam gives about a 40% airflow increase,its puts you in the range of a 2.8 litre engine which is about 40% more displacement. The only problem is getting the tires to bite.

[Guy Croft]

Marc, hi

Thanks, interesting, and if I had derived that solution I'd be quite excited too.

But, since I run the site I have to be very exact (ing) about everything. I would not recommend it (not yet, note!) as such without sight of a detailed and most thorough examination of the fuelling and torque curve, which of course is a rolling-road op. Moreover there as many 40/80 cams as the day is long, all with different lift integrals, and one simply cannot correlate their characteristic to any increase in airflow, far less an enhanced intake airflow of 40%.



GC

[TR-Spider]

That sounds interessting.

I well remember the difficulties I had integrating a BMW (Bosch) air flow meter some years ago. I had to use the Fiat internals (the resistor plate + spring) and increase fuel pressure...lots of full blast tests runs with a lambda meter to get the fuelling right.

Can we get more details, please.

Thomas

[sumplug]

Cams need the head to be gasflowed and the inlet and exhaust manifolds flowing as well. If you are going to increase airflow with the 280ZX Airflow meter, then the head needs to be able to use it, so it needs proper gasflowing with matching manifolds with as little CFM flow loss as possible. Then a good rolling road session to see improvements. Sounds promising .

Andy.

[Guy Croft]

Not the whole story Andy. I suppose what I'm going to write belongs more correctly in GC V/W but it is associated with Marc's idea to an extent.

1. Fitting camshafts of higher lift might only work better than std lift if the head flows enough air at that higher lift to make the change worthwhile. That needs testing on a flowbench. A std TC Fiat head airflow typically peaks at about 10mm lift - around the std lift figure. And I've found that on many heads. And of course it's obvious why, that's one of the factors in how cams are chosen. So putting a cam of higher lift (per se).

Just putting in a cam of lift much over 10mm won't assure more power on its own.

There are other factors:

2. Without changing the full lift characteristic of the cam the valve event of the inlet can be extended so that the valve is opened earlier before TDC and closed later after BDC. The former gives 'overlap' which generates superior cylinder scavenging and ram-charging in the inlet tract and the latter allows air to continue flowing even as the piston is going up the bore. Both enhance the torque.

3. The exhaust valve can be opened earlier potentially allowing a more powerful primary pulse and superior cylinder evacuation. This has the dual benefit of enhancing beneficial pressure waves inlet-exhaust and also reducing the pumping work the piston has to do on the ex stroke. Closing the ex later helps with scavenge by allowing incoming charge to purge the cylinder. Etc etc.

Both those changes will increase the lift at top dead centre of the cams and within limits can be built into a cam of same-peak lift-as-standard and can yield improved torque and power.

It must be understood that two profiles which appear the same in terms of timing and duration can have markedly different behaviour. I'm afraid when it comes to cams people will believe what they want to but a 300 deg cam (eg an often cited 40/80 cam) from the 'Acme cam grinders' and one of proven pedigree could be as different as, well 'chalk and cheese'. I happen to have reams of great dyno feedback on my own TC profiles, so I know exactly what they do and that's how I can sell them and rarely comment on other peoples' TC profiles - except C&B who are good in my book. There are some truly rotten TC profiles in circulation - mainly regrinds with weak lift integrals but I am not naming names.
When it comes to other engines my friends at Piper and Kent in the UK have profiles (Peugeot, Ford Vauxhall eg) whose performance I can only describe as 'proven and stunning'. Those guys really know what they are doing.

Make a careful study of the graph.

Those graphs are cam degrees vs lift in inches with flat-follower and zero clearance. The (so-called and much copied 40/80) Pittatore cam (a good albeit mild StII profile as it happens) seems to have longer duration than the GC cam. From my data I can tell you that if set at 0.016" clearance the true duration of the 77 is 312 deg and the 3A 300 degrees. So in fact, the 77 might look to some like a more powerful profile. It's not, nowhere near.

For argument sake let's suppose the cams are both set with 0.020" running clearance. From start of the blue line to contact of cam with tappet at 105 the cam isn't acting on the valve at all. Then the cam goes into the opening ramp phase that lasts from a full 5 degrees of cam rotation 10 crank degrees. The 3A if set at 20 thou clearance pretty well just goes straight into the lift phase. I don't set them at 20 thou as it happens. So at 16 thou, harsh ramp? Not a bit of it. You cannot have a powerful competition cam with long ramps, there isn't time for that. Some are even more aggresive. 3A is my top selling cam and I've never had a ramp wear issue with it. So you can see that a generous proportion of the 'duration' of the 77 and any cam - come to that - can easily be eaten up by a completely wasted and non-productive phase - the ramp.

From my data I can tell you that if set at 0.016" clearance the true duration of the 77 is 312 deg and the 3A 300 degrees. So in fact, the 77 might look to some like a more powerful profile. it's not, nowhere near.

Another important feature that is rarely discussed is the dwell at full lift. The 77 has 6 cam degrees and the 3A 10 degrees. May not sound much but a valve at full lift can flow a lot of air in 4 degrees of crank rotation.
Take a look at the lift rates too. The 3A from contact to 2mm valve lift takes 13 deg of cam rotation. The 77 takes 23 degrees.

And remember above all that both these cams are billet not regrind. The incremental lift determines how much air can flow thru the valve region under the influence of the net cylinder-to-inlet port pressure differential at any given time and the acceleration from start-lift to full-lift largely governs how effective the cam is going to be on generating those vital cylinder-filling pressure pulses in the inlet tract. The bigger the lift integral (measured in inch-degrees) is a measure of this. For equal full lift the cam with the bigger area under the curve is always going to be more powerful. Here the 3A has true integral of 39.8 in-deg and the 77 34.4 in-deg, but I have integral data on some regrinds where it is as low as 28in-deg even for a grind supposedly of '300 deg' duration and 10.5mm lift. Might sell but only on paper and quite useless in fact!

Now the cam characteristic could well have a bearing on the behaviour of the Bosch L Jetronic or other fitted airflow meter well beyond the issue of whether the meter is calibrated to deliver enough fuel and that is the response of the unit to the much-more-powerful pressure pulses (pos and neg) that emanate from the valve region with more aggressive cams. And make no mistake, powerful they are! I'm afraid I don't know enough about these units to be able to comment on their damping and transient response but I would certainly be concerned enough to want to investigate it.

GC

[sumplug]

Guy.
Can this last post be put in your engine build area please? I think it answers a lot of questions people have, at choosing a cam for their engine.

Andy.

[Guy Croft]

Sure Andy, it needs a lot adding to it but I'll post it as-is for now, thanks for the 'heads up',

GC

[Marc]

Hi Guy,
I understand your position ,as I am the tech advisor for FLU's web site with my Downshift forum.Every word is scrutinized!Its not easy,I admire you abilities and knowledge.
After I tried this,I had to tell you,and see what you thought about it.The AFM seems to be the weak link in terms of airflow(size) .What is your opinion about the throttle body size on a stock 2ltr.
By the way people in the club regard you as the worlds foremost expert and your opinion is Gospel.You deserve this praise,and I feel the same,especially trying to moderate the FLU website makes me even more impressed at what you do.
All I can tell you is I giggle everytime I mash down on the throttle.The nippondenso unit looks the same as a bosch unit size wise,I beleive its internally electronically calibrated for the bigger engine.Yes, I stumbled across this,but technically I need to fully understand what makes this system work so I can convey this info to everyone Regards ,Marc

[Guy Croft]

Marc, hi

it's kind of you to write in that way and I do appreciate it.

Insofar as how the L Jet works I could never describe its mode of operation as well as Charles Probst SAE does in his 'Bosch Fuel Injection and Engine Management' (Bentley ISBN 0 - 8376 - 0300 -5) But I can tell you one or two pointers that could be pertinent.

Let's assume your engine is taking in more air because it's been tuned up by one means or another. The worry is that it will not respond with stock L Jetronic because the fuelling will be weak. This is of course exactly what can happen. The air flow sensor (flap valve) is fully open at max torque because this is where the airflow is highest and the region where most fuel is required. Above peak torque it is shorter because although the engine is cycling faster the actual airflow is lower. This is due to volumetric efficiency dropping as the engine revs beyond peak torque - all engines do this.
Up to the point where the valve is full open the fuelling is adjusted by the ecu in accordance with the position-related signal from the flow meter. The pulse width (time) of the injection is longest at peak torque.

So what you might have achieved by fitting a flow meter from a bigger engine is effectively tell the 2liter engine's ecu that the engine is flowing more air at peak torque, whereas what's actually happening is that the new flowmeter is calibrated differently, and sending a signal that makes the peak torque pulse width longer. It is possible that you have got the match right.

As the engine with the std flowmeter goes past peak torque the fuelling is controlled by the ecu alone. The injector pulse width (time) is reduced because less fuel per intake cycle is needed, but the injector has to cycle a lot faster because the engine speed has gone up. It can be the case that even though the injector pulse width is shorter the thing has to open and close so fast because of the high crank speed that it barely has time to close before it opens again. The danger here is obvious. I don't know how much spare capacity there is in the std injectors. A rather brief pulse width might be fine on a standard engine but if the pulse width is very short there may be insufficient fuel delivered to satisfy the extra demand for a tuned engine. However if you know that the fuelling and power is also good at high rpm then maybe the injectors are big enough to cope.

I note your comment about increased torque (wheelspin) and this suggests that the engine is working very well there. I don't know what peak rpm your motor is pulling though. It may be that the new air flowmeter (being from a bigger engine) is continuing to signal to the ecu at higher engine speeds and there is sufficient injector capacity in hand. Maybe it¢ž¢s covering the whole upper speed range.

As long as the flap valve doesn't open fully it's going to effectively try and meter the fuel. However if it reaches full open the engine could go lean, it depends on the injector size, fuel pressure and pulse width. My main concern might be that the top-end is leaning out and I certainly suggest you get a wide-band Lambda installed to determine this. It is easier to pick up an improvement of peak torque than it is to discover whether the engine is going lean at high rpm ¢‚¬Å“ or over-rich at low rpm..


The other reason why the new flowmeter might be better is that it is a stiffer or better-damped unit and it is unaffected by the strong pressure waves in the tuned engine's inlet tract. I am not too sure on that latter aspect - so if anyone wants to comment feel free.

Is it that the replacement 280Z flowmeter is bigger in area at narrowest section and that is why the engine is more powerful? I cannot say for sure because I don't know the flowrate of either. It's possible I guess, although that's not to say that simply bolting on any bigger unit is a good thing per-se. The mechanism of the new air sensor flap would in any case have to tell the ecu accurately 'this is amount of air X' so that the ecu can inject a matching amount of fuel.


GC

(revision_01)

[chicken]

Bosch licensed Denso to clone the system.

Changing an AFM can change the signal to the ECU of course. But there is little likelyhood that you will be able to select the correct curve sans magic.

The "Box" may have a wider channel but unless you reach the CFM limit of the stock AFM there is no genuine advantage to this.

The "Flap" may have a greater surface but that increases its resistance to the incoming mass of air. The air mass passing through the channel must now first overcome a larger obstacle. Then have you gained or lost performance from the swap?

That cannot be answered yet because the spring tension may be lesser or greater. It may also have a different rate of increase as specified by engineering. The resulting voltage can vary quite widely.

Most engines of an exact same displacement often by the same manufacturer and sometimes only differing by market have very different power curves. These aforementioned differences only in the AFM affect quite a lot. A HUGE range of air is perfectly matched to a small range of voltage to control milliseconds of time on injectors measured by cc's delivered. That is not an insignificant sentence.

So does that mean your car will blow up? Not likely. The engineers know how much faith to put in the end user and have mapped the system to run with a gigantic margin for unforeseeable error. Even if its a deliberate error. (*This is somewhat less true now where emission and consumer demands have forced systems which calculate fractions of a degree from Crank Sensors mounted in Bearing Caps for accuracy).

The velocity of the air between entering the AFM and your valve is a whole 'nother tricky issue that changes at least once at the throttle plate. How great you decide to have that velocity between the AFM and the T/B is like choosing your chokes. Where do you drive your motor most of the time? Compromises have to be made again and again.

(edited for punctuation)

[per]

I have the 122hp Argenta 120ie 1985, which I have bought Pittatore cams for. I also plan to insert the pistons from a Regatta 100S/Lancia Prisma 1600.
I have many questions for my future conversion, but I don't have the cam details in front of me now. They do have a 10,1mm inlet lift though.

But as for the AFM, isn't a surgery of the existing Bosch AFM a possible way of raising the fuel? I've been having some other issues with this car in the past years, and one trick was opening it up and loose/tense the spring a notch.
Loosening it a notch or two, would'nt that be enough? (considering you also measure the lambda of course)

[vandor]

Thought you would like this one...an airflow meter from a Datsun 280ZX fits perfect on a Fiat 2 liter Spider and works with 40/80 cams without tricking the ecu!

My previous 124 Spider ran an injected 2000 engine with 9.8:1 CR, Pittatore 40-80 cams, and Bosch Fuel injection. The injection was stock, except for the air filter box. As measured with a lambda meter, the mixture stayed rich up to 7000 rpm at WOT.
It's not worth fooling with L-jet, if you want real gains use a fully programmable system, they are very affordable for what they offer.

Csaba

[per]

Thanks a lot Vandor, that's reassuring. I too will use the 40/80 ones, with 10.35 lift.
Was just about to hoover round for old 280ZX parts.. Will leave it as it is and then do lambda checks.
If anything, what you say would mean only small adjustments are necessary.